Applied Evidence

Evidence supports what family physicians know to be true: Knee pain is an exceedingly common presenting problem in the primary care office. Estimates of lifetime incidence reach as high as 54%,¹ and the prevalence of knee pain in the general population is increasing.² Knee disability can result from acute or traumatic injuries as well as chronic, degenerative conditions such as osteoarthritis (OA). The decision to pursue orthopedic consultation for a particular injury or painful knee condition can be challenging. To address this, we highlight specific knee diagnoses known to cause pain, with the aim of describing which conditions likely will necessitate surgical consultation—and which won’t.

Acute or nondegenerative knee injuries and pain

Acute knee injuries range in severity from simple contusions and sprains to high-energy, traumatic injuries with resulting joint instability and potential neurovascular compromise. While conservative treatment often is successful for many simple injuries, surgical management—sometimes urgently or emergently—is needed in other cases, as will be detailed shortly.

Neurovascular injury associated with knee dislocations

Acute neurovascular injuries often require emergent surgical intervention. Although rare, tibiofemoral (knee) dislocations pose a significant challenge to the clinician in both diagnosis and management. The reported frequency of popliteal artery injury or rupture following a dislocation varies widely, with rates ranging from 5% to 64%, according to older studies; more recent data, however, suggest the rate is actually as low as 3.3%.³ Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified. Identifying a knee dislocation can prove challenging, as spontaneous joint reduction occurs in as many as 50% of cases, potentially shrouding the severity of the injury on initial evaluation.⁴

Immediate immobilization and emergency department transport for monitoring, orthopedics consultation, and vascular ­studies or vascular surgery consultation is recommended in the case of a suspected knee dislocation. In one cross-sectional cohort study, the surgical management of knee ­dislocations yielded favorable outcomes in > 75% of cases.⁵

Tibial plateau fracture

This fracture often occurs as a result of high-energy trauma, such as contact sports or motor vehicle accidents, and is characterized by a proximal tibial fracture line with extension to the articular surface. X-rays often are sufficient for initial diagnosis. Computed tomo­graphy can help rule out a fracture line when clinical suspicion is high and x-rays are nondiagnostic. As noted earlier, any suggestion of neurovascular compromise on physical exam requires an emergent orthopedic surgeon consultation for a possible displaced and unstable (or more complex) injury (FIGURE 1).^6-8

Nondisplaced tibial plateau fractures without supraphysiologic ligamentous laxity on valgus or varus stress testing can be managed safely with protection and early mobilization, gradual progression of weight-bearing, and serial x-rays to ensure fracture healing and stability. Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise. Suspicion of these complications is raised by distal neuropathic symptoms of paresthesia or skin anesthesia, progressively worsening pain distal to the knee, or vascular signs of pallor, delayed or lost capillary refill, or decreased or absent distal pulses.

Gross joint instability identified by positive valgus or varus stress testing, positive anterior or posterior drawer testing, or patient inability to tolerate these maneuvers due to pain similarly should raise suspicion for a more significant fracture at risk for concurrent neurovascular injury. Acute compartment syndrome also is a known complication of tibial plateau fractures and similarly requires emergent operative management. Urgent surgical consultation is recommended for fractures with displaced fracture fragments, tibial articular surface step-off or depression, fractures with concurrent joint laxity, or medial plateau fractures.^6-8

Continue to: Patella fractures

Patella fractures

These fractures occur as a direct blow to the front of the knee, such as falling forward onto a hard surface, or indirectly due to a sudden extreme eccentric contraction of the quadriceps muscle. Nondisplaced fractures with an intact knee extension mechanism, which is examined via a supine straight-leg raise or seated knee extension, are managed with weight-bearing as tolerated in strict immobilization in full extension for 4 to 6 weeks, with active range-of-motion and isometric quadriceps exercises beginning in 1 to 2 weeks. Serial x-rays also are obtained to ensure fracture displacement does not occur during the rehabilitation process.⁹

High-quality evidence guiding follow-up care and comparing outcomes of surgical and nonsurgical management of patella fractures is lacking, and studies comparing different surgical techniques are of lower methodological quality.¹⁰ Nevertheless, displaced or comminuted patellar fractures are referred urgently to orthopedic surgical care for fixation, as are those with concurrent loose bodies, chondral surface injuries or articular step-off, or osteochondral fractures.⁹ Inability to perform a straight-leg raise (ie, clinical loss of the knee extension mechanism) suggests a fracture under tension that likely also requires surgical fixation for successful recovery. Neurovascular injuries are unlikely in most patellar fractures but would require emergent surgical consultation.⁹

Ligamentous injury

Tibiofemoral joint laxity occurs as a result of ligamentous injury, with or without tibial plateau fracture. The anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), medial collateral ligament (MCL), and lateral collateral ligament (LCL) comprise the 4 main ligaments of the knee. The ACL resists anterior tibial translation and rotational forces, while the PCL resists posterior tibial translation. The MCL and LCL resist valgus and varus stress, respectively.

Ligament injuries are classified as Grades 1 to 3¹¹:

• Grade 1 sprains. The ligament is stretched, but there is no macroscopic tearing; joint stability is maintained.
• Grade 2 sprains. There are partial macroscopic ligament tears. There is joint laxity due to the partial loss of the ligament’s structural integrity.
• Grade 3 sprains. The ligament is fully avulsed or ruptured with resultant gross joint instability.

Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified.

The decision to pursue surgical repair of a knee sprain depends heavily on the likelihood of keeping or regaining and maintaining functional joint stability during the injury recovery and postinjury time periods. Injuries that do not result in joint instability or injuries with a high likelihood of returning to a stable state with conservative measures often do not require surgical intervention.

Continue to: ACL tears

ACL tears occur most commonly via a noncontact event, as when an individual plants their foot and suddenly changes direction during sport or other physical activity. Treatment hinges on patient activity levels and participation in sports. Patients who do not plan to engage in athletic movements (that require changes in direction or planting and twisting) and who otherwise maintain satisfactory joint stability during activities of daily living may elect to defer or even altogether avoid surgical reconstruction of isolated ACL tears. One pair of studies demonstrated equivalent outcomes in surgical and nonsurgical management in 121 young, nonelite athletes at 2- and 5-year follow-up, although the crossover from the nonsurgical to surgical groups was high.^12,13 Athletes who regain satisfactory function and stability nonoperatively can defer surgical intervention. However, the majority of active patients and athletes will require surgical ACL reconstruction to return to pre-injury functional levels.¹⁴

PCL sprains occur as a result of sudden posteriorly directed force on the tibia, such as when the knee is hyperextended or a patient falls directly onto a flexed knee. Patients with Grade 1 and 2 isolated sprains generally will recover with conservative care, as will patients with some Grade 3 complete tears that do not fully compromise joint stability. However, high-grade PCL injuries often are comorbid with posterolateral corner or other injuries, leading to a higher likelihood of joint instability and thus the need for surgical intervention for the best chance at an optimal outcome.¹⁵

MCL sprain. Surgical management is not required in an isolated Grade 1 or 2 MCL sprain, as the hallmarks of recovery—return of joint stability, knee strength and range of motion, and pain ­reduction—can be achieved successfully with conservative management. Isolated Grade 3 MCL sprains are also successfully managed nonoperatively¹⁶ except in specific cases, such as a concurrent large avulsion fracture.¹⁷

LCL sprain. Similarly, isolated Grade 1 and 2 LCL sprains generally do not require surgical intervention. However, Grade 3 LCL injuries usually do, as persistent joint instability and poor functional outcomes are more common with nonsurgical management.^18-20 Additionally, high-grade LCL injuries frequently manifest with comorbid meniscus injuries or sprains of the posterolateral corner of the knee, a complex anatomic structure that provides both static and dynamic tibiofemoral joint stability. Surgical repair or reconstruction of the posterolateral corner frequently is necessary for optimal functional outcomes.²¹

Multiligamentous sprains frequently lead to gross joint instability and necessitate orthopedic surgeon consultation to determine the best treatment plan; this should be done emergently if neurovascular compromise is suspected. A common injury combination is simultaneous ACL and MCL sprains with or without meniscus injury. In these cases, some surgeons will choose to defer ACL reconstruction until after MCL healing is achieved. This allows the patient to regain valgus stability of the joint prior to performing ACL reconstruction to regain rotational and anterior stability.²⁰

Continue to: Patellar dislocations

Patellar dislocations represent a relatively common knee injury in young active patients, often occurring in a noncontact fashion when a valgus force is applied to an externally rotated and planted lower leg. A chief risk factor for a patellar dislocation is a history of prior dislocation. If rehabilitation following a dislocation is insufficient to regain patellofemoral joint stability, or if certain risk factors for recurrent dislocation are present, surgical intervention, such as medial patellofemoral ligament (MPFL) reconstruction or tibial tubercle transfer, is considered.²² A systematic review concluded that MPFL reconstruction following a first-time dislocation yielded lower redislocation rates of 7% compared to 30% with nonoperative treatment.²³

Major tendon rupture

Patellar tendon ruptures occur when a sudden eccentric force is applied to the knee, such as when landing from a jump with the knee flexed. Patellar tendon ruptures frequently are clinically apparent, with patients demonstrating a high-riding patella and loss of active knee extension. Quadriceps tendon ruptures often result from a similar injury mechanism in older patients, with a similar loss of active knee extension and a palpable gap superior to the patella.²⁴

Partial tears in patients who can maintain full extension of the knee against gravity are treated nonoperatively, but early surgical repair is indicated for complete quadriceps or patellar tendon ruptures to achieve optimal outcomes. Prompt diagnosis and treatment are critical, as repair delayed beyond 1 to 2 weeks postinjury is associated with worse outcomes.^25-28

Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise.

Even with prompt treatment, return to sport is not guaranteed. According to a recent systematic review, athletes returned to play 88.9% and 89.8% of the time following patellar and quadriceps tendon repairs, respectively. However, returning to the same level of play was less common and achieved 80.8% (patellar tendon repair) and 70% (quadriceps tendon repair) of the time. Return-to-work rates were higher, at 96% for both surgical treatments.²⁹

Locked knee and acute meniscus tears in younger patients

In some acute knee injuries, meniscus tears, loose cartilage bodies or osteochondral defects, or other internal structures can become interposed between the femoral and tibial surfaces, preventing both active and passive knee extension. Such injuries are often severely painful and functionally debilitating. While manipulation under anesthesia can acutely restore joint function,³⁰ diagnostic and therapeutic arthroscopy often is pursued for definitive treatment.³¹ Compared to the gold standard of diagnostic arthroscopy, preoperative magnetic resonance imaging (MRI) carries positive and negative predictive values of 85% and 77%, respectively, in identifying or ruling out the anatomic structure responsible for a locked knee. ³² As such, MRI has been proposed as a method to avoid performing arthroscopy on a patient with a “pseudo-locked” knee, or loss of range of motion due to pain but without a true mechanical block.³²

Continue to: Depending on the location...

Depending on the location, size, and shape of an acute meniscus tear in younger patients, surgical repair may be an option to preserve long-term joint function. In one case series of patients younger than 20 years, 62% of meniscus repairs yielded good outcomes after a mean follow-up period of 16.8 years.³³

Osteochondritis dissecans

Osteochondritis dissecans is characterized by subchondral bone osteonecrosis that most often occurs in pediatric patients, potentially causing the separation of a fragment of articular cartilage and subchondral bone into the joint space (FIGURE 2). In early stages, nonoperative treatment consisting of prolonged rest followed by physical therapy to gradually return to activity is recommended to prevent small, low-grade lesions from progressing to unstable or separated fragments. Arthroscopy, which consists of microfracture or other surgical resurfacing techniques to restore joint integrity, is pursued in more advanced cases of unstable or separated fragments.

High-quality data guiding the management of osteochondritis dissecans are lacking, and these recommendations are based on consensus guidelines.³⁴

Septic arthritis

Septic arthritis is a medical emergency caused by the hematogenous spread of microorganisms, most often staphylococci and streptococci species. Less commonly, it arises from direct inoculation through an open wound or, rarely, iatrogenically following a joint injection procedure. Clinical signs of septic arthritis include joint pain, joint swelling, and fever. Passive range of motion of the joint is often severely painful. Synovial fluid studies consistent with septic arthritis include an elevated white blood cell count greater than 25,000/mcL with polymorphonuclear cell predominance.³⁵ The knee accounts for more than 50% of septic arthritis cases, and surgical drainage usually is required to achieve infection source control and decrease morbidity and mortality due to destruction of articular cartilage when treatment is delayed.³⁶

Chronic knee injuries and pain

Surgical intervention for chronic knee injuries and pain generally is considered when patients demonstrate significant functional impairment and persistent symptoms despite pursuing numerous nonsurgical treatment options. A significant portion of chronic knee pain is due to degenerative processes such as OA or meniscus injuries, or tears without a history of trauma that do not cause locking of the knee. Treatments for degenerative knee pain include supervised exercise, physical therapy, bracing, offloading with a cane or other equipment, topical or oral ­anti-inflammatories or analgesics, and injectable therapies such as intra-articular ­corticosteroids.³⁷

Continue to: Other common causes...

Other common causes of chronic knee pain include chronic tendinopathy or biomechanical syndromes such as patellofemoral pain syndrome or iliotibial band syndrome. Surgical treatment of these conditions is pursued in select cases and only after exhausting nonoperative treatment programs, as recommended by international consensus statements,³⁸ societal guidelines,³⁹ and expert opinion.⁴⁰ High-quality data on the effectiveness, or ineffectiveness, of surgical intervention for these conditions are lacking.

A chief risk factor for a patellar dislocation is a history of prior dislocation.

Despite being one of the most commonly performed surgical procedures in the United States,⁴¹ arthroscopic partial meniscectomy treatment of degenerative meniscus tears does not lead to improved outcomes compared to nonsurgical management, according to multiple recent studies.^42-45 Evidence does not support routine arthroscopic intervention for degenerative meniscus tears or OA,⁴² and recent guidelines recommend against it⁴⁶ or to pursue it only after nonsurgical treatments have failed.³⁷

Surgical management of degenerative knee conditions generally consists of partial or total arthroplasty and is similarly considered after failure of conservative measures. Appropriate use criteria that account for multiple clinical and patient factors are used to enhance patient selection for the procedure.⁴⁷

Takeaways

Primary care clinicians will treat patients sustaining knee injuries and see many patients with knee pain in the outpatient setting. Treatment options vary considerably depending on the underlying diagnosis and resulting functional losses. Several categories of clinical presentation, including neurovascular injury, unstable or displaced fractures, joint instability, major tendon rupture, significant mechanical symptoms such as a locked knee, certain osteochondral injuries, and septic arthritis, likely or almost always warrant surgical consultation ­(TABLE^3-10,12-36). Occasionally, as in the case of neurovascular injury or septic arthritis, such consultation should be emergent.

CORRESPONDENCE
David M. Siebert, MD, Sports Medicine Center at Husky Stadium, 3800 Montlake Boulevard NE, Seattle, WA 98195; siebert@uw.edu

Evidence supports what family physicians know to be true: Knee pain is an exceedingly common presenting problem in the primary care office. Estimates of lifetime incidence reach as high as 54%,¹ and the prevalence of knee pain in the general population is increasing.² Knee disability can result from acute or traumatic injuries as well as chronic, degenerative conditions such as osteoarthritis (OA). The decision to pursue orthopedic consultation for a particular injury or painful knee condition can be challenging. To address this, we highlight specific knee diagnoses known to cause pain, with the aim of describing which conditions likely will necessitate surgical consultation—and which won’t.

Acute or nondegenerative knee injuries and pain

Acute knee injuries range in severity from simple contusions and sprains to high-energy, traumatic injuries with resulting joint instability and potential neurovascular compromise. While conservative treatment often is successful for many simple injuries, surgical management—sometimes urgently or emergently—is needed in other cases, as will be detailed shortly.

Neurovascular injury associated with knee dislocations

Acute neurovascular injuries often require emergent surgical intervention. Although rare, tibiofemoral (knee) dislocations pose a significant challenge to the clinician in both diagnosis and management. The reported frequency of popliteal artery injury or rupture following a dislocation varies widely, with rates ranging from 5% to 64%, according to older studies; more recent data, however, suggest the rate is actually as low as 3.3%.³ Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified. Identifying a knee dislocation can prove challenging, as spontaneous joint reduction occurs in as many as 50% of cases, potentially shrouding the severity of the injury on initial evaluation.⁴

Immediate immobilization and emergency department transport for monitoring, orthopedics consultation, and vascular ­studies or vascular surgery consultation is recommended in the case of a suspected knee dislocation. In one cross-sectional cohort study, the surgical management of knee ­dislocations yielded favorable outcomes in > 75% of cases.⁵

Tibial plateau fracture

This fracture often occurs as a result of high-energy trauma, such as contact sports or motor vehicle accidents, and is characterized by a proximal tibial fracture line with extension to the articular surface. X-rays often are sufficient for initial diagnosis. Computed tomo­graphy can help rule out a fracture line when clinical suspicion is high and x-rays are nondiagnostic. As noted earlier, any suggestion of neurovascular compromise on physical exam requires an emergent orthopedic surgeon consultation for a possible displaced and unstable (or more complex) injury (FIGURE 1).^6-8

Nondisplaced tibial plateau fractures without supraphysiologic ligamentous laxity on valgus or varus stress testing can be managed safely with protection and early mobilization, gradual progression of weight-bearing, and serial x-rays to ensure fracture healing and stability. Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise. Suspicion of these complications is raised by distal neuropathic symptoms of paresthesia or skin anesthesia, progressively worsening pain distal to the knee, or vascular signs of pallor, delayed or lost capillary refill, or decreased or absent distal pulses.

Gross joint instability identified by positive valgus or varus stress testing, positive anterior or posterior drawer testing, or patient inability to tolerate these maneuvers due to pain similarly should raise suspicion for a more significant fracture at risk for concurrent neurovascular injury. Acute compartment syndrome also is a known complication of tibial plateau fractures and similarly requires emergent operative management. Urgent surgical consultation is recommended for fractures with displaced fracture fragments, tibial articular surface step-off or depression, fractures with concurrent joint laxity, or medial plateau fractures.^6-8

Continue to: Patella fractures

Patella fractures

These fractures occur as a direct blow to the front of the knee, such as falling forward onto a hard surface, or indirectly due to a sudden extreme eccentric contraction of the quadriceps muscle. Nondisplaced fractures with an intact knee extension mechanism, which is examined via a supine straight-leg raise or seated knee extension, are managed with weight-bearing as tolerated in strict immobilization in full extension for 4 to 6 weeks, with active range-of-motion and isometric quadriceps exercises beginning in 1 to 2 weeks. Serial x-rays also are obtained to ensure fracture displacement does not occur during the rehabilitation process.⁹

High-quality evidence guiding follow-up care and comparing outcomes of surgical and nonsurgical management of patella fractures is lacking, and studies comparing different surgical techniques are of lower methodological quality.¹⁰ Nevertheless, displaced or comminuted patellar fractures are referred urgently to orthopedic surgical care for fixation, as are those with concurrent loose bodies, chondral surface injuries or articular step-off, or osteochondral fractures.⁹ Inability to perform a straight-leg raise (ie, clinical loss of the knee extension mechanism) suggests a fracture under tension that likely also requires surgical fixation for successful recovery. Neurovascular injuries are unlikely in most patellar fractures but would require emergent surgical consultation.⁹

Ligamentous injury

Tibiofemoral joint laxity occurs as a result of ligamentous injury, with or without tibial plateau fracture. The anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), medial collateral ligament (MCL), and lateral collateral ligament (LCL) comprise the 4 main ligaments of the knee. The ACL resists anterior tibial translation and rotational forces, while the PCL resists posterior tibial translation. The MCL and LCL resist valgus and varus stress, respectively.

Ligament injuries are classified as Grades 1 to 3¹¹:

• Grade 1 sprains. The ligament is stretched, but there is no macroscopic tearing; joint stability is maintained.
• Grade 2 sprains. There are partial macroscopic ligament tears. There is joint laxity due to the partial loss of the ligament’s structural integrity.
• Grade 3 sprains. The ligament is fully avulsed or ruptured with resultant gross joint instability.

Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified.

The decision to pursue surgical repair of a knee sprain depends heavily on the likelihood of keeping or regaining and maintaining functional joint stability during the injury recovery and postinjury time periods. Injuries that do not result in joint instability or injuries with a high likelihood of returning to a stable state with conservative measures often do not require surgical intervention.

Continue to: ACL tears

ACL tears occur most commonly via a noncontact event, as when an individual plants their foot and suddenly changes direction during sport or other physical activity. Treatment hinges on patient activity levels and participation in sports. Patients who do not plan to engage in athletic movements (that require changes in direction or planting and twisting) and who otherwise maintain satisfactory joint stability during activities of daily living may elect to defer or even altogether avoid surgical reconstruction of isolated ACL tears. One pair of studies demonstrated equivalent outcomes in surgical and nonsurgical management in 121 young, nonelite athletes at 2- and 5-year follow-up, although the crossover from the nonsurgical to surgical groups was high.^12,13 Athletes who regain satisfactory function and stability nonoperatively can defer surgical intervention. However, the majority of active patients and athletes will require surgical ACL reconstruction to return to pre-injury functional levels.¹⁴

PCL sprains occur as a result of sudden posteriorly directed force on the tibia, such as when the knee is hyperextended or a patient falls directly onto a flexed knee. Patients with Grade 1 and 2 isolated sprains generally will recover with conservative care, as will patients with some Grade 3 complete tears that do not fully compromise joint stability. However, high-grade PCL injuries often are comorbid with posterolateral corner or other injuries, leading to a higher likelihood of joint instability and thus the need for surgical intervention for the best chance at an optimal outcome.¹⁵

MCL sprain. Surgical management is not required in an isolated Grade 1 or 2 MCL sprain, as the hallmarks of recovery—return of joint stability, knee strength and range of motion, and pain ­reduction—can be achieved successfully with conservative management. Isolated Grade 3 MCL sprains are also successfully managed nonoperatively¹⁶ except in specific cases, such as a concurrent large avulsion fracture.¹⁷

LCL sprain. Similarly, isolated Grade 1 and 2 LCL sprains generally do not require surgical intervention. However, Grade 3 LCL injuries usually do, as persistent joint instability and poor functional outcomes are more common with nonsurgical management.^18-20 Additionally, high-grade LCL injuries frequently manifest with comorbid meniscus injuries or sprains of the posterolateral corner of the knee, a complex anatomic structure that provides both static and dynamic tibiofemoral joint stability. Surgical repair or reconstruction of the posterolateral corner frequently is necessary for optimal functional outcomes.²¹

Multiligamentous sprains frequently lead to gross joint instability and necessitate orthopedic surgeon consultation to determine the best treatment plan; this should be done emergently if neurovascular compromise is suspected. A common injury combination is simultaneous ACL and MCL sprains with or without meniscus injury. In these cases, some surgeons will choose to defer ACL reconstruction until after MCL healing is achieved. This allows the patient to regain valgus stability of the joint prior to performing ACL reconstruction to regain rotational and anterior stability.²⁰

Continue to: Patellar dislocations

Patellar dislocations represent a relatively common knee injury in young active patients, often occurring in a noncontact fashion when a valgus force is applied to an externally rotated and planted lower leg. A chief risk factor for a patellar dislocation is a history of prior dislocation. If rehabilitation following a dislocation is insufficient to regain patellofemoral joint stability, or if certain risk factors for recurrent dislocation are present, surgical intervention, such as medial patellofemoral ligament (MPFL) reconstruction or tibial tubercle transfer, is considered.²² A systematic review concluded that MPFL reconstruction following a first-time dislocation yielded lower redislocation rates of 7% compared to 30% with nonoperative treatment.²³

Major tendon rupture

Patellar tendon ruptures occur when a sudden eccentric force is applied to the knee, such as when landing from a jump with the knee flexed. Patellar tendon ruptures frequently are clinically apparent, with patients demonstrating a high-riding patella and loss of active knee extension. Quadriceps tendon ruptures often result from a similar injury mechanism in older patients, with a similar loss of active knee extension and a palpable gap superior to the patella.²⁴

Partial tears in patients who can maintain full extension of the knee against gravity are treated nonoperatively, but early surgical repair is indicated for complete quadriceps or patellar tendon ruptures to achieve optimal outcomes. Prompt diagnosis and treatment are critical, as repair delayed beyond 1 to 2 weeks postinjury is associated with worse outcomes.^25-28

Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise.

Even with prompt treatment, return to sport is not guaranteed. According to a recent systematic review, athletes returned to play 88.9% and 89.8% of the time following patellar and quadriceps tendon repairs, respectively. However, returning to the same level of play was less common and achieved 80.8% (patellar tendon repair) and 70% (quadriceps tendon repair) of the time. Return-to-work rates were higher, at 96% for both surgical treatments.²⁹

Locked knee and acute meniscus tears in younger patients

In some acute knee injuries, meniscus tears, loose cartilage bodies or osteochondral defects, or other internal structures can become interposed between the femoral and tibial surfaces, preventing both active and passive knee extension. Such injuries are often severely painful and functionally debilitating. While manipulation under anesthesia can acutely restore joint function,³⁰ diagnostic and therapeutic arthroscopy often is pursued for definitive treatment.³¹ Compared to the gold standard of diagnostic arthroscopy, preoperative magnetic resonance imaging (MRI) carries positive and negative predictive values of 85% and 77%, respectively, in identifying or ruling out the anatomic structure responsible for a locked knee. ³² As such, MRI has been proposed as a method to avoid performing arthroscopy on a patient with a “pseudo-locked” knee, or loss of range of motion due to pain but without a true mechanical block.³²

Continue to: Depending on the location...

Depending on the location, size, and shape of an acute meniscus tear in younger patients, surgical repair may be an option to preserve long-term joint function. In one case series of patients younger than 20 years, 62% of meniscus repairs yielded good outcomes after a mean follow-up period of 16.8 years.³³

Osteochondritis dissecans

Osteochondritis dissecans is characterized by subchondral bone osteonecrosis that most often occurs in pediatric patients, potentially causing the separation of a fragment of articular cartilage and subchondral bone into the joint space (FIGURE 2). In early stages, nonoperative treatment consisting of prolonged rest followed by physical therapy to gradually return to activity is recommended to prevent small, low-grade lesions from progressing to unstable or separated fragments. Arthroscopy, which consists of microfracture or other surgical resurfacing techniques to restore joint integrity, is pursued in more advanced cases of unstable or separated fragments.

High-quality data guiding the management of osteochondritis dissecans are lacking, and these recommendations are based on consensus guidelines.³⁴

Septic arthritis

Septic arthritis is a medical emergency caused by the hematogenous spread of microorganisms, most often staphylococci and streptococci species. Less commonly, it arises from direct inoculation through an open wound or, rarely, iatrogenically following a joint injection procedure. Clinical signs of septic arthritis include joint pain, joint swelling, and fever. Passive range of motion of the joint is often severely painful. Synovial fluid studies consistent with septic arthritis include an elevated white blood cell count greater than 25,000/mcL with polymorphonuclear cell predominance.³⁵ The knee accounts for more than 50% of septic arthritis cases, and surgical drainage usually is required to achieve infection source control and decrease morbidity and mortality due to destruction of articular cartilage when treatment is delayed.³⁶

Chronic knee injuries and pain

Surgical intervention for chronic knee injuries and pain generally is considered when patients demonstrate significant functional impairment and persistent symptoms despite pursuing numerous nonsurgical treatment options. A significant portion of chronic knee pain is due to degenerative processes such as OA or meniscus injuries, or tears without a history of trauma that do not cause locking of the knee. Treatments for degenerative knee pain include supervised exercise, physical therapy, bracing, offloading with a cane or other equipment, topical or oral ­anti-inflammatories or analgesics, and injectable therapies such as intra-articular ­corticosteroids.³⁷

Continue to: Other common causes...

Other common causes of chronic knee pain include chronic tendinopathy or biomechanical syndromes such as patellofemoral pain syndrome or iliotibial band syndrome. Surgical treatment of these conditions is pursued in select cases and only after exhausting nonoperative treatment programs, as recommended by international consensus statements,³⁸ societal guidelines,³⁹ and expert opinion.⁴⁰ High-quality data on the effectiveness, or ineffectiveness, of surgical intervention for these conditions are lacking.

A chief risk factor for a patellar dislocation is a history of prior dislocation.

Despite being one of the most commonly performed surgical procedures in the United States,⁴¹ arthroscopic partial meniscectomy treatment of degenerative meniscus tears does not lead to improved outcomes compared to nonsurgical management, according to multiple recent studies.^42-45 Evidence does not support routine arthroscopic intervention for degenerative meniscus tears or OA,⁴² and recent guidelines recommend against it⁴⁶ or to pursue it only after nonsurgical treatments have failed.³⁷

Surgical management of degenerative knee conditions generally consists of partial or total arthroplasty and is similarly considered after failure of conservative measures. Appropriate use criteria that account for multiple clinical and patient factors are used to enhance patient selection for the procedure.⁴⁷

Takeaways

Primary care clinicians will treat patients sustaining knee injuries and see many patients with knee pain in the outpatient setting. Treatment options vary considerably depending on the underlying diagnosis and resulting functional losses. Several categories of clinical presentation, including neurovascular injury, unstable or displaced fractures, joint instability, major tendon rupture, significant mechanical symptoms such as a locked knee, certain osteochondral injuries, and septic arthritis, likely or almost always warrant surgical consultation ­(TABLE^3-10,12-36). Occasionally, as in the case of neurovascular injury or septic arthritis, such consultation should be emergent.

CORRESPONDENCE
David M. Siebert, MD, Sports Medicine Center at Husky Stadium, 3800 Montlake Boulevard NE, Seattle, WA 98195; siebert@uw.edu

Evidence supports what family physicians know to be true: Knee pain is an exceedingly common presenting problem in the primary care office. Estimates of lifetime incidence reach as high as 54%,¹ and the prevalence of knee pain in the general population is increasing.² Knee disability can result from acute or traumatic injuries as well as chronic, degenerative conditions such as osteoarthritis (OA). The decision to pursue orthopedic consultation for a particular injury or painful knee condition can be challenging. To address this, we highlight specific knee diagnoses known to cause pain, with the aim of describing which conditions likely will necessitate surgical consultation—and which won’t.

Acute or nondegenerative knee injuries and pain

Acute knee injuries range in severity from simple contusions and sprains to high-energy, traumatic injuries with resulting joint instability and potential neurovascular compromise. While conservative treatment often is successful for many simple injuries, surgical management—sometimes urgently or emergently—is needed in other cases, as will be detailed shortly.

Neurovascular injury associated with knee dislocations

Acute neurovascular injuries often require emergent surgical intervention. Although rare, tibiofemoral (knee) dislocations pose a significant challenge to the clinician in both diagnosis and management. The reported frequency of popliteal artery injury or rupture following a dislocation varies widely, with rates ranging from 5% to 64%, according to older studies; more recent data, however, suggest the rate is actually as low as 3.3%.³ Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified. Identifying a knee dislocation can prove challenging, as spontaneous joint reduction occurs in as many as 50% of cases, potentially shrouding the severity of the injury on initial evaluation.⁴

Immediate immobilization and emergency department transport for monitoring, orthopedics consultation, and vascular ­studies or vascular surgery consultation is recommended in the case of a suspected knee dislocation. In one cross-sectional cohort study, the surgical management of knee ­dislocations yielded favorable outcomes in > 75% of cases.⁵

Tibial plateau fracture

This fracture often occurs as a result of high-energy trauma, such as contact sports or motor vehicle accidents, and is characterized by a proximal tibial fracture line with extension to the articular surface. X-rays often are sufficient for initial diagnosis. Computed tomo­graphy can help rule out a fracture line when clinical suspicion is high and x-rays are nondiagnostic. As noted earlier, any suggestion of neurovascular compromise on physical exam requires an emergent orthopedic surgeon consultation for a possible displaced and unstable (or more complex) injury (FIGURE 1).^6-8

Nondisplaced tibial plateau fractures without supraphysiologic ligamentous laxity on valgus or varus stress testing can be managed safely with protection and early mobilization, gradual progression of weight-bearing, and serial x-rays to ensure fracture healing and stability. Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise. Suspicion of these complications is raised by distal neuropathic symptoms of paresthesia or skin anesthesia, progressively worsening pain distal to the knee, or vascular signs of pallor, delayed or lost capillary refill, or decreased or absent distal pulses.

Gross joint instability identified by positive valgus or varus stress testing, positive anterior or posterior drawer testing, or patient inability to tolerate these maneuvers due to pain similarly should raise suspicion for a more significant fracture at risk for concurrent neurovascular injury. Acute compartment syndrome also is a known complication of tibial plateau fractures and similarly requires emergent operative management. Urgent surgical consultation is recommended for fractures with displaced fracture fragments, tibial articular surface step-off or depression, fractures with concurrent joint laxity, or medial plateau fractures.^6-8

Continue to: Patella fractures

Patella fractures

These fractures occur as a direct blow to the front of the knee, such as falling forward onto a hard surface, or indirectly due to a sudden extreme eccentric contraction of the quadriceps muscle. Nondisplaced fractures with an intact knee extension mechanism, which is examined via a supine straight-leg raise or seated knee extension, are managed with weight-bearing as tolerated in strict immobilization in full extension for 4 to 6 weeks, with active range-of-motion and isometric quadriceps exercises beginning in 1 to 2 weeks. Serial x-rays also are obtained to ensure fracture displacement does not occur during the rehabilitation process.⁹

High-quality evidence guiding follow-up care and comparing outcomes of surgical and nonsurgical management of patella fractures is lacking, and studies comparing different surgical techniques are of lower methodological quality.¹⁰ Nevertheless, displaced or comminuted patellar fractures are referred urgently to orthopedic surgical care for fixation, as are those with concurrent loose bodies, chondral surface injuries or articular step-off, or osteochondral fractures.⁹ Inability to perform a straight-leg raise (ie, clinical loss of the knee extension mechanism) suggests a fracture under tension that likely also requires surgical fixation for successful recovery. Neurovascular injuries are unlikely in most patellar fractures but would require emergent surgical consultation.⁹

Ligamentous injury

Tibiofemoral joint laxity occurs as a result of ligamentous injury, with or without tibial plateau fracture. The anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), medial collateral ligament (MCL), and lateral collateral ligament (LCL) comprise the 4 main ligaments of the knee. The ACL resists anterior tibial translation and rotational forces, while the PCL resists posterior tibial translation. The MCL and LCL resist valgus and varus stress, respectively.

Ligament injuries are classified as Grades 1 to 3¹¹:

• Grade 1 sprains. The ligament is stretched, but there is no macroscopic tearing; joint stability is maintained.
• Grade 2 sprains. There are partial macroscopic ligament tears. There is joint laxity due to the partial loss of the ligament’s structural integrity.
• Grade 3 sprains. The ligament is fully avulsed or ruptured with resultant gross joint instability.

Vascular injury can lead to irreversible tissue damage and even limb loss if not promptly identified.

The decision to pursue surgical repair of a knee sprain depends heavily on the likelihood of keeping or regaining and maintaining functional joint stability during the injury recovery and postinjury time periods. Injuries that do not result in joint instability or injuries with a high likelihood of returning to a stable state with conservative measures often do not require surgical intervention.

Continue to: ACL tears

ACL tears occur most commonly via a noncontact event, as when an individual plants their foot and suddenly changes direction during sport or other physical activity. Treatment hinges on patient activity levels and participation in sports. Patients who do not plan to engage in athletic movements (that require changes in direction or planting and twisting) and who otherwise maintain satisfactory joint stability during activities of daily living may elect to defer or even altogether avoid surgical reconstruction of isolated ACL tears. One pair of studies demonstrated equivalent outcomes in surgical and nonsurgical management in 121 young, nonelite athletes at 2- and 5-year follow-up, although the crossover from the nonsurgical to surgical groups was high.^12,13 Athletes who regain satisfactory function and stability nonoperatively can defer surgical intervention. However, the majority of active patients and athletes will require surgical ACL reconstruction to return to pre-injury functional levels.¹⁴

PCL sprains occur as a result of sudden posteriorly directed force on the tibia, such as when the knee is hyperextended or a patient falls directly onto a flexed knee. Patients with Grade 1 and 2 isolated sprains generally will recover with conservative care, as will patients with some Grade 3 complete tears that do not fully compromise joint stability. However, high-grade PCL injuries often are comorbid with posterolateral corner or other injuries, leading to a higher likelihood of joint instability and thus the need for surgical intervention for the best chance at an optimal outcome.¹⁵

MCL sprain. Surgical management is not required in an isolated Grade 1 or 2 MCL sprain, as the hallmarks of recovery—return of joint stability, knee strength and range of motion, and pain ­reduction—can be achieved successfully with conservative management. Isolated Grade 3 MCL sprains are also successfully managed nonoperatively¹⁶ except in specific cases, such as a concurrent large avulsion fracture.¹⁷

LCL sprain. Similarly, isolated Grade 1 and 2 LCL sprains generally do not require surgical intervention. However, Grade 3 LCL injuries usually do, as persistent joint instability and poor functional outcomes are more common with nonsurgical management.^18-20 Additionally, high-grade LCL injuries frequently manifest with comorbid meniscus injuries or sprains of the posterolateral corner of the knee, a complex anatomic structure that provides both static and dynamic tibiofemoral joint stability. Surgical repair or reconstruction of the posterolateral corner frequently is necessary for optimal functional outcomes.²¹

Multiligamentous sprains frequently lead to gross joint instability and necessitate orthopedic surgeon consultation to determine the best treatment plan; this should be done emergently if neurovascular compromise is suspected. A common injury combination is simultaneous ACL and MCL sprains with or without meniscus injury. In these cases, some surgeons will choose to defer ACL reconstruction until after MCL healing is achieved. This allows the patient to regain valgus stability of the joint prior to performing ACL reconstruction to regain rotational and anterior stability.²⁰

Continue to: Patellar dislocations

Patellar dislocations represent a relatively common knee injury in young active patients, often occurring in a noncontact fashion when a valgus force is applied to an externally rotated and planted lower leg. A chief risk factor for a patellar dislocation is a history of prior dislocation. If rehabilitation following a dislocation is insufficient to regain patellofemoral joint stability, or if certain risk factors for recurrent dislocation are present, surgical intervention, such as medial patellofemoral ligament (MPFL) reconstruction or tibial tubercle transfer, is considered.²² A systematic review concluded that MPFL reconstruction following a first-time dislocation yielded lower redislocation rates of 7% compared to 30% with nonoperative treatment.²³

Major tendon rupture

Patellar tendon ruptures occur when a sudden eccentric force is applied to the knee, such as when landing from a jump with the knee flexed. Patellar tendon ruptures frequently are clinically apparent, with patients demonstrating a high-riding patella and loss of active knee extension. Quadriceps tendon ruptures often result from a similar injury mechanism in older patients, with a similar loss of active knee extension and a palpable gap superior to the patella.²⁴

Partial tears in patients who can maintain full extension of the knee against gravity are treated nonoperatively, but early surgical repair is indicated for complete quadriceps or patellar tendon ruptures to achieve optimal outcomes. Prompt diagnosis and treatment are critical, as repair delayed beyond 1 to 2 weeks postinjury is associated with worse outcomes.^25-28

Surgical management and fixation are required emergently for open fractures or gross joint instability with vascular or neurologic compromise.

Even with prompt treatment, return to sport is not guaranteed. According to a recent systematic review, athletes returned to play 88.9% and 89.8% of the time following patellar and quadriceps tendon repairs, respectively. However, returning to the same level of play was less common and achieved 80.8% (patellar tendon repair) and 70% (quadriceps tendon repair) of the time. Return-to-work rates were higher, at 96% for both surgical treatments.²⁹

Locked knee and acute meniscus tears in younger patients

In some acute knee injuries, meniscus tears, loose cartilage bodies or osteochondral defects, or other internal structures can become interposed between the femoral and tibial surfaces, preventing both active and passive knee extension. Such injuries are often severely painful and functionally debilitating. While manipulation under anesthesia can acutely restore joint function,³⁰ diagnostic and therapeutic arthroscopy often is pursued for definitive treatment.³¹ Compared to the gold standard of diagnostic arthroscopy, preoperative magnetic resonance imaging (MRI) carries positive and negative predictive values of 85% and 77%, respectively, in identifying or ruling out the anatomic structure responsible for a locked knee. ³² As such, MRI has been proposed as a method to avoid performing arthroscopy on a patient with a “pseudo-locked” knee, or loss of range of motion due to pain but without a true mechanical block.³²

Continue to: Depending on the location...

Depending on the location, size, and shape of an acute meniscus tear in younger patients, surgical repair may be an option to preserve long-term joint function. In one case series of patients younger than 20 years, 62% of meniscus repairs yielded good outcomes after a mean follow-up period of 16.8 years.³³

Osteochondritis dissecans

Osteochondritis dissecans is characterized by subchondral bone osteonecrosis that most often occurs in pediatric patients, potentially causing the separation of a fragment of articular cartilage and subchondral bone into the joint space (FIGURE 2). In early stages, nonoperative treatment consisting of prolonged rest followed by physical therapy to gradually return to activity is recommended to prevent small, low-grade lesions from progressing to unstable or separated fragments. Arthroscopy, which consists of microfracture or other surgical resurfacing techniques to restore joint integrity, is pursued in more advanced cases of unstable or separated fragments.

High-quality data guiding the management of osteochondritis dissecans are lacking, and these recommendations are based on consensus guidelines.³⁴

Septic arthritis

Septic arthritis is a medical emergency caused by the hematogenous spread of microorganisms, most often staphylococci and streptococci species. Less commonly, it arises from direct inoculation through an open wound or, rarely, iatrogenically following a joint injection procedure. Clinical signs of septic arthritis include joint pain, joint swelling, and fever. Passive range of motion of the joint is often severely painful. Synovial fluid studies consistent with septic arthritis include an elevated white blood cell count greater than 25,000/mcL with polymorphonuclear cell predominance.³⁵ The knee accounts for more than 50% of septic arthritis cases, and surgical drainage usually is required to achieve infection source control and decrease morbidity and mortality due to destruction of articular cartilage when treatment is delayed.³⁶

Chronic knee injuries and pain

Surgical intervention for chronic knee injuries and pain generally is considered when patients demonstrate significant functional impairment and persistent symptoms despite pursuing numerous nonsurgical treatment options. A significant portion of chronic knee pain is due to degenerative processes such as OA or meniscus injuries, or tears without a history of trauma that do not cause locking of the knee. Treatments for degenerative knee pain include supervised exercise, physical therapy, bracing, offloading with a cane or other equipment, topical or oral ­anti-inflammatories or analgesics, and injectable therapies such as intra-articular ­corticosteroids.³⁷

Continue to: Other common causes...

Other common causes of chronic knee pain include chronic tendinopathy or biomechanical syndromes such as patellofemoral pain syndrome or iliotibial band syndrome. Surgical treatment of these conditions is pursued in select cases and only after exhausting nonoperative treatment programs, as recommended by international consensus statements,³⁸ societal guidelines,³⁹ and expert opinion.⁴⁰ High-quality data on the effectiveness, or ineffectiveness, of surgical intervention for these conditions are lacking.

A chief risk factor for a patellar dislocation is a history of prior dislocation.

Despite being one of the most commonly performed surgical procedures in the United States,⁴¹ arthroscopic partial meniscectomy treatment of degenerative meniscus tears does not lead to improved outcomes compared to nonsurgical management, according to multiple recent studies.^42-45 Evidence does not support routine arthroscopic intervention for degenerative meniscus tears or OA,⁴² and recent guidelines recommend against it⁴⁶ or to pursue it only after nonsurgical treatments have failed.³⁷

Surgical management of degenerative knee conditions generally consists of partial or total arthroplasty and is similarly considered after failure of conservative measures. Appropriate use criteria that account for multiple clinical and patient factors are used to enhance patient selection for the procedure.⁴⁷

Takeaways

Primary care clinicians will treat patients sustaining knee injuries and see many patients with knee pain in the outpatient setting. Treatment options vary considerably depending on the underlying diagnosis and resulting functional losses. Several categories of clinical presentation, including neurovascular injury, unstable or displaced fractures, joint instability, major tendon rupture, significant mechanical symptoms such as a locked knee, certain osteochondral injuries, and septic arthritis, likely or almost always warrant surgical consultation ­(TABLE^3-10,12-36). Occasionally, as in the case of neurovascular injury or septic arthritis, such consultation should be emergent.

CORRESPONDENCE
David M. Siebert, MD, Sports Medicine Center at Husky Stadium, 3800 Montlake Boulevard NE, Seattle, WA 98195; siebert@uw.edu

Four medications comprise the drug category known as direct oral anticoagulants (DOACs). Dabigatran (Pradaxa)¹ was the first to gain approval. It was approved by the US Food and Drug Administration (FDA) in 2010 for the reduction of stroke and systemic embolism in patients with nonvalvular atrial fibrillation (AF). This was followed by approvals for rivaroxaban (Xarelto)² in 2011, apixaban (Eliquis)³ in 2012, and edoxaban (Savaysa)⁴ in 2015. Betrixaban (Bevyxxa)⁵ was approved in 2017 for venous thromboembolism (VTE) prophylaxis in acutely ill hospitalized patients with restricted mobility, but it was removed from the market in 2020.

IMAGE: © KO STUDIOS

In addition to stroke prevention in nonvalvular AF, each DOAC has been approved for other indications and has been addressed further in guideline-based recommendations outside FDA-approved indications. This review highlights the evolving use of DOACs and the expansion of applications for multiple adult patient populations.

Overview of DOACs

Dabigatran is the only direct thrombin inhibitor; the other agents inhibit factor Xa. TABLE 1^1-4 summarizes FDA-­approved indications and dosing and guideline-based dosing. Dabigatran and edoxaban require parenteral anticoagulation for 5 to 10 days prior to initiation for acute VTE, limiting their use.^1,4TABLE 2^1-4 highlights pharmacokinetic differences among the agents. For example, dabigatran is 80% renally cleared, is somewhat dialyzable, and can accumulate in patients with renal dysfunction.¹ Edoxaban is contraindicated for nonvalvular AF in patients with a creatinine clearance (CrCl) > 95 mL/min because an increased stroke risk was demonstrated.⁴ Therefore, rivaroxaban and apixaban are prescribed most often in the United States.^6,7

Applications in special patient populations

Obesity

As of 2020, more than 40% of adults in the United States were obese (body mass index [BMI] ≥ 30), with 9% classified as class 3 or severely obese (BMI ≥ 40).⁸ Altered drug pharmacokinetics in patients with severe obesity raises concern for undertreatment with fixed-dose DOACs. Phase III DOAC approval trials included patients with obesity, but weight cutoffs differed, making extrapolating efficacy and safety data difficult across different obesity stages.⁹ Although no FDA-labeled dosing adjustments exist for patients with obesity, the International Society on Thrombosis and Haemostasis (ISTH) does provide such recommendations.

ISTH changes position on measuring drug levels. ISTH previously recommended avoiding DOACs in those with a BMI > 40 or body weight > 120 kg. If a DOAC was used, ISTH advised obtaining peak and trough drug levels.¹⁰ However, DOAC drug levels have not been associated with clinical outcomes or sufficient degrees of anticoagulation.¹¹

Men and women are affected equally by fibrolipomas. Prevalence does not differ by race or ethnicity.

In April 2021, ISTH updated guidance on DOACs in obesity, indicating standard doses of rivaroxaban or apixaban can be used for the treatment and prevention of VTE in all patients regardless of weight or BMI. Because data in obesity are lacking for dabigatran and edoxaban, avoid using these agents in patients with a BMI > 40 or weight > 120 kg. Additionally, assessing drug levels is no longer recommended, as there is insufficient evidence that these impact clinical outcomes.¹²

The 2021 American College of Chest Physicians (CHEST) guideline update does not recommend adjustments based on weight,¹³ and the 2019 European Society of Cardiology (ESC) guidelines make no mention of weight when treating acute pulmonary embolism.¹⁴

Continue to: Effectiveness of DOACs for AF in patients with obesity isn't clear

Effectiveness of DOACs for AF in patients with obesity isn’t clear, as most data are from retrospective cohort analyses. In patients weighing > 120 kg, dabigatran has shown efficacy in thrombosis prevention similar to that achieved in those weighing ≤ 120 kg, but it has increased the risk for gastrointestinal (GI) bleeding.¹⁵ Another study indicated a 15-mg dose of rivaroxaban may be associated with increased thromboembolic complications in patients with a BMI ≥ 35.¹⁶ Alternatively, another retrospective study of rivaroxaban demonstrated a small absolute risk reduction in ischemic stroke among patients in all stages of obesity and no difference in significant bleeding events.¹⁷ One further retrospective cohort showed that, in patients with a BMI ≥ 50 kg, the effectiveness of rivaroxaban and apixaban in thrombosis prevention and bleeding safety outcomes was comparable to that seen in those with a BMI < 30.¹⁸

As a result of conflicting data, and a lack of prospective randomized controlled trials (RCTs), ISTH continued recommending international normalized ratio (INR)–based dosing of warfarin for class 3 or severely obese patients with AF. The 2018 CHEST guidelines¹⁹ and the 2020 ESC guidelines²⁰ make no mention of DOAC avoidance in patients with obesity and AF.

Advanced and end-stage renal disease

DOACs are renally dosed based on indication, drug-drug interactions, and degree of renal function (TABLE 3^1-4). For example, patients with AF who are anticoagulated with apixaban are prescribed 2.5 mg twice daily when 2 of the 3 following criteria are met: age ≥ 80 years, body weight ≤ 60 kg, serum creatinine ≥ 1.5 mg/dL. However, no dosage adjustment is necessary for VTE treatment or prophylaxis with apixaban regardless of renal function.³

Data supporting the safety and efficacy of DOACs in end-stage renal disease (ESRD) are sparse. All DOACs are renally cleared to varying degrees (TABLE 2^1-4), theoretically increasing bleeding risk as kidney disease progresses. Apixaban is the least renally cleared of the DOACs and has been evaluated in the greatest number of trials for patients with ESRD for both VTE treatment and prevention and nonvalvular AF.²¹ As a result, the FDA approved standard-dose apixaban (5 mg twice daily) for VTE treatment and prevention and nonvalvular AF in patients with ESRD, even those requiring dialysis. Use the reduced apixaban dose (2.5 mg twice daily) in patients with ESRD and AF only if they are ≥ 80 years of age or their body weight is ≤ 60 kg.³

Patients with cancer

Cancer-associated acute VTE treatment. Cancer is an established risk factor for acute VTE but it also increases the risk for treatment-­associated bleeding compared with patients without cancer.²² Historically, low-molecular-weight heparin (LMWH) was recommended over warfarin and DOACs for cancer-associated thromboses (CAT).²³ Compared with warfarin, LMWH reduced the rate of recurrent VTE and had similar or reduced bleeding rates at 6 to 12 months.^24-26 However, clinicians and patients often chose warfarin to avoid subcutaneous injections.²⁷

CHEST guidelines recommend oral Xa inhibitors over LMWH for the treatment of CAT.¹³ The 2020 guidelines of the National Institute for Health and Care Excellence (NICE) recommend DOACs as an option for CAT along with LMWH or LMWH transitioned to warfarin.²⁸ The American Society of Clinical Oncology (ASCO) recommends rivaroxaban for acute VTE treatment in CAT. No head-to-head trials have evaluated comparative efficacy of DOACs for CAT. However, edoxaban and rivaroxaban are associated with a greater risk for GI bleeding; therefore, apixaban is preferred in patients with GI malignancies.²⁹ Standard DOAC VTE treatment dosing is recommended for all 3 agents.^2-4

When using DOACs for patients with CAT, consider potential drug-drug interactions with chemotherapy regimens. All DOACs­ are transported by p-glycoprotein, while rivaroxaban and apixaban are substrates of cytochrome P450, leading to potentially significant drug-drug interactions.³⁰ These interactions could affect the patient’s chemotherapeutic regimen, decrease the efficacy of the DOAC, or increase the risk for bleeding. Therefore, anticoagulation choice should be made in collaboration with the ­hematology/oncology team.

Continue to: Cancer-associated VTE prophylaxis...

Cancer-associated VTE prophylaxis. VTE prophylaxis for patients with cancer is complex and necessitates a global assessment of cancer location and treatment regimen and setting. Hospitalized patients receiving chemotherapy are at high risk for VTE if mobility is reduced or if other VTE risk factors are present. The International Initiative on Thrombosis and Cancer (ITAC)³¹ and ISTH³² recommend VTE prophylaxis with unfractionated heparin or LMWH (ISTH recommends LMWH more strongly). The 2020 ASCO Guidelines recommend pharmacologic anticoagulation but make no drug-specific recommendation.²⁹ Parenteral treatment in hospitalized patients is not as burdensome as it is in ambulatory patients; therefore, these recommendations are less likely to elicit inpatient opposition.

In the ambulatory setting, patient avoidance of subcutaneous injections necessitates consideration of DOACs for CAT prophylaxis. The Khorana Risk Score (KRS) is a validated tool (scale, 0-7) to predict VTE risk in ambulatory patients receiving chemotherapy.³³ KRS scores ≥ 2 indicate high thrombotic risk and the need for prophylactic anticoagulation. ASCO recommends apixaban, rivaroxaban, or LMWH.²⁹ ISTH and ITAC both recommend apixaban or rivaroxaban over LMWH.^31,34 An RCT published in June 2023 confirmed that, for adults with cancer and VTE, DOACs were noninferior to LMWH for preventing recurrent VTE for 6 months.³⁵ The recommended doses for apixaban (2.5 mg twice daily) and rivaroxaban (10 mg daily) for CAT VTE prophylaxis are lower than FDA-approved treatment doses.³¹

Patients with thrombophilia: VTE prevention

Thrombophilias are broadly categorized as inherited or acquired, with inherited thrombophilia being more prevalent. The Factor V Leiden (FVL) variant affects 2% to 7% of the population, and prothrombin gene mutation (PGM) affects 1% to 2% of the population.³⁶ Other forms of inherited thrombophilia, such as protein C deficiency, protein S deficiency, and antithrombin deficiency, occur less commonly (< 0.7% of the population).³⁶ Antiphospholipid syndrome (APS), the most common acquired thrombophilia, affects approximately 2% of the population.³⁶ APS involves multiple antibodies: anticardiolipin antibodies, lupus anticoagulant, and anti-beta-2 glycoprotein 1 antibodies. Establishing risk for thrombosis across the varying types of thrombophilia has proven difficult, but APS is considered the most thrombogenic thrombophilia apart from extremely rare homozygous inherited thrombophilias.³⁶ Therefore, DOAC recommendations are thrombophilia specific.

Assessing DOAC blood levels is no longer recommended for patients with obesity, as there is insufficient evidence that these measures affect clinical outcomes.

A prospective cohort study evaluated DOACs compared with heparin/warfarin for VTE treatment in patients with inherited thrombophilias.³⁷ Although all 4 available DOACs were included, most patients (61.1%) received rivaroxaban. Patients with an array of inherited thrombophilias, including rare homozygous mutations, were enrolled in this trial. While most patients (66.9%) had a “mild thrombophilia” defined as either FVL or PGM, the remainder had more severe thrombophilias.³⁷ VTE recurrence was similar and uncommon in the DOAC and heparin/warfarin groups, consistent with a previous meta-analysis.³⁸ Surprisingly, an increase in the cumulative risk for bleeding was seen in the DOAC group compared with the warfarin group, a finding inconsistent with prior trials.³⁸ There were no major bleeding events in the DOAC group, but 3 such events occurred in the heparin/warfarin group, including 2 intracranial hemorrhages.

Currently NICE, CHEST, and ISTH do not make a recommendation for a preferred agent in patients with an acute VTE and inherited thrombophilia; however, DOACs would not be inappropriate.^23,28,32 The American Society of Hematology (ASH) had planned to release recommendations related to the treatment of thrombophilia in 2020, but they were delayed by the COVID-19 pandemic.³⁹

APS presents challenges for acute VTE anticoagulation. First, it causes a strongly thrombogenic state necessitating therapeutic anticoagulation. Second, for patients with positive lupus anticoagulant, INR monitoring and standardized INR goals may be inadequate.⁴⁰ Therefore, using fixed-dose DOACs without the need for therapeutic monitoring is appealing, but significant concerns exist for using DOACs in patients with APS.^41-45 ISTH and CHEST recommend warfarin for the treatment and prevention of acute VTE in patients with APS, especially those with triple-­positive (anticardiolipin, lupus anticoagulant, and anti-beta-2 glycoprotein 1) APS.^13,46 Package labeling for all DOACs recommends avoidance in triple-positive APS.^1-4

ASTRO-APS is the most recent RCT to compare apixaban and warfarin for patients with APS,⁴⁷ and it was terminated early after 6 of 23 patients in the apixaban group had thrombotic events, while no one in the warfarin group had such an event.⁴⁸ Subsequently, a meta-analysis⁴⁹ demonstrated that patients with thrombotic APS appear to have a greater risk for arterial thrombosis when treated with DOACs compared with warfarin. These 2 studies may lead to changes in recommendations to avoid DOACs in all patients with APS or may prompt more focused trials for DOAC use in patients with APS plus an antiplatelet to mitigate arterial thrombotic risk.

Continue to: Expanded clinical indications

Expanded clinical indications

Superficial vein thrombosis

Superficial thrombophlebitis or superficial vein thrombosis (SVT) is estimated to occur 6 times more frequently than VTE.⁵⁰ Management of patients with isolated, uncomplicated thrombophlebitis who are at low risk for extension of the SVT involves symptomatic treatment with nonsteroidal anti-inflammatory drugs, topical agents, or compression therapy. However, depending on risk for progression, anticoagulation may be recommended.⁵¹

Patients at intermediate risk for extension or propagation of SVT are candidates for anticoagulation. The CHEST guidelines recommend fondaparinux 2.5 mg subcutaneous injections daily for 45 days instead of LMWH or warfarin.¹³ However, if patients decline 6 weeks of daily injections, the guidelines acknowledge that rivaroxaban 10 mg daily may be an alternative.¹³

Certain situations should prompt one to consider using a treatment dose of a DOAC for 3 months. These include cases in which the SVT is located within 3 cm of the deep venous system, expands despite an appropriate prophylactic regimen, or recurs after discontinuation of prophylactic anticoagulation.^13,50

Acute coronary syndrome

The American College of Cardiology/­American Heart Association (ACC/AHA) recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.⁵² Data are mixed regarding longer-term anticoagulation in addition to dual antiplatelet therapy in outpatient settings to prevent thrombosis recurrence in the absence of AF.

For patients at intermediate risk for extension of superficial vein thrombosis who decline daily subcutaneous injections of fondaparinux, rivaroxaban 10 mg/d may be an alternative.

The APPRAISE-2 trial enrolled high-risk patients with ACS within 7 days of the event.⁵³ Apixaban 5 mg twice daily was compared with placebo in patients taking aspirin or aspirin plus clopidogrel. The trial was terminated early because major bleeding events increased with apixaban without reduction in recurrent ischemic events. The ATLAS ACS-TIMI 46 trial evaluated different rivaroxaban doses (5-20 mg daily) in ACS patients.⁵⁴ The study revealed possible thrombosis benefit but also increased risk for bleeding, particularly at higher doses. As a result, another study—ATLAS ACS 2-TIMI 51—was conducted and compared the use of low-dose rivaroxaban (2.5 mg twice daily or 5 mg twice daily) vs placebo for patients with recent ACS.⁵⁵ All patients were receiving low-dose aspirin, and approximately 93% of patients in each group also were receiving clopidogrel or ticlopidine. As in the APPRAISE-2 trial, rivaroxaban increased the rate of major bleeding and intracranial hemorrhage; however, it did not increase the incidence of fatal bleeding. Unlike APPRAISE-2, rivaroxaban significantly reduced the primary efficacy end point, a composite of death from cardiovascular causes, myocardial infarction, or stroke (absolute risk reduction = 1.8%; number needed to treat = 56 for combined rivaroxaban doses).⁵⁵

A secondary subgroup analysis combined data from the ATLAS ACMS-TIMI 46 and ATLAS ACS 2-TIMI 51 trials to evaluate outcomes in patients receiving aspirin monotherapy when combined with rivaroxaban 2.5 mg twice daily or 5 mg twice daily or with placebo.⁵⁶ The primary efficacy end point was a composite of cardiovascular death, myocardial infarction, or stroke. When the 2 trials were evaluated separately, neither rivaroxaban dose was associated with reduction of the primary efficacy outcomes compared with aspirin alone. However, when the data were pooled, both the combined rivaroxaban doses (particularly the 5-mg dose) were associated with reduced cardiovascular outcomes. From a safety perspective, the 2.5-mg twice-daily dose of rivaroxaban was the only dose not associated with increased major bleeding risk. Thus, the 2.5-mg twice-daily dose of rivaroxaban may not provide sufficient cardiovascular benefit in patients with ACS, while the larger dose may increase the risk for nonfatal major bleeding events.⁵⁶

The European Medicines Agency⁵⁷ approved rivaroxaban 2.5 mg twice daily for ACS, and the 2020 ESC guidelines⁵⁸ consider it an appropriate therapeutic option in addition to aspirin for patients at high ischemic risk and low bleeding risk. ACS is not an FDA-approved indication for DOACs, and the ACC/AHA Guideline for the Management of ACS, last updated in 2014, does not include DOACs for ACS unless patients have AF.⁵² Ongoing trials are further investigating rivaroxaban for ACS, so the use of DOACs in the post-acute phase of ACS may become clearer in the future.⁵⁹

Continue to: Heparin-induced thrombocytopenia

Heparin-induced thrombocytopenia

Historically, nonheparin parenteral anticoagulants argatroban, bivalirudin, and fondaparinux were recommended for patients at risk for or who had heparin-induced thrombocytopenia (HIT). Argatroban is the only drug FDA approved for the treatment and prophylaxis of HIT; recommendations for the others are based on guideline recommendations.^23,60,61 The nonheparin parenteral anticoagulants cost between $700 and $1500 per day; therefore most patients with HIT are transitioned to warfarin.⁶² However, protein C and S inhibition and a subsequent prothrombotic state conveyed by warfarin initiation necessitates a minimum 5-day bridge to therapeutic warfarin with a nonheparin parenteral anticoagulant.

In vitro tests show that DOACs do not promote development of HIT antibodies⁶³ or affect platelet activation or aggregation.⁶⁴ A literature summary of DOACs for HIT determined that in 104 patients, all but 1 achieved platelet recovery (defined as > 150,000/mcL) within a median time of 7 days. Therapeutically, DOACs prevented new or recurrent VTE in 102/104 cases (98%), and only 3% of patients experienced significant bleeding events.⁶²

The ACC/AHA recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.

The 2018 ASH guidelines for VTE management in HIT include (with very low certainty of evidence) dabigatran, rivaroxaban, or apixaban for consideration in addition to previously recommended nonheparin parenteral anticoagulants.⁶¹ The dosing of each agent is contingent upon treatment of patients with HIT and an acute thrombosis (HITT) or HIT in the absence of VTE. For patients with HITT, treatment doses for acute VTE should be used for the appropriate duration of therapy (ie, 3 months). Importantly, dabigatran requires a 5-day pretreatment period with a parenteral anticoagulant, so it is not an ideal option. When treating isolated HIT (in the absence of VTE), ASH recommends all agents be dosed twice daily—dabigatran 150 mg twice daily (no 5-day parenteral pretreatment necessary), rivaroxaban 15 mg twice daily, or apixaban 5 mg twice daily—until platelet recovery (≥ 150,000/mcL) is achieved.⁶¹

CORRESPONDENCE
Kevin Schleich, PharmD, BCACP, Departments of Pharmaceutical Care and Family Medicine, University of Iowa, 200 Hawkins Drive, 01102-D PFP, Iowa City, IA, 52242; kevin-schleich@uiowa.edu

Four medications comprise the drug category known as direct oral anticoagulants (DOACs). Dabigatran (Pradaxa)¹ was the first to gain approval. It was approved by the US Food and Drug Administration (FDA) in 2010 for the reduction of stroke and systemic embolism in patients with nonvalvular atrial fibrillation (AF). This was followed by approvals for rivaroxaban (Xarelto)² in 2011, apixaban (Eliquis)³ in 2012, and edoxaban (Savaysa)⁴ in 2015. Betrixaban (Bevyxxa)⁵ was approved in 2017 for venous thromboembolism (VTE) prophylaxis in acutely ill hospitalized patients with restricted mobility, but it was removed from the market in 2020.

IMAGE: © KO STUDIOS

In addition to stroke prevention in nonvalvular AF, each DOAC has been approved for other indications and has been addressed further in guideline-based recommendations outside FDA-approved indications. This review highlights the evolving use of DOACs and the expansion of applications for multiple adult patient populations.

Overview of DOACs

Dabigatran is the only direct thrombin inhibitor; the other agents inhibit factor Xa. TABLE 1^1-4 summarizes FDA-­approved indications and dosing and guideline-based dosing. Dabigatran and edoxaban require parenteral anticoagulation for 5 to 10 days prior to initiation for acute VTE, limiting their use.^1,4TABLE 2^1-4 highlights pharmacokinetic differences among the agents. For example, dabigatran is 80% renally cleared, is somewhat dialyzable, and can accumulate in patients with renal dysfunction.¹ Edoxaban is contraindicated for nonvalvular AF in patients with a creatinine clearance (CrCl) > 95 mL/min because an increased stroke risk was demonstrated.⁴ Therefore, rivaroxaban and apixaban are prescribed most often in the United States.^6,7

Applications in special patient populations

Obesity

As of 2020, more than 40% of adults in the United States were obese (body mass index [BMI] ≥ 30), with 9% classified as class 3 or severely obese (BMI ≥ 40).⁸ Altered drug pharmacokinetics in patients with severe obesity raises concern for undertreatment with fixed-dose DOACs. Phase III DOAC approval trials included patients with obesity, but weight cutoffs differed, making extrapolating efficacy and safety data difficult across different obesity stages.⁹ Although no FDA-labeled dosing adjustments exist for patients with obesity, the International Society on Thrombosis and Haemostasis (ISTH) does provide such recommendations.

ISTH changes position on measuring drug levels. ISTH previously recommended avoiding DOACs in those with a BMI > 40 or body weight > 120 kg. If a DOAC was used, ISTH advised obtaining peak and trough drug levels.¹⁰ However, DOAC drug levels have not been associated with clinical outcomes or sufficient degrees of anticoagulation.¹¹

Men and women are affected equally by fibrolipomas. Prevalence does not differ by race or ethnicity.

In April 2021, ISTH updated guidance on DOACs in obesity, indicating standard doses of rivaroxaban or apixaban can be used for the treatment and prevention of VTE in all patients regardless of weight or BMI. Because data in obesity are lacking for dabigatran and edoxaban, avoid using these agents in patients with a BMI > 40 or weight > 120 kg. Additionally, assessing drug levels is no longer recommended, as there is insufficient evidence that these impact clinical outcomes.¹²

The 2021 American College of Chest Physicians (CHEST) guideline update does not recommend adjustments based on weight,¹³ and the 2019 European Society of Cardiology (ESC) guidelines make no mention of weight when treating acute pulmonary embolism.¹⁴

Continue to: Effectiveness of DOACs for AF in patients with obesity isn't clear

Effectiveness of DOACs for AF in patients with obesity isn’t clear, as most data are from retrospective cohort analyses. In patients weighing > 120 kg, dabigatran has shown efficacy in thrombosis prevention similar to that achieved in those weighing ≤ 120 kg, but it has increased the risk for gastrointestinal (GI) bleeding.¹⁵ Another study indicated a 15-mg dose of rivaroxaban may be associated with increased thromboembolic complications in patients with a BMI ≥ 35.¹⁶ Alternatively, another retrospective study of rivaroxaban demonstrated a small absolute risk reduction in ischemic stroke among patients in all stages of obesity and no difference in significant bleeding events.¹⁷ One further retrospective cohort showed that, in patients with a BMI ≥ 50 kg, the effectiveness of rivaroxaban and apixaban in thrombosis prevention and bleeding safety outcomes was comparable to that seen in those with a BMI < 30.¹⁸

As a result of conflicting data, and a lack of prospective randomized controlled trials (RCTs), ISTH continued recommending international normalized ratio (INR)–based dosing of warfarin for class 3 or severely obese patients with AF. The 2018 CHEST guidelines¹⁹ and the 2020 ESC guidelines²⁰ make no mention of DOAC avoidance in patients with obesity and AF.

Advanced and end-stage renal disease

DOACs are renally dosed based on indication, drug-drug interactions, and degree of renal function (TABLE 3^1-4). For example, patients with AF who are anticoagulated with apixaban are prescribed 2.5 mg twice daily when 2 of the 3 following criteria are met: age ≥ 80 years, body weight ≤ 60 kg, serum creatinine ≥ 1.5 mg/dL. However, no dosage adjustment is necessary for VTE treatment or prophylaxis with apixaban regardless of renal function.³

Data supporting the safety and efficacy of DOACs in end-stage renal disease (ESRD) are sparse. All DOACs are renally cleared to varying degrees (TABLE 2^1-4), theoretically increasing bleeding risk as kidney disease progresses. Apixaban is the least renally cleared of the DOACs and has been evaluated in the greatest number of trials for patients with ESRD for both VTE treatment and prevention and nonvalvular AF.²¹ As a result, the FDA approved standard-dose apixaban (5 mg twice daily) for VTE treatment and prevention and nonvalvular AF in patients with ESRD, even those requiring dialysis. Use the reduced apixaban dose (2.5 mg twice daily) in patients with ESRD and AF only if they are ≥ 80 years of age or their body weight is ≤ 60 kg.³

Patients with cancer

Cancer-associated acute VTE treatment. Cancer is an established risk factor for acute VTE but it also increases the risk for treatment-­associated bleeding compared with patients without cancer.²² Historically, low-molecular-weight heparin (LMWH) was recommended over warfarin and DOACs for cancer-associated thromboses (CAT).²³ Compared with warfarin, LMWH reduced the rate of recurrent VTE and had similar or reduced bleeding rates at 6 to 12 months.^24-26 However, clinicians and patients often chose warfarin to avoid subcutaneous injections.²⁷

CHEST guidelines recommend oral Xa inhibitors over LMWH for the treatment of CAT.¹³ The 2020 guidelines of the National Institute for Health and Care Excellence (NICE) recommend DOACs as an option for CAT along with LMWH or LMWH transitioned to warfarin.²⁸ The American Society of Clinical Oncology (ASCO) recommends rivaroxaban for acute VTE treatment in CAT. No head-to-head trials have evaluated comparative efficacy of DOACs for CAT. However, edoxaban and rivaroxaban are associated with a greater risk for GI bleeding; therefore, apixaban is preferred in patients with GI malignancies.²⁹ Standard DOAC VTE treatment dosing is recommended for all 3 agents.^2-4

When using DOACs for patients with CAT, consider potential drug-drug interactions with chemotherapy regimens. All DOACs­ are transported by p-glycoprotein, while rivaroxaban and apixaban are substrates of cytochrome P450, leading to potentially significant drug-drug interactions.³⁰ These interactions could affect the patient’s chemotherapeutic regimen, decrease the efficacy of the DOAC, or increase the risk for bleeding. Therefore, anticoagulation choice should be made in collaboration with the ­hematology/oncology team.

Continue to: Cancer-associated VTE prophylaxis...

Cancer-associated VTE prophylaxis. VTE prophylaxis for patients with cancer is complex and necessitates a global assessment of cancer location and treatment regimen and setting. Hospitalized patients receiving chemotherapy are at high risk for VTE if mobility is reduced or if other VTE risk factors are present. The International Initiative on Thrombosis and Cancer (ITAC)³¹ and ISTH³² recommend VTE prophylaxis with unfractionated heparin or LMWH (ISTH recommends LMWH more strongly). The 2020 ASCO Guidelines recommend pharmacologic anticoagulation but make no drug-specific recommendation.²⁹ Parenteral treatment in hospitalized patients is not as burdensome as it is in ambulatory patients; therefore, these recommendations are less likely to elicit inpatient opposition.

In the ambulatory setting, patient avoidance of subcutaneous injections necessitates consideration of DOACs for CAT prophylaxis. The Khorana Risk Score (KRS) is a validated tool (scale, 0-7) to predict VTE risk in ambulatory patients receiving chemotherapy.³³ KRS scores ≥ 2 indicate high thrombotic risk and the need for prophylactic anticoagulation. ASCO recommends apixaban, rivaroxaban, or LMWH.²⁹ ISTH and ITAC both recommend apixaban or rivaroxaban over LMWH.^31,34 An RCT published in June 2023 confirmed that, for adults with cancer and VTE, DOACs were noninferior to LMWH for preventing recurrent VTE for 6 months.³⁵ The recommended doses for apixaban (2.5 mg twice daily) and rivaroxaban (10 mg daily) for CAT VTE prophylaxis are lower than FDA-approved treatment doses.³¹

Patients with thrombophilia: VTE prevention

Thrombophilias are broadly categorized as inherited or acquired, with inherited thrombophilia being more prevalent. The Factor V Leiden (FVL) variant affects 2% to 7% of the population, and prothrombin gene mutation (PGM) affects 1% to 2% of the population.³⁶ Other forms of inherited thrombophilia, such as protein C deficiency, protein S deficiency, and antithrombin deficiency, occur less commonly (< 0.7% of the population).³⁶ Antiphospholipid syndrome (APS), the most common acquired thrombophilia, affects approximately 2% of the population.³⁶ APS involves multiple antibodies: anticardiolipin antibodies, lupus anticoagulant, and anti-beta-2 glycoprotein 1 antibodies. Establishing risk for thrombosis across the varying types of thrombophilia has proven difficult, but APS is considered the most thrombogenic thrombophilia apart from extremely rare homozygous inherited thrombophilias.³⁶ Therefore, DOAC recommendations are thrombophilia specific.

Assessing DOAC blood levels is no longer recommended for patients with obesity, as there is insufficient evidence that these measures affect clinical outcomes.

A prospective cohort study evaluated DOACs compared with heparin/warfarin for VTE treatment in patients with inherited thrombophilias.³⁷ Although all 4 available DOACs were included, most patients (61.1%) received rivaroxaban. Patients with an array of inherited thrombophilias, including rare homozygous mutations, were enrolled in this trial. While most patients (66.9%) had a “mild thrombophilia” defined as either FVL or PGM, the remainder had more severe thrombophilias.³⁷ VTE recurrence was similar and uncommon in the DOAC and heparin/warfarin groups, consistent with a previous meta-analysis.³⁸ Surprisingly, an increase in the cumulative risk for bleeding was seen in the DOAC group compared with the warfarin group, a finding inconsistent with prior trials.³⁸ There were no major bleeding events in the DOAC group, but 3 such events occurred in the heparin/warfarin group, including 2 intracranial hemorrhages.

Currently NICE, CHEST, and ISTH do not make a recommendation for a preferred agent in patients with an acute VTE and inherited thrombophilia; however, DOACs would not be inappropriate.^23,28,32 The American Society of Hematology (ASH) had planned to release recommendations related to the treatment of thrombophilia in 2020, but they were delayed by the COVID-19 pandemic.³⁹

APS presents challenges for acute VTE anticoagulation. First, it causes a strongly thrombogenic state necessitating therapeutic anticoagulation. Second, for patients with positive lupus anticoagulant, INR monitoring and standardized INR goals may be inadequate.⁴⁰ Therefore, using fixed-dose DOACs without the need for therapeutic monitoring is appealing, but significant concerns exist for using DOACs in patients with APS.^41-45 ISTH and CHEST recommend warfarin for the treatment and prevention of acute VTE in patients with APS, especially those with triple-­positive (anticardiolipin, lupus anticoagulant, and anti-beta-2 glycoprotein 1) APS.^13,46 Package labeling for all DOACs recommends avoidance in triple-positive APS.^1-4

ASTRO-APS is the most recent RCT to compare apixaban and warfarin for patients with APS,⁴⁷ and it was terminated early after 6 of 23 patients in the apixaban group had thrombotic events, while no one in the warfarin group had such an event.⁴⁸ Subsequently, a meta-analysis⁴⁹ demonstrated that patients with thrombotic APS appear to have a greater risk for arterial thrombosis when treated with DOACs compared with warfarin. These 2 studies may lead to changes in recommendations to avoid DOACs in all patients with APS or may prompt more focused trials for DOAC use in patients with APS plus an antiplatelet to mitigate arterial thrombotic risk.

Continue to: Expanded clinical indications

Expanded clinical indications

Superficial vein thrombosis

Superficial thrombophlebitis or superficial vein thrombosis (SVT) is estimated to occur 6 times more frequently than VTE.⁵⁰ Management of patients with isolated, uncomplicated thrombophlebitis who are at low risk for extension of the SVT involves symptomatic treatment with nonsteroidal anti-inflammatory drugs, topical agents, or compression therapy. However, depending on risk for progression, anticoagulation may be recommended.⁵¹

Patients at intermediate risk for extension or propagation of SVT are candidates for anticoagulation. The CHEST guidelines recommend fondaparinux 2.5 mg subcutaneous injections daily for 45 days instead of LMWH or warfarin.¹³ However, if patients decline 6 weeks of daily injections, the guidelines acknowledge that rivaroxaban 10 mg daily may be an alternative.¹³

Certain situations should prompt one to consider using a treatment dose of a DOAC for 3 months. These include cases in which the SVT is located within 3 cm of the deep venous system, expands despite an appropriate prophylactic regimen, or recurs after discontinuation of prophylactic anticoagulation.^13,50

Acute coronary syndrome

The American College of Cardiology/­American Heart Association (ACC/AHA) recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.⁵² Data are mixed regarding longer-term anticoagulation in addition to dual antiplatelet therapy in outpatient settings to prevent thrombosis recurrence in the absence of AF.

For patients at intermediate risk for extension of superficial vein thrombosis who decline daily subcutaneous injections of fondaparinux, rivaroxaban 10 mg/d may be an alternative.

The APPRAISE-2 trial enrolled high-risk patients with ACS within 7 days of the event.⁵³ Apixaban 5 mg twice daily was compared with placebo in patients taking aspirin or aspirin plus clopidogrel. The trial was terminated early because major bleeding events increased with apixaban without reduction in recurrent ischemic events. The ATLAS ACS-TIMI 46 trial evaluated different rivaroxaban doses (5-20 mg daily) in ACS patients.⁵⁴ The study revealed possible thrombosis benefit but also increased risk for bleeding, particularly at higher doses. As a result, another study—ATLAS ACS 2-TIMI 51—was conducted and compared the use of low-dose rivaroxaban (2.5 mg twice daily or 5 mg twice daily) vs placebo for patients with recent ACS.⁵⁵ All patients were receiving low-dose aspirin, and approximately 93% of patients in each group also were receiving clopidogrel or ticlopidine. As in the APPRAISE-2 trial, rivaroxaban increased the rate of major bleeding and intracranial hemorrhage; however, it did not increase the incidence of fatal bleeding. Unlike APPRAISE-2, rivaroxaban significantly reduced the primary efficacy end point, a composite of death from cardiovascular causes, myocardial infarction, or stroke (absolute risk reduction = 1.8%; number needed to treat = 56 for combined rivaroxaban doses).⁵⁵

A secondary subgroup analysis combined data from the ATLAS ACMS-TIMI 46 and ATLAS ACS 2-TIMI 51 trials to evaluate outcomes in patients receiving aspirin monotherapy when combined with rivaroxaban 2.5 mg twice daily or 5 mg twice daily or with placebo.⁵⁶ The primary efficacy end point was a composite of cardiovascular death, myocardial infarction, or stroke. When the 2 trials were evaluated separately, neither rivaroxaban dose was associated with reduction of the primary efficacy outcomes compared with aspirin alone. However, when the data were pooled, both the combined rivaroxaban doses (particularly the 5-mg dose) were associated with reduced cardiovascular outcomes. From a safety perspective, the 2.5-mg twice-daily dose of rivaroxaban was the only dose not associated with increased major bleeding risk. Thus, the 2.5-mg twice-daily dose of rivaroxaban may not provide sufficient cardiovascular benefit in patients with ACS, while the larger dose may increase the risk for nonfatal major bleeding events.⁵⁶

The European Medicines Agency⁵⁷ approved rivaroxaban 2.5 mg twice daily for ACS, and the 2020 ESC guidelines⁵⁸ consider it an appropriate therapeutic option in addition to aspirin for patients at high ischemic risk and low bleeding risk. ACS is not an FDA-approved indication for DOACs, and the ACC/AHA Guideline for the Management of ACS, last updated in 2014, does not include DOACs for ACS unless patients have AF.⁵² Ongoing trials are further investigating rivaroxaban for ACS, so the use of DOACs in the post-acute phase of ACS may become clearer in the future.⁵⁹

Continue to: Heparin-induced thrombocytopenia

Heparin-induced thrombocytopenia

Historically, nonheparin parenteral anticoagulants argatroban, bivalirudin, and fondaparinux were recommended for patients at risk for or who had heparin-induced thrombocytopenia (HIT). Argatroban is the only drug FDA approved for the treatment and prophylaxis of HIT; recommendations for the others are based on guideline recommendations.^23,60,61 The nonheparin parenteral anticoagulants cost between $700 and $1500 per day; therefore most patients with HIT are transitioned to warfarin.⁶² However, protein C and S inhibition and a subsequent prothrombotic state conveyed by warfarin initiation necessitates a minimum 5-day bridge to therapeutic warfarin with a nonheparin parenteral anticoagulant.

In vitro tests show that DOACs do not promote development of HIT antibodies⁶³ or affect platelet activation or aggregation.⁶⁴ A literature summary of DOACs for HIT determined that in 104 patients, all but 1 achieved platelet recovery (defined as > 150,000/mcL) within a median time of 7 days. Therapeutically, DOACs prevented new or recurrent VTE in 102/104 cases (98%), and only 3% of patients experienced significant bleeding events.⁶²

The ACC/AHA recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.

The 2018 ASH guidelines for VTE management in HIT include (with very low certainty of evidence) dabigatran, rivaroxaban, or apixaban for consideration in addition to previously recommended nonheparin parenteral anticoagulants.⁶¹ The dosing of each agent is contingent upon treatment of patients with HIT and an acute thrombosis (HITT) or HIT in the absence of VTE. For patients with HITT, treatment doses for acute VTE should be used for the appropriate duration of therapy (ie, 3 months). Importantly, dabigatran requires a 5-day pretreatment period with a parenteral anticoagulant, so it is not an ideal option. When treating isolated HIT (in the absence of VTE), ASH recommends all agents be dosed twice daily—dabigatran 150 mg twice daily (no 5-day parenteral pretreatment necessary), rivaroxaban 15 mg twice daily, or apixaban 5 mg twice daily—until platelet recovery (≥ 150,000/mcL) is achieved.⁶¹

CORRESPONDENCE
Kevin Schleich, PharmD, BCACP, Departments of Pharmaceutical Care and Family Medicine, University of Iowa, 200 Hawkins Drive, 01102-D PFP, Iowa City, IA, 52242; kevin-schleich@uiowa.edu

Four medications comprise the drug category known as direct oral anticoagulants (DOACs). Dabigatran (Pradaxa)¹ was the first to gain approval. It was approved by the US Food and Drug Administration (FDA) in 2010 for the reduction of stroke and systemic embolism in patients with nonvalvular atrial fibrillation (AF). This was followed by approvals for rivaroxaban (Xarelto)² in 2011, apixaban (Eliquis)³ in 2012, and edoxaban (Savaysa)⁴ in 2015. Betrixaban (Bevyxxa)⁵ was approved in 2017 for venous thromboembolism (VTE) prophylaxis in acutely ill hospitalized patients with restricted mobility, but it was removed from the market in 2020.

IMAGE: © KO STUDIOS

In addition to stroke prevention in nonvalvular AF, each DOAC has been approved for other indications and has been addressed further in guideline-based recommendations outside FDA-approved indications. This review highlights the evolving use of DOACs and the expansion of applications for multiple adult patient populations.

Overview of DOACs

Dabigatran is the only direct thrombin inhibitor; the other agents inhibit factor Xa. TABLE 1^1-4 summarizes FDA-­approved indications and dosing and guideline-based dosing. Dabigatran and edoxaban require parenteral anticoagulation for 5 to 10 days prior to initiation for acute VTE, limiting their use.^1,4TABLE 2^1-4 highlights pharmacokinetic differences among the agents. For example, dabigatran is 80% renally cleared, is somewhat dialyzable, and can accumulate in patients with renal dysfunction.¹ Edoxaban is contraindicated for nonvalvular AF in patients with a creatinine clearance (CrCl) > 95 mL/min because an increased stroke risk was demonstrated.⁴ Therefore, rivaroxaban and apixaban are prescribed most often in the United States.^6,7

Applications in special patient populations

Obesity

As of 2020, more than 40% of adults in the United States were obese (body mass index [BMI] ≥ 30), with 9% classified as class 3 or severely obese (BMI ≥ 40).⁸ Altered drug pharmacokinetics in patients with severe obesity raises concern for undertreatment with fixed-dose DOACs. Phase III DOAC approval trials included patients with obesity, but weight cutoffs differed, making extrapolating efficacy and safety data difficult across different obesity stages.⁹ Although no FDA-labeled dosing adjustments exist for patients with obesity, the International Society on Thrombosis and Haemostasis (ISTH) does provide such recommendations.

ISTH changes position on measuring drug levels. ISTH previously recommended avoiding DOACs in those with a BMI > 40 or body weight > 120 kg. If a DOAC was used, ISTH advised obtaining peak and trough drug levels.¹⁰ However, DOAC drug levels have not been associated with clinical outcomes or sufficient degrees of anticoagulation.¹¹

Men and women are affected equally by fibrolipomas. Prevalence does not differ by race or ethnicity.

In April 2021, ISTH updated guidance on DOACs in obesity, indicating standard doses of rivaroxaban or apixaban can be used for the treatment and prevention of VTE in all patients regardless of weight or BMI. Because data in obesity are lacking for dabigatran and edoxaban, avoid using these agents in patients with a BMI > 40 or weight > 120 kg. Additionally, assessing drug levels is no longer recommended, as there is insufficient evidence that these impact clinical outcomes.¹²

The 2021 American College of Chest Physicians (CHEST) guideline update does not recommend adjustments based on weight,¹³ and the 2019 European Society of Cardiology (ESC) guidelines make no mention of weight when treating acute pulmonary embolism.¹⁴

Continue to: Effectiveness of DOACs for AF in patients with obesity isn't clear

Effectiveness of DOACs for AF in patients with obesity isn’t clear, as most data are from retrospective cohort analyses. In patients weighing > 120 kg, dabigatran has shown efficacy in thrombosis prevention similar to that achieved in those weighing ≤ 120 kg, but it has increased the risk for gastrointestinal (GI) bleeding.¹⁵ Another study indicated a 15-mg dose of rivaroxaban may be associated with increased thromboembolic complications in patients with a BMI ≥ 35.¹⁶ Alternatively, another retrospective study of rivaroxaban demonstrated a small absolute risk reduction in ischemic stroke among patients in all stages of obesity and no difference in significant bleeding events.¹⁷ One further retrospective cohort showed that, in patients with a BMI ≥ 50 kg, the effectiveness of rivaroxaban and apixaban in thrombosis prevention and bleeding safety outcomes was comparable to that seen in those with a BMI < 30.¹⁸

As a result of conflicting data, and a lack of prospective randomized controlled trials (RCTs), ISTH continued recommending international normalized ratio (INR)–based dosing of warfarin for class 3 or severely obese patients with AF. The 2018 CHEST guidelines¹⁹ and the 2020 ESC guidelines²⁰ make no mention of DOAC avoidance in patients with obesity and AF.

Advanced and end-stage renal disease

DOACs are renally dosed based on indication, drug-drug interactions, and degree of renal function (TABLE 3^1-4). For example, patients with AF who are anticoagulated with apixaban are prescribed 2.5 mg twice daily when 2 of the 3 following criteria are met: age ≥ 80 years, body weight ≤ 60 kg, serum creatinine ≥ 1.5 mg/dL. However, no dosage adjustment is necessary for VTE treatment or prophylaxis with apixaban regardless of renal function.³

Data supporting the safety and efficacy of DOACs in end-stage renal disease (ESRD) are sparse. All DOACs are renally cleared to varying degrees (TABLE 2^1-4), theoretically increasing bleeding risk as kidney disease progresses. Apixaban is the least renally cleared of the DOACs and has been evaluated in the greatest number of trials for patients with ESRD for both VTE treatment and prevention and nonvalvular AF.²¹ As a result, the FDA approved standard-dose apixaban (5 mg twice daily) for VTE treatment and prevention and nonvalvular AF in patients with ESRD, even those requiring dialysis. Use the reduced apixaban dose (2.5 mg twice daily) in patients with ESRD and AF only if they are ≥ 80 years of age or their body weight is ≤ 60 kg.³

Patients with cancer

Cancer-associated acute VTE treatment. Cancer is an established risk factor for acute VTE but it also increases the risk for treatment-­associated bleeding compared with patients without cancer.²² Historically, low-molecular-weight heparin (LMWH) was recommended over warfarin and DOACs for cancer-associated thromboses (CAT).²³ Compared with warfarin, LMWH reduced the rate of recurrent VTE and had similar or reduced bleeding rates at 6 to 12 months.^24-26 However, clinicians and patients often chose warfarin to avoid subcutaneous injections.²⁷

CHEST guidelines recommend oral Xa inhibitors over LMWH for the treatment of CAT.¹³ The 2020 guidelines of the National Institute for Health and Care Excellence (NICE) recommend DOACs as an option for CAT along with LMWH or LMWH transitioned to warfarin.²⁸ The American Society of Clinical Oncology (ASCO) recommends rivaroxaban for acute VTE treatment in CAT. No head-to-head trials have evaluated comparative efficacy of DOACs for CAT. However, edoxaban and rivaroxaban are associated with a greater risk for GI bleeding; therefore, apixaban is preferred in patients with GI malignancies.²⁹ Standard DOAC VTE treatment dosing is recommended for all 3 agents.^2-4

When using DOACs for patients with CAT, consider potential drug-drug interactions with chemotherapy regimens. All DOACs­ are transported by p-glycoprotein, while rivaroxaban and apixaban are substrates of cytochrome P450, leading to potentially significant drug-drug interactions.³⁰ These interactions could affect the patient’s chemotherapeutic regimen, decrease the efficacy of the DOAC, or increase the risk for bleeding. Therefore, anticoagulation choice should be made in collaboration with the ­hematology/oncology team.

Continue to: Cancer-associated VTE prophylaxis...

Cancer-associated VTE prophylaxis. VTE prophylaxis for patients with cancer is complex and necessitates a global assessment of cancer location and treatment regimen and setting. Hospitalized patients receiving chemotherapy are at high risk for VTE if mobility is reduced or if other VTE risk factors are present. The International Initiative on Thrombosis and Cancer (ITAC)³¹ and ISTH³² recommend VTE prophylaxis with unfractionated heparin or LMWH (ISTH recommends LMWH more strongly). The 2020 ASCO Guidelines recommend pharmacologic anticoagulation but make no drug-specific recommendation.²⁹ Parenteral treatment in hospitalized patients is not as burdensome as it is in ambulatory patients; therefore, these recommendations are less likely to elicit inpatient opposition.

In the ambulatory setting, patient avoidance of subcutaneous injections necessitates consideration of DOACs for CAT prophylaxis. The Khorana Risk Score (KRS) is a validated tool (scale, 0-7) to predict VTE risk in ambulatory patients receiving chemotherapy.³³ KRS scores ≥ 2 indicate high thrombotic risk and the need for prophylactic anticoagulation. ASCO recommends apixaban, rivaroxaban, or LMWH.²⁹ ISTH and ITAC both recommend apixaban or rivaroxaban over LMWH.^31,34 An RCT published in June 2023 confirmed that, for adults with cancer and VTE, DOACs were noninferior to LMWH for preventing recurrent VTE for 6 months.³⁵ The recommended doses for apixaban (2.5 mg twice daily) and rivaroxaban (10 mg daily) for CAT VTE prophylaxis are lower than FDA-approved treatment doses.³¹

Patients with thrombophilia: VTE prevention

Thrombophilias are broadly categorized as inherited or acquired, with inherited thrombophilia being more prevalent. The Factor V Leiden (FVL) variant affects 2% to 7% of the population, and prothrombin gene mutation (PGM) affects 1% to 2% of the population.³⁶ Other forms of inherited thrombophilia, such as protein C deficiency, protein S deficiency, and antithrombin deficiency, occur less commonly (< 0.7% of the population).³⁶ Antiphospholipid syndrome (APS), the most common acquired thrombophilia, affects approximately 2% of the population.³⁶ APS involves multiple antibodies: anticardiolipin antibodies, lupus anticoagulant, and anti-beta-2 glycoprotein 1 antibodies. Establishing risk for thrombosis across the varying types of thrombophilia has proven difficult, but APS is considered the most thrombogenic thrombophilia apart from extremely rare homozygous inherited thrombophilias.³⁶ Therefore, DOAC recommendations are thrombophilia specific.

Assessing DOAC blood levels is no longer recommended for patients with obesity, as there is insufficient evidence that these measures affect clinical outcomes.

A prospective cohort study evaluated DOACs compared with heparin/warfarin for VTE treatment in patients with inherited thrombophilias.³⁷ Although all 4 available DOACs were included, most patients (61.1%) received rivaroxaban. Patients with an array of inherited thrombophilias, including rare homozygous mutations, were enrolled in this trial. While most patients (66.9%) had a “mild thrombophilia” defined as either FVL or PGM, the remainder had more severe thrombophilias.³⁷ VTE recurrence was similar and uncommon in the DOAC and heparin/warfarin groups, consistent with a previous meta-analysis.³⁸ Surprisingly, an increase in the cumulative risk for bleeding was seen in the DOAC group compared with the warfarin group, a finding inconsistent with prior trials.³⁸ There were no major bleeding events in the DOAC group, but 3 such events occurred in the heparin/warfarin group, including 2 intracranial hemorrhages.

Currently NICE, CHEST, and ISTH do not make a recommendation for a preferred agent in patients with an acute VTE and inherited thrombophilia; however, DOACs would not be inappropriate.^23,28,32 The American Society of Hematology (ASH) had planned to release recommendations related to the treatment of thrombophilia in 2020, but they were delayed by the COVID-19 pandemic.³⁹

APS presents challenges for acute VTE anticoagulation. First, it causes a strongly thrombogenic state necessitating therapeutic anticoagulation. Second, for patients with positive lupus anticoagulant, INR monitoring and standardized INR goals may be inadequate.⁴⁰ Therefore, using fixed-dose DOACs without the need for therapeutic monitoring is appealing, but significant concerns exist for using DOACs in patients with APS.^41-45 ISTH and CHEST recommend warfarin for the treatment and prevention of acute VTE in patients with APS, especially those with triple-­positive (anticardiolipin, lupus anticoagulant, and anti-beta-2 glycoprotein 1) APS.^13,46 Package labeling for all DOACs recommends avoidance in triple-positive APS.^1-4

ASTRO-APS is the most recent RCT to compare apixaban and warfarin for patients with APS,⁴⁷ and it was terminated early after 6 of 23 patients in the apixaban group had thrombotic events, while no one in the warfarin group had such an event.⁴⁸ Subsequently, a meta-analysis⁴⁹ demonstrated that patients with thrombotic APS appear to have a greater risk for arterial thrombosis when treated with DOACs compared with warfarin. These 2 studies may lead to changes in recommendations to avoid DOACs in all patients with APS or may prompt more focused trials for DOAC use in patients with APS plus an antiplatelet to mitigate arterial thrombotic risk.

Continue to: Expanded clinical indications

Expanded clinical indications

Superficial vein thrombosis

Superficial thrombophlebitis or superficial vein thrombosis (SVT) is estimated to occur 6 times more frequently than VTE.⁵⁰ Management of patients with isolated, uncomplicated thrombophlebitis who are at low risk for extension of the SVT involves symptomatic treatment with nonsteroidal anti-inflammatory drugs, topical agents, or compression therapy. However, depending on risk for progression, anticoagulation may be recommended.⁵¹

Patients at intermediate risk for extension or propagation of SVT are candidates for anticoagulation. The CHEST guidelines recommend fondaparinux 2.5 mg subcutaneous injections daily for 45 days instead of LMWH or warfarin.¹³ However, if patients decline 6 weeks of daily injections, the guidelines acknowledge that rivaroxaban 10 mg daily may be an alternative.¹³

Certain situations should prompt one to consider using a treatment dose of a DOAC for 3 months. These include cases in which the SVT is located within 3 cm of the deep venous system, expands despite an appropriate prophylactic regimen, or recurs after discontinuation of prophylactic anticoagulation.^13,50

Acute coronary syndrome

The American College of Cardiology/­American Heart Association (ACC/AHA) recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.⁵² Data are mixed regarding longer-term anticoagulation in addition to dual antiplatelet therapy in outpatient settings to prevent thrombosis recurrence in the absence of AF.

For patients at intermediate risk for extension of superficial vein thrombosis who decline daily subcutaneous injections of fondaparinux, rivaroxaban 10 mg/d may be an alternative.

The APPRAISE-2 trial enrolled high-risk patients with ACS within 7 days of the event.⁵³ Apixaban 5 mg twice daily was compared with placebo in patients taking aspirin or aspirin plus clopidogrel. The trial was terminated early because major bleeding events increased with apixaban without reduction in recurrent ischemic events. The ATLAS ACS-TIMI 46 trial evaluated different rivaroxaban doses (5-20 mg daily) in ACS patients.⁵⁴ The study revealed possible thrombosis benefit but also increased risk for bleeding, particularly at higher doses. As a result, another study—ATLAS ACS 2-TIMI 51—was conducted and compared the use of low-dose rivaroxaban (2.5 mg twice daily or 5 mg twice daily) vs placebo for patients with recent ACS.⁵⁵ All patients were receiving low-dose aspirin, and approximately 93% of patients in each group also were receiving clopidogrel or ticlopidine. As in the APPRAISE-2 trial, rivaroxaban increased the rate of major bleeding and intracranial hemorrhage; however, it did not increase the incidence of fatal bleeding. Unlike APPRAISE-2, rivaroxaban significantly reduced the primary efficacy end point, a composite of death from cardiovascular causes, myocardial infarction, or stroke (absolute risk reduction = 1.8%; number needed to treat = 56 for combined rivaroxaban doses).⁵⁵

A secondary subgroup analysis combined data from the ATLAS ACMS-TIMI 46 and ATLAS ACS 2-TIMI 51 trials to evaluate outcomes in patients receiving aspirin monotherapy when combined with rivaroxaban 2.5 mg twice daily or 5 mg twice daily or with placebo.⁵⁶ The primary efficacy end point was a composite of cardiovascular death, myocardial infarction, or stroke. When the 2 trials were evaluated separately, neither rivaroxaban dose was associated with reduction of the primary efficacy outcomes compared with aspirin alone. However, when the data were pooled, both the combined rivaroxaban doses (particularly the 5-mg dose) were associated with reduced cardiovascular outcomes. From a safety perspective, the 2.5-mg twice-daily dose of rivaroxaban was the only dose not associated with increased major bleeding risk. Thus, the 2.5-mg twice-daily dose of rivaroxaban may not provide sufficient cardiovascular benefit in patients with ACS, while the larger dose may increase the risk for nonfatal major bleeding events.⁵⁶

The European Medicines Agency⁵⁷ approved rivaroxaban 2.5 mg twice daily for ACS, and the 2020 ESC guidelines⁵⁸ consider it an appropriate therapeutic option in addition to aspirin for patients at high ischemic risk and low bleeding risk. ACS is not an FDA-approved indication for DOACs, and the ACC/AHA Guideline for the Management of ACS, last updated in 2014, does not include DOACs for ACS unless patients have AF.⁵² Ongoing trials are further investigating rivaroxaban for ACS, so the use of DOACs in the post-acute phase of ACS may become clearer in the future.⁵⁹

Continue to: Heparin-induced thrombocytopenia

Heparin-induced thrombocytopenia

Historically, nonheparin parenteral anticoagulants argatroban, bivalirudin, and fondaparinux were recommended for patients at risk for or who had heparin-induced thrombocytopenia (HIT). Argatroban is the only drug FDA approved for the treatment and prophylaxis of HIT; recommendations for the others are based on guideline recommendations.^23,60,61 The nonheparin parenteral anticoagulants cost between $700 and $1500 per day; therefore most patients with HIT are transitioned to warfarin.⁶² However, protein C and S inhibition and a subsequent prothrombotic state conveyed by warfarin initiation necessitates a minimum 5-day bridge to therapeutic warfarin with a nonheparin parenteral anticoagulant.

In vitro tests show that DOACs do not promote development of HIT antibodies⁶³ or affect platelet activation or aggregation.⁶⁴ A literature summary of DOACs for HIT determined that in 104 patients, all but 1 achieved platelet recovery (defined as > 150,000/mcL) within a median time of 7 days. Therapeutically, DOACs prevented new or recurrent VTE in 102/104 cases (98%), and only 3% of patients experienced significant bleeding events.⁶²

The ACC/AHA recommend combination antiplatelet therapy and anticoagulation for management of acute coronary syndrome in hospitalized patients.

The 2018 ASH guidelines for VTE management in HIT include (with very low certainty of evidence) dabigatran, rivaroxaban, or apixaban for consideration in addition to previously recommended nonheparin parenteral anticoagulants.⁶¹ The dosing of each agent is contingent upon treatment of patients with HIT and an acute thrombosis (HITT) or HIT in the absence of VTE. For patients with HITT, treatment doses for acute VTE should be used for the appropriate duration of therapy (ie, 3 months). Importantly, dabigatran requires a 5-day pretreatment period with a parenteral anticoagulant, so it is not an ideal option. When treating isolated HIT (in the absence of VTE), ASH recommends all agents be dosed twice daily—dabigatran 150 mg twice daily (no 5-day parenteral pretreatment necessary), rivaroxaban 15 mg twice daily, or apixaban 5 mg twice daily—until platelet recovery (≥ 150,000/mcL) is achieved.⁶¹

CORRESPONDENCE
Kevin Schleich, PharmD, BCACP, Departments of Pharmaceutical Care and Family Medicine, University of Iowa, 200 Hawkins Drive, 01102-D PFP, Iowa City, IA, 52242; kevin-schleich@uiowa.edu

Social distancing measures instituted during the ­COVID-19 pandemic challenged the usual way of operating in primary care. To continue delivering medical services, physicians had to transition quickly to forms of remote interaction with patients. Use of technology appeared to be the answer. And it gave clinicians the ability to do what many had long hoped for: offer patients the option of telehealth.

The terms telemedicine and telehealth have similar definitions and are commonly used interchangeably. We think most practices probably would have adopted telehealth earlier were it not for reimbursement barriers. In this article, we adopt the World Health Organization’s definition of telemedicine as: “The delivery of healthcare services, where distance is a critical factor, by all healthcare professionals using information and communication technologies for the exchange of valid information for the diagnosis, treatment, and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”¹

To provide family medicine clinicians with evidence-based recommendations about telehealth, we conducted a critical review of the literature published through April 30, 2021. The scope of this review includes studies found using the PubMed and Google Scholar databases. In addition, we used the keywords “telehealth,” “telemedicine,” “family medicine,” and “primary care.” We divided this review into 6 sections, including focus areas on implementation in primary care, remote diagnostic accuracy, conditions lending themselves to telehealth, physician and patient perceptions, disparities in telehealth, and finally, the conclusions.

Telehealth implementation in primary care

Telehealth in various forms had been around for years before the pandemic, mainly in the form of commercial telehealth businesses. Telehealth was being used in rural and remote areas where it could be difficult to see a primary care provider—let alone a specialist. The family medicine department of the University of Colorado was an early adopter of telehealth and had navigated this transition since 2017, with clinical champions guiding the process. By 2019, 54% of their clinicians were conducting telehealth encounters.²

However, telehealth implementation elsewhere was not accepted so readily. Before the pandemic, a cross-sectional study of more than 1.1 million patients in Northern California showed that 86% preferred in-­person care over video.³ Even as the pandemic began and social distancing measures were implemented, a quality improvement project at a family medicine residency clinic in Florida documented that clinicians still preferred telephone interviews despite the capacity for video visits.⁴ And many primary care systems were simply unprepared to adopt telehealth technologies.

With time, however, family physicians began to improvise using popular videoconferencing technologies (eg, Zoom) that were readily available and familiar to patients, and medical centers began to repurpose their existing videoconferencing systems.⁵ The Ohio State University Wexner Medical Center launched a virtual health initiative just before the pandemic struck, at which time fewer than 5% of patient visits were conducted through telehealth. Weeks later, nearly 93% of patient visits were offered through telehealth.⁶

Reimbursement. Another significant impediment to early telehealth uptake was the late reaction by the Centers for Medicare and Medicaid Services (CMS) in changing the payment system. Hectic expansion of telehealth in response to the crisis pointed to the lack of policies that supported primary care with payments based on outcomes rather than fee-for-service models.⁷ By the end of April 2020, CMS finally announced that video visits would be reimbursed at the same rate as in-person visits. However, telephone-only visits are still very limited in coverage, and appropriate codes should be verified with payers.

Continue to: Remote diagnosis comes with a caveat

Remote diagnosis comes with a caveat

Some primary care practices have found that images of skin lesions submitted by patients (usually by cell phone) suffice for accurate diagnosis in lieu of office visits.⁸ With chronic conditions, home-based remote monitoring of vital signs may assist in diagnosing and managing acute issues. More efficient triage of patients is increasingly possible with the receipt of still images or video files of concerning lesions (eg, burns, rash, chronic wounds) sent from smartphones alone^9,10 or with devices attached to smartphones (eg, parent-managed otoscopes).^11,12

CMS reimburses for video visits at the same rate it does for in-person visits, but telephone-only visits are still limited in coverage.

Family physicians historically have relied on in-person visits for holistic assessment and diagnosis. Telehealth video visits have the potential to assist with this goal, but there are risks. For example, one patient cut her foot while swimming and the wound became infected. In a video telehealth visit with a physician assistant, the patient was prescribed an oral antibiotic for cellulitis. However, redness and swelling of the wound continued to increase. Subsequent messages left by the patient went unheeded, and she then visited her local emergency department where she received intravenous antibiotics.¹³ Another patient, who had cervical neck discomfort, had a video visit with an urgent care clinic. The initial diagnosis was sciatica; however, the pain continued with the addition of chills, sweats, and subjective fever.¹⁴ Physical examination at the hospital and lab tests revealed endocarditis. These case reports show the limitations of video visits.

Specific conditions usually suitable for telehealth evaluation

The pandemic helped us understand that some situations and conditions are better suited than others to coverage by telehealth. The National Ambulatory Medical Care Survey examined 850 million patient–­physician encounters and found that 66% of all ambulatory primary care visits required in-­office care,¹⁵ suggesting that about one-third of patient encounters could be treated via ­telehealth.

As an example, our southeastern Wisconsin urban clinic has about 20,000 office visits per year. We launched telehealth in March 2020 in direct response to the pandemic. Telehealth usage peaked at the beginning of the pandemic (FIGURE), fell gradually, hit a lower peak in November and December as COVID case counts increased, and then decreased again as our community changed from a “quarantine/lockdown” mentality to “opening up/back to new normal.”

Some conditions can be managed favorably with the telehealth format:

Infectious diseases may be treatable remotely.^16,17 Following an initial telehealth visit, the physician can evaluate and recommend further care.

Stable, chronic conditions. Telehealth can be used for stable, chronic conditions such as diabetes, chronic obstructive pulmonary disease, and heart failure when lab or imaging studies are not needed.¹⁸

Mental health. Telehealth can be useful in counseling and providing mental health and social support.¹⁸ Safeguards can be put in place to protect patient privacy in this setting.¹⁹

Behavioral change. Telehealth can be effective in providing support for patients actively trying to quit smoking or lose weight, and for caregivers. A physician who “checks in” can be a positive motivator and can promote a patient’s continued success.²⁰

Continue to: Telehealth is less beneficial...

Telehealth is less beneficial when a physical exam is needed to assess pain, tenderness, strength, or other sensations. Office visits also are required for lab assays and imaging, as in periodic checks of A1C levels in patients with diabetes. As technology advances, home-based laboratory kits and sensors likely will change this picture. New patients may be better served through an initial office visit to develop the patient–physician relationship.

Visual assessment of conditions may be limited by telehealth depending on the quality of the devices used. For example, rashes may be difficult to assess given the clarity of the picture on the device and the ability to see only in 2D. There is still a need for more controlled trials to clarify which conditions can be evaluated and managed by telehealth and which ones need in-person care.²¹

Physician and patient perceptions of telehealth encounters

Research into family physicians’ perceptions of telehealth is scant. However, 3 studies published in 2021 reveal some advantages and challenges for telehealth adoption.

• A qualitative study found that physicians valued the increased access to care for some patients, changes to reimbursement practices not covered before, and the opportunity to see patients’ home environments.²² Disadvantages included an inability to examine the patient, problems with diagnostic accuracy, hindrances to developing personal connections, and the potential for burnout with on-demand care.²² The researchers suggested that telehealth might better serve to augment in-person care.
• A second study found that clinicians are satisfied with the use of telehealth in general. However, it also noted that the lack of physical examination could hinder accurate diagnosis and treatment.²³
• A third study surveyed 109 family physicians, reinforcing the importance of physical exams and highlighting the lack of body language as another barrier.²⁴

In addition, all 3 studies noted that video visits are typically briefer than in-person visits. Previous research predominantly done in specialty and mental health care showed that the benefits of telehealth for physicians include an increase in efficiency, reduced commute time, and improved work-life balance.²⁵

Patient perspectives. Many patients have reported that they prefer telehealth because of lower costs, decreased travel time, and faster health care access.^26,27 However, patients also have expressed concerns that the telehealth environment may reduce physician attention, can limit personal interaction (and impart a sense of being rushed), and lacks the physical examination that may be key to an adequate diagnosis.²⁸

Continue to: A survey of 223 patients showed...

A survey of 223 patients showed that sicker patients choose in-person care because they want more in-depth visits with more attention to detail than healthier patients do.²⁹ In a Veterans Affairs health care system qualitative study, patients voiced concerns about communicating with physicians via telehealth, including the potential for errors, less attention paid to their needs, audio difficulties, and challenges to establishing a physician–patient relationship.³⁰ Some patients thought telehealth inhibited their personal expression or that the clinician was not attentive enough. These patient reports underscore the importance of patient–clinician relationships developed in person.³¹ The perceived level of complexity involved in a visit appears to be an essential factor in a patient opting for telehealth—or not.

Efficient triage of patients is increasingly possible with still images or videos sent from smartphones.

In light of these known physician and patient perspectives, it seems wise to develop a hybrid model approach in which visits alternate between telehealth and office.

Patient disparities that may limit the use of telehealth

Race and ethnicity is a major factor in telehealth use. Patients who are Black or Hispanic use telehealth services less often than patients who are White.^32,33 A study that looked at patients with chronic conditions—hypertension and diabetes—that disproportionately affect Black and Hispanic patients found that patients in these populations with either of these conditions had a lower prevalence of Internet use when compared with White patients.³⁴ However, subpopulations can vary in their usage. For example, a study in East Harlem, New York, found that Hispanic pregnant women used telehealth frequently for prenatal care and perceived the care as satisfactory.³⁵

Age is also a significant variable in the adoption of telehealth, with pre-­COVID-19 studies finding lower use of technology among older adults. However, a study performed at the University of Missouri during the first months of the pandemic found an increase in telehealth use in seniors,³² although the increase was in telephone use and not full video sessions.

Many patients in need of health care services may have older devices and/or low-speed or no Internet access; they also may lack the technical know-how to conduct a telehealth visit.^4,36 For example, regardless of race or ethnicity, patients on government insurance (Medicaid and Medicare) have been shown to complete more telephone than video visits,³⁷ underscoring the importance of telehealth practice flexibility and the need for increased technology support to decrease the digital divide. Even with adequate technological support and patient training, telehealth may be more complicated if patients have such comorbidities as hearing, visual, or cognitive impairment.³¹ Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.³⁸

Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.

Overall, incorporating telehealth for the care of older and/or vulnerable patients will present a unique set of challenges that organizations must address. Efforts must be made to understand the available technologies and patients’ comfort in using them. A hybrid model offering telehealth and in-office encounters may be the best solution.

CORRESPONDENCE
Hernan Barenboim, PhD, KPC Health Group, 301 North San Jacinto Street, Hemet, CA 92543; hbarenboim@gmail.com

Social distancing measures instituted during the ­COVID-19 pandemic challenged the usual way of operating in primary care. To continue delivering medical services, physicians had to transition quickly to forms of remote interaction with patients. Use of technology appeared to be the answer. And it gave clinicians the ability to do what many had long hoped for: offer patients the option of telehealth.

The terms telemedicine and telehealth have similar definitions and are commonly used interchangeably. We think most practices probably would have adopted telehealth earlier were it not for reimbursement barriers. In this article, we adopt the World Health Organization’s definition of telemedicine as: “The delivery of healthcare services, where distance is a critical factor, by all healthcare professionals using information and communication technologies for the exchange of valid information for the diagnosis, treatment, and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”¹

To provide family medicine clinicians with evidence-based recommendations about telehealth, we conducted a critical review of the literature published through April 30, 2021. The scope of this review includes studies found using the PubMed and Google Scholar databases. In addition, we used the keywords “telehealth,” “telemedicine,” “family medicine,” and “primary care.” We divided this review into 6 sections, including focus areas on implementation in primary care, remote diagnostic accuracy, conditions lending themselves to telehealth, physician and patient perceptions, disparities in telehealth, and finally, the conclusions.

Telehealth implementation in primary care

Telehealth in various forms had been around for years before the pandemic, mainly in the form of commercial telehealth businesses. Telehealth was being used in rural and remote areas where it could be difficult to see a primary care provider—let alone a specialist. The family medicine department of the University of Colorado was an early adopter of telehealth and had navigated this transition since 2017, with clinical champions guiding the process. By 2019, 54% of their clinicians were conducting telehealth encounters.²

However, telehealth implementation elsewhere was not accepted so readily. Before the pandemic, a cross-sectional study of more than 1.1 million patients in Northern California showed that 86% preferred in-­person care over video.³ Even as the pandemic began and social distancing measures were implemented, a quality improvement project at a family medicine residency clinic in Florida documented that clinicians still preferred telephone interviews despite the capacity for video visits.⁴ And many primary care systems were simply unprepared to adopt telehealth technologies.

With time, however, family physicians began to improvise using popular videoconferencing technologies (eg, Zoom) that were readily available and familiar to patients, and medical centers began to repurpose their existing videoconferencing systems.⁵ The Ohio State University Wexner Medical Center launched a virtual health initiative just before the pandemic struck, at which time fewer than 5% of patient visits were conducted through telehealth. Weeks later, nearly 93% of patient visits were offered through telehealth.⁶

Reimbursement. Another significant impediment to early telehealth uptake was the late reaction by the Centers for Medicare and Medicaid Services (CMS) in changing the payment system. Hectic expansion of telehealth in response to the crisis pointed to the lack of policies that supported primary care with payments based on outcomes rather than fee-for-service models.⁷ By the end of April 2020, CMS finally announced that video visits would be reimbursed at the same rate as in-person visits. However, telephone-only visits are still very limited in coverage, and appropriate codes should be verified with payers.

Continue to: Remote diagnosis comes with a caveat

Remote diagnosis comes with a caveat

Some primary care practices have found that images of skin lesions submitted by patients (usually by cell phone) suffice for accurate diagnosis in lieu of office visits.⁸ With chronic conditions, home-based remote monitoring of vital signs may assist in diagnosing and managing acute issues. More efficient triage of patients is increasingly possible with the receipt of still images or video files of concerning lesions (eg, burns, rash, chronic wounds) sent from smartphones alone^9,10 or with devices attached to smartphones (eg, parent-managed otoscopes).^11,12

CMS reimburses for video visits at the same rate it does for in-person visits, but telephone-only visits are still limited in coverage.

Family physicians historically have relied on in-person visits for holistic assessment and diagnosis. Telehealth video visits have the potential to assist with this goal, but there are risks. For example, one patient cut her foot while swimming and the wound became infected. In a video telehealth visit with a physician assistant, the patient was prescribed an oral antibiotic for cellulitis. However, redness and swelling of the wound continued to increase. Subsequent messages left by the patient went unheeded, and she then visited her local emergency department where she received intravenous antibiotics.¹³ Another patient, who had cervical neck discomfort, had a video visit with an urgent care clinic. The initial diagnosis was sciatica; however, the pain continued with the addition of chills, sweats, and subjective fever.¹⁴ Physical examination at the hospital and lab tests revealed endocarditis. These case reports show the limitations of video visits.

Specific conditions usually suitable for telehealth evaluation

The pandemic helped us understand that some situations and conditions are better suited than others to coverage by telehealth. The National Ambulatory Medical Care Survey examined 850 million patient–­physician encounters and found that 66% of all ambulatory primary care visits required in-­office care,¹⁵ suggesting that about one-third of patient encounters could be treated via ­telehealth.

As an example, our southeastern Wisconsin urban clinic has about 20,000 office visits per year. We launched telehealth in March 2020 in direct response to the pandemic. Telehealth usage peaked at the beginning of the pandemic (FIGURE), fell gradually, hit a lower peak in November and December as COVID case counts increased, and then decreased again as our community changed from a “quarantine/lockdown” mentality to “opening up/back to new normal.”

Some conditions can be managed favorably with the telehealth format:

Infectious diseases may be treatable remotely.^16,17 Following an initial telehealth visit, the physician can evaluate and recommend further care.

Stable, chronic conditions. Telehealth can be used for stable, chronic conditions such as diabetes, chronic obstructive pulmonary disease, and heart failure when lab or imaging studies are not needed.¹⁸

Mental health. Telehealth can be useful in counseling and providing mental health and social support.¹⁸ Safeguards can be put in place to protect patient privacy in this setting.¹⁹

Behavioral change. Telehealth can be effective in providing support for patients actively trying to quit smoking or lose weight, and for caregivers. A physician who “checks in” can be a positive motivator and can promote a patient’s continued success.²⁰

Continue to: Telehealth is less beneficial...

Telehealth is less beneficial when a physical exam is needed to assess pain, tenderness, strength, or other sensations. Office visits also are required for lab assays and imaging, as in periodic checks of A1C levels in patients with diabetes. As technology advances, home-based laboratory kits and sensors likely will change this picture. New patients may be better served through an initial office visit to develop the patient–physician relationship.

Visual assessment of conditions may be limited by telehealth depending on the quality of the devices used. For example, rashes may be difficult to assess given the clarity of the picture on the device and the ability to see only in 2D. There is still a need for more controlled trials to clarify which conditions can be evaluated and managed by telehealth and which ones need in-person care.²¹

Physician and patient perceptions of telehealth encounters

Research into family physicians’ perceptions of telehealth is scant. However, 3 studies published in 2021 reveal some advantages and challenges for telehealth adoption.

• A qualitative study found that physicians valued the increased access to care for some patients, changes to reimbursement practices not covered before, and the opportunity to see patients’ home environments.²² Disadvantages included an inability to examine the patient, problems with diagnostic accuracy, hindrances to developing personal connections, and the potential for burnout with on-demand care.²² The researchers suggested that telehealth might better serve to augment in-person care.
• A second study found that clinicians are satisfied with the use of telehealth in general. However, it also noted that the lack of physical examination could hinder accurate diagnosis and treatment.²³
• A third study surveyed 109 family physicians, reinforcing the importance of physical exams and highlighting the lack of body language as another barrier.²⁴

In addition, all 3 studies noted that video visits are typically briefer than in-person visits. Previous research predominantly done in specialty and mental health care showed that the benefits of telehealth for physicians include an increase in efficiency, reduced commute time, and improved work-life balance.²⁵

Patient perspectives. Many patients have reported that they prefer telehealth because of lower costs, decreased travel time, and faster health care access.^26,27 However, patients also have expressed concerns that the telehealth environment may reduce physician attention, can limit personal interaction (and impart a sense of being rushed), and lacks the physical examination that may be key to an adequate diagnosis.²⁸

Continue to: A survey of 223 patients showed...

A survey of 223 patients showed that sicker patients choose in-person care because they want more in-depth visits with more attention to detail than healthier patients do.²⁹ In a Veterans Affairs health care system qualitative study, patients voiced concerns about communicating with physicians via telehealth, including the potential for errors, less attention paid to their needs, audio difficulties, and challenges to establishing a physician–patient relationship.³⁰ Some patients thought telehealth inhibited their personal expression or that the clinician was not attentive enough. These patient reports underscore the importance of patient–clinician relationships developed in person.³¹ The perceived level of complexity involved in a visit appears to be an essential factor in a patient opting for telehealth—or not.

Efficient triage of patients is increasingly possible with still images or videos sent from smartphones.

In light of these known physician and patient perspectives, it seems wise to develop a hybrid model approach in which visits alternate between telehealth and office.

Patient disparities that may limit the use of telehealth

Race and ethnicity is a major factor in telehealth use. Patients who are Black or Hispanic use telehealth services less often than patients who are White.^32,33 A study that looked at patients with chronic conditions—hypertension and diabetes—that disproportionately affect Black and Hispanic patients found that patients in these populations with either of these conditions had a lower prevalence of Internet use when compared with White patients.³⁴ However, subpopulations can vary in their usage. For example, a study in East Harlem, New York, found that Hispanic pregnant women used telehealth frequently for prenatal care and perceived the care as satisfactory.³⁵

Age is also a significant variable in the adoption of telehealth, with pre-­COVID-19 studies finding lower use of technology among older adults. However, a study performed at the University of Missouri during the first months of the pandemic found an increase in telehealth use in seniors,³² although the increase was in telephone use and not full video sessions.

Many patients in need of health care services may have older devices and/or low-speed or no Internet access; they also may lack the technical know-how to conduct a telehealth visit.^4,36 For example, regardless of race or ethnicity, patients on government insurance (Medicaid and Medicare) have been shown to complete more telephone than video visits,³⁷ underscoring the importance of telehealth practice flexibility and the need for increased technology support to decrease the digital divide. Even with adequate technological support and patient training, telehealth may be more complicated if patients have such comorbidities as hearing, visual, or cognitive impairment.³¹ Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.³⁸

Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.

Overall, incorporating telehealth for the care of older and/or vulnerable patients will present a unique set of challenges that organizations must address. Efforts must be made to understand the available technologies and patients’ comfort in using them. A hybrid model offering telehealth and in-office encounters may be the best solution.

CORRESPONDENCE
Hernan Barenboim, PhD, KPC Health Group, 301 North San Jacinto Street, Hemet, CA 92543; hbarenboim@gmail.com

Social distancing measures instituted during the ­COVID-19 pandemic challenged the usual way of operating in primary care. To continue delivering medical services, physicians had to transition quickly to forms of remote interaction with patients. Use of technology appeared to be the answer. And it gave clinicians the ability to do what many had long hoped for: offer patients the option of telehealth.

The terms telemedicine and telehealth have similar definitions and are commonly used interchangeably. We think most practices probably would have adopted telehealth earlier were it not for reimbursement barriers. In this article, we adopt the World Health Organization’s definition of telemedicine as: “The delivery of healthcare services, where distance is a critical factor, by all healthcare professionals using information and communication technologies for the exchange of valid information for the diagnosis, treatment, and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”¹

To provide family medicine clinicians with evidence-based recommendations about telehealth, we conducted a critical review of the literature published through April 30, 2021. The scope of this review includes studies found using the PubMed and Google Scholar databases. In addition, we used the keywords “telehealth,” “telemedicine,” “family medicine,” and “primary care.” We divided this review into 6 sections, including focus areas on implementation in primary care, remote diagnostic accuracy, conditions lending themselves to telehealth, physician and patient perceptions, disparities in telehealth, and finally, the conclusions.

Telehealth implementation in primary care

Telehealth in various forms had been around for years before the pandemic, mainly in the form of commercial telehealth businesses. Telehealth was being used in rural and remote areas where it could be difficult to see a primary care provider—let alone a specialist. The family medicine department of the University of Colorado was an early adopter of telehealth and had navigated this transition since 2017, with clinical champions guiding the process. By 2019, 54% of their clinicians were conducting telehealth encounters.²

However, telehealth implementation elsewhere was not accepted so readily. Before the pandemic, a cross-sectional study of more than 1.1 million patients in Northern California showed that 86% preferred in-­person care over video.³ Even as the pandemic began and social distancing measures were implemented, a quality improvement project at a family medicine residency clinic in Florida documented that clinicians still preferred telephone interviews despite the capacity for video visits.⁴ And many primary care systems were simply unprepared to adopt telehealth technologies.

With time, however, family physicians began to improvise using popular videoconferencing technologies (eg, Zoom) that were readily available and familiar to patients, and medical centers began to repurpose their existing videoconferencing systems.⁵ The Ohio State University Wexner Medical Center launched a virtual health initiative just before the pandemic struck, at which time fewer than 5% of patient visits were conducted through telehealth. Weeks later, nearly 93% of patient visits were offered through telehealth.⁶

Reimbursement. Another significant impediment to early telehealth uptake was the late reaction by the Centers for Medicare and Medicaid Services (CMS) in changing the payment system. Hectic expansion of telehealth in response to the crisis pointed to the lack of policies that supported primary care with payments based on outcomes rather than fee-for-service models.⁷ By the end of April 2020, CMS finally announced that video visits would be reimbursed at the same rate as in-person visits. However, telephone-only visits are still very limited in coverage, and appropriate codes should be verified with payers.

Continue to: Remote diagnosis comes with a caveat

Remote diagnosis comes with a caveat

Some primary care practices have found that images of skin lesions submitted by patients (usually by cell phone) suffice for accurate diagnosis in lieu of office visits.⁸ With chronic conditions, home-based remote monitoring of vital signs may assist in diagnosing and managing acute issues. More efficient triage of patients is increasingly possible with the receipt of still images or video files of concerning lesions (eg, burns, rash, chronic wounds) sent from smartphones alone^9,10 or with devices attached to smartphones (eg, parent-managed otoscopes).^11,12

CMS reimburses for video visits at the same rate it does for in-person visits, but telephone-only visits are still limited in coverage.

Family physicians historically have relied on in-person visits for holistic assessment and diagnosis. Telehealth video visits have the potential to assist with this goal, but there are risks. For example, one patient cut her foot while swimming and the wound became infected. In a video telehealth visit with a physician assistant, the patient was prescribed an oral antibiotic for cellulitis. However, redness and swelling of the wound continued to increase. Subsequent messages left by the patient went unheeded, and she then visited her local emergency department where she received intravenous antibiotics.¹³ Another patient, who had cervical neck discomfort, had a video visit with an urgent care clinic. The initial diagnosis was sciatica; however, the pain continued with the addition of chills, sweats, and subjective fever.¹⁴ Physical examination at the hospital and lab tests revealed endocarditis. These case reports show the limitations of video visits.

Specific conditions usually suitable for telehealth evaluation

The pandemic helped us understand that some situations and conditions are better suited than others to coverage by telehealth. The National Ambulatory Medical Care Survey examined 850 million patient–­physician encounters and found that 66% of all ambulatory primary care visits required in-­office care,¹⁵ suggesting that about one-third of patient encounters could be treated via ­telehealth.

As an example, our southeastern Wisconsin urban clinic has about 20,000 office visits per year. We launched telehealth in March 2020 in direct response to the pandemic. Telehealth usage peaked at the beginning of the pandemic (FIGURE), fell gradually, hit a lower peak in November and December as COVID case counts increased, and then decreased again as our community changed from a “quarantine/lockdown” mentality to “opening up/back to new normal.”

Some conditions can be managed favorably with the telehealth format:

Infectious diseases may be treatable remotely.^16,17 Following an initial telehealth visit, the physician can evaluate and recommend further care.

Stable, chronic conditions. Telehealth can be used for stable, chronic conditions such as diabetes, chronic obstructive pulmonary disease, and heart failure when lab or imaging studies are not needed.¹⁸

Mental health. Telehealth can be useful in counseling and providing mental health and social support.¹⁸ Safeguards can be put in place to protect patient privacy in this setting.¹⁹

Behavioral change. Telehealth can be effective in providing support for patients actively trying to quit smoking or lose weight, and for caregivers. A physician who “checks in” can be a positive motivator and can promote a patient’s continued success.²⁰

Continue to: Telehealth is less beneficial...

Telehealth is less beneficial when a physical exam is needed to assess pain, tenderness, strength, or other sensations. Office visits also are required for lab assays and imaging, as in periodic checks of A1C levels in patients with diabetes. As technology advances, home-based laboratory kits and sensors likely will change this picture. New patients may be better served through an initial office visit to develop the patient–physician relationship.

Visual assessment of conditions may be limited by telehealth depending on the quality of the devices used. For example, rashes may be difficult to assess given the clarity of the picture on the device and the ability to see only in 2D. There is still a need for more controlled trials to clarify which conditions can be evaluated and managed by telehealth and which ones need in-person care.²¹

Physician and patient perceptions of telehealth encounters

Research into family physicians’ perceptions of telehealth is scant. However, 3 studies published in 2021 reveal some advantages and challenges for telehealth adoption.

• A qualitative study found that physicians valued the increased access to care for some patients, changes to reimbursement practices not covered before, and the opportunity to see patients’ home environments.²² Disadvantages included an inability to examine the patient, problems with diagnostic accuracy, hindrances to developing personal connections, and the potential for burnout with on-demand care.²² The researchers suggested that telehealth might better serve to augment in-person care.
• A second study found that clinicians are satisfied with the use of telehealth in general. However, it also noted that the lack of physical examination could hinder accurate diagnosis and treatment.²³
• A third study surveyed 109 family physicians, reinforcing the importance of physical exams and highlighting the lack of body language as another barrier.²⁴

In addition, all 3 studies noted that video visits are typically briefer than in-person visits. Previous research predominantly done in specialty and mental health care showed that the benefits of telehealth for physicians include an increase in efficiency, reduced commute time, and improved work-life balance.²⁵

Patient perspectives. Many patients have reported that they prefer telehealth because of lower costs, decreased travel time, and faster health care access.^26,27 However, patients also have expressed concerns that the telehealth environment may reduce physician attention, can limit personal interaction (and impart a sense of being rushed), and lacks the physical examination that may be key to an adequate diagnosis.²⁸

Continue to: A survey of 223 patients showed...

A survey of 223 patients showed that sicker patients choose in-person care because they want more in-depth visits with more attention to detail than healthier patients do.²⁹ In a Veterans Affairs health care system qualitative study, patients voiced concerns about communicating with physicians via telehealth, including the potential for errors, less attention paid to their needs, audio difficulties, and challenges to establishing a physician–patient relationship.³⁰ Some patients thought telehealth inhibited their personal expression or that the clinician was not attentive enough. These patient reports underscore the importance of patient–clinician relationships developed in person.³¹ The perceived level of complexity involved in a visit appears to be an essential factor in a patient opting for telehealth—or not.

Efficient triage of patients is increasingly possible with still images or videos sent from smartphones.

In light of these known physician and patient perspectives, it seems wise to develop a hybrid model approach in which visits alternate between telehealth and office.

Patient disparities that may limit the use of telehealth

Race and ethnicity is a major factor in telehealth use. Patients who are Black or Hispanic use telehealth services less often than patients who are White.^32,33 A study that looked at patients with chronic conditions—hypertension and diabetes—that disproportionately affect Black and Hispanic patients found that patients in these populations with either of these conditions had a lower prevalence of Internet use when compared with White patients.³⁴ However, subpopulations can vary in their usage. For example, a study in East Harlem, New York, found that Hispanic pregnant women used telehealth frequently for prenatal care and perceived the care as satisfactory.³⁵

Age is also a significant variable in the adoption of telehealth, with pre-­COVID-19 studies finding lower use of technology among older adults. However, a study performed at the University of Missouri during the first months of the pandemic found an increase in telehealth use in seniors,³² although the increase was in telephone use and not full video sessions.

Many patients in need of health care services may have older devices and/or low-speed or no Internet access; they also may lack the technical know-how to conduct a telehealth visit.^4,36 For example, regardless of race or ethnicity, patients on government insurance (Medicaid and Medicare) have been shown to complete more telephone than video visits,³⁷ underscoring the importance of telehealth practice flexibility and the need for increased technology support to decrease the digital divide. Even with adequate technological support and patient training, telehealth may be more complicated if patients have such comorbidities as hearing, visual, or cognitive impairment.³¹ Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.³⁸

Patients from a lower socioeconomic status may feel uncomfortable with providers seeing their home environment on video.

Overall, incorporating telehealth for the care of older and/or vulnerable patients will present a unique set of challenges that organizations must address. Efforts must be made to understand the available technologies and patients’ comfort in using them. A hybrid model offering telehealth and in-office encounters may be the best solution.

CORRESPONDENCE
Hernan Barenboim, PhD, KPC Health Group, 301 North San Jacinto Street, Hemet, CA 92543; hbarenboim@gmail.com

Nearly 40 antihyperglycemic agents have been approved by the US Food and Drug Administration (FDA) since the approval of human insulin in 1982.¹ In addition, existing antihyperglycemic medications are constantly gaining FDA approval for new indications for common type 2 diabetes (T2D) comorbidities. For example, in addition to their glycemic benefits, the sodium-glucose cotransporter-2 (SGLT2) inhibitors have been approved for use in patients with T2D and established atherosclerotic cardiovascular disease (ASCVD) to reduce the risk for major adverse cardiovascular events (MACE; canagliflozin), risk for hospitalization for heart failure (dapagliflozin), and cardiovascular death (empagliflozin).^2-4

The plethora of new agents and new data for existing agents, coupled with the annual release of guidelines from the American Diabetes Association (ADA) and practice recommendations from several other professional organizations,^5-7 make it challenging for family physicians to stay current and provide the most up-to-date, evidence-based care. In this article, we provide advice on how to approach the screening, diagnosis, and evaluation of T2D, and on how to manage newly diagnosed T2D.

Screening, Dx, and evaluation: A quick review

Screening

Screening recommendations vary among professional organizations (TABLE 1^5,6,8). The US Preventive Services Task Force (USPSTF) recommends screening adults ages 35 to 70 years who are overweight or obese. Clinicians also can consider screening patients with a higher risk for diabetes.⁵ The ADA suggests screening all adults starting at 35 years, regardless of risk factors.⁸ Asymptomatic adults of any age with overweight or obesity and 1 or more risk factors should be screened.⁸The American Association of Clinical Endocrinology (AACE) recommends screening adults of any age who have risk factors.⁶ If the screening result is normal, repeat testing in 3 years is appropriate, unless there is a change in symptoms or risks.^5,8 Annual testing can be considered in patients with ≥ 2 risk factors or with prediabetes (glycosylated hemoglobin [A1C] ≥ 5.7%).^6,8

Making the diagnosis

The initial diagnosis of diabetes can be made by a fasting plasma glucose level ≥ 126 mg/dL (7.0 mmol/L); a 2-hour plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) following an oral glucose tolerance test; or an A1C level ≥ 6.5%. Prioritize lab-drawn A1C measurements over point-of-care tests to diagnose T2D. In ­patients with classic symptoms of hyper­glycemia, a random plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) is also diagnostic. Generally, these tests are considered equally appropriate in screening for diabetes and may be used to detect prediabetes. In the absence of clear symptoms of hyperglycemia, the diagnosis of diabetes requires 2 abnormal screening test results, either via 1 blood sample (such as an abnormal A1C and glucose) or 2 separate blood samples of the same test. Further evaluation is advised if there is discordance between the 2 samples.⁸

Extended evaluations

Patients with newly diagnosed T2D require a thorough evaluation for comorbidities and complications of diabetes. Refer patients to an ophthalmologist for a dilated eye examination, with subsequent exams occurring every 1 to 2 years.^6,9 Additional referrals for diabetes education, family planning for women of reproductive age, and dental, social, or mental health services may be clinically appropriate.⁹

Setting goals for glycemic control

Glycemic control is commonly monitored by the A1C level and by blood glucose monitoring either through traditional point-of-care glucometers or continuous glucose monitors (CGMs).¹⁰ Generally, CGMs provide more glycemic data than traditional glucometers and may cue patients to choose healthier dietary options and engage in physical exercise.¹¹ Patients with T2D who use CGMs exhibit lower A1Cs, greater time in glycemic range, and reduced hypoglycemic episodes.¹¹ Generally, CGMs are reserved for patients with type 1 diabetes and patients with T2D who use multiple daily injections, subcutaneous insulin infusions, or basal insulin only.¹² Most professional organizations recommend that clinicians consider patient-specific factors to set individualized glycemic goals.^6,10,13,14 For example, more stringent glycemic goals could be pursued for patients with longer life expectancy, shorter disease duration, absence of complications (eg, nephropathy, neuropathy, or cardiovascular disease), fewer comorbid conditions, lower hypoglycemia risk, or higher cognitive function.⁶

With newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.

More specific A1C goals vary by professional organization. For nonpregnant adults, the ADA recommends an A1C goal of < 7% and a preprandial blood glucose level of 80 to 130 mg/dL (4.4-7.2 mmol/L).¹⁰ However, a lower A1C goal may be appropriate if it can be attained safely without causing hypoglycemia or other adverse effects.¹⁰ The AACE suggests an A1C goal of ≤ 6.5% and a fasting blood glucose level of < 110 mg/dL when it can be achieved safely.⁶ More stringent A1C goals may reduce long-term micro- and macrovascular complications—especially in patients with newly diagnosed T2D.¹⁰ While older studies such as the ACCORD trial found increased mortality in groups with more stringent glycemic targets, they did not include newer agents (SGLT2 inhibitors or glucagon-like peptide-1 [GLP-1] receptor agonists) that reduce cardiovascular events by mechanisms outside their glycemic-lowering effect. With these newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.¹⁰ Both the ADA and AACE recommend a less stringent A1C goal of 7% to 8% for patients with limited life expectancy or risks (eg, a history of hypoglycemia) that outweigh expected benefits.^6,10

Continue to: Lifestyle modifications

Nutrition

The energy-dense Western diet, combined with sedentary behavior, are thought to be a primary cause of T2D.¹⁵ Therefore, include lifestyle modifications in the initial management of newly diagnosed T2D. Diets that replace carbohydrates with saturated and trans fats are related to increased mortality in patients with T2D.¹⁶ Increased consumption of vegetables, fruits, legumes, nuts, fish, cereal, and oils reduces concentrations of saturated and trans fats and increases dietary intake of monounsaturated fatty acids, fiber, antioxidants, and polyphenols.¹⁷

Combined endurance and resistance training is superior for improving glycemic control, cardiorespiratory fitness, and body composition, compared with either type of training alone.

Increasing the intake of fiber, an undigestible carbohydrate, offers numerous benefits in T2D management. High-fiber diets can help regulate blood sugar and lipid levels, increase satiety, reduce inflammation, aid in weight management, and reduce premature mortality.¹⁸ Insoluble fiber, found in foods such as whole wheat flour, nuts, and cauliflower, helps food pass more quickly through the stomach and intestines and adds bulk to stool. Soluble fiber, found in foods such as chickpeas, lentils, and Brussels sprouts, absorbs water and forms a gel-like substance that protects nutrients from digestive enzymes and slows down digestion. The result is a more gradual rise in postprandial glucose levels and improved insulin sensitivity.¹⁹ Dietary fiber may produce short-chain fatty acids which in turn activate incretin secretion and stimulate a glucose-dependent release of insulin from the pancreas.²⁰

Simple dietary substitutions, such as whole grains and legumes for white rice, can reduce fasting blood glucose and A1C levels.²¹ In a randomized controlled trial (RCT), increasing whole grain oat intake improved measures of glycemic control, reducing A1C by 1% at 1-year follow-up.¹⁹ Encourage patients with T2D to increase consumption of high-fiber foods and replace animal fats and refined grains with vegetable fats (eg, nuts, avocados, olives).
Nutritional therapies should be individualized, taking into account personal preferences and cultural customs.²² Nutritional habits may be based on race/ethnicity, ­religion/spirituality, or even the city in which an individual resides. Nutrition recommendations should account for these differences as well as access to healthy foods. For instance, ethnic groups whose dietary patterns include tortillas could be counseled to choose high-fiber options such as corn instead of flour tortillas and to incorporate vegetables in place of high-fat foods. Additionally, ethnic groups who favor using animal fats in foods such as greens could be advised on ways to add flavor to vegetables without adding saturated fats. Taking this approach may lessen barriers to change and increase ability to make dietary modifications.²³

Exercise

Encourage all patients with T2D to exercise regularly. The atherosclerotic plaques found in patients with T2D have increased inflammatory properties and result in worse cardiovascular outcomes compared with plaques in individuals without T2D.²⁴ Regular exercise reduces levels of pro-inflammatory markers—C-reactive protein, interleukin (IL)-6, and tumor necrosis factor alpha—and increases levels of anti-­inflammatory markers (IL-4 and IL-10).²⁴ Regular exercise can improve body composition, physical fitness, lipid and glucose metabolism, and insulin sensitivity.^25,26

A meta-analysis of RCTs demonstrated that structured exercise > 150 minutes per week resulted in A1C reductions of 0.89%,²⁷ which is comparable to the effect of many oral antihyperglycemic medications.²⁶ The Health Benefits of Aerobic and Resistance Training in individuals with T2D (HART-D) and Diabetes Aerobic and Resistance Exercise (DARE) studies demonstrated that combining endurance and resistance training was superior for improving glycemic control, cardiorespiratory fitness, and body composition, than using either type of training alone.²⁵ Both the American College of Sports Medicine (ACSM) and the ADA recommend that adults engage in at least 150 total minutes of moderate-intensity aerobic activity per week and resistance training 2 to 3 times weekly.²⁶ ACSM defines moderate-intensity exercise as 65% to 75% of maximal heart rate, a rating of perceived exertion of 3 to 4, or a step rate of 100 steps per minute.²⁸

Continue to: Because of their longitudinal relationships...

Because of their longitudinal relationships with patients, family physicians are in an optimal position to assess a patient’s physical capacity level and provide individualized counseling. Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.²⁹ Encourage your patients with T2D to exercise regularly, considering each individual’s ability to engage in physical activity.

Weight loss

Include weight management in the initial treatment of patients with newly diagnosed T2D. Weight loss decreases hepatic glucose production and increases peripheral insulin sensitivity and insulin secretion.³⁰ Moderate decreases in weight (5%-10%) can reduce complications related to diabetes, and sustained significant weight loss (> 10%) can potentially cause T2D remission (A1C < 6.5% after stopping diabetes medications).^31,32

Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.

Diabetes self-management education supports patients by giving them tools for making and maintaining lifestyle changes. Understanding individual barriers to change and addressing these during motivational interviews is important. Through a qualitative interview study, participants in a diabetes self-management program revealed 4 factors that motivated them to maintain lifestyle changes: support from others, experiencing the impact of the changes they made, fear of T2D complications, and forming new habits.³³ Family physicians are key in helping patients acquire knowledge and support to make the lifestyle modifications needed to manage newly diagnosed T2D.

Individualized pharmacotherapy considerations

For decades, the initial pharmacotherapeutic regimen for patients with newly diagnosed T2D considered the patient’s baseline A1C as a major driver for therapy. Metformin has been the mainstay in T2D treatment due to its clinical efficacy, minimal risk for hypoglycemia, and low cost. Regardless of the regimen, pharmacotherapy should be initiated at the time of T2D diagnosis in conjunction with the aforementioned lifestyle modifications.³⁴

When selecting pharmacotherapy, practice guidelines recommend considering the efficacy and adverse effects of medications, patient-specific comorbidities, adherence, cost, and a patient’s lifestyle factors.³⁴ Drug classes with pertinent information are listed in TABLE 2.^34-54 After starting medication, monitor the A1C level every 3 months to determine whether therapy should be intensified. Patients should have their labs drawn ahead of the quarterly visit, or point-of-care measurements may be used to facilitate in-person patient–provider discussions.

Continue to: Consider patient-specific factors when starting pharmacotherapy

Consider patient-specific factors when starting pharmacotherapy

ASCVD. Regardless of baseline glycemic control, offer patients who have ASCVD, or who are at high risk for it, an SGLT2 inhibitor (canagliflozin, dapagliflozin, or empagliflozin) or a long-acting GLP-1 receptor agonist (dulaglutide, liraglutide, or semaglutide).^34,35 SGLT2 inhibitors reduced the risk for MACE by 11% in patients with established ASCVD.⁵⁵ They also reduced a composite outcome of cardiovascular death or hospitalization for heart failure by 23% in patients with or without ASCVD or heart failure at baseline.⁵⁵ GLP-1 receptor agonists offer a similar reduction in MACE to SGLT2 inhibitors, but they do not have significant effects in heart failure.⁵⁶ Thiazolidinediones (TZDs), saxagliptin, and alogliptin should be avoided in patients with heart failure.⁵⁷ TZDs may reduce the risk for recurrent stroke in patients with T2D.⁵⁸

Chronic kidney disease (CKD). As with ASCVD, prioritize SGLT2 inhibitors and ­GLP-1 receptor agonists in patients with CKD. While both classes reduced the risk for progression of kidney disease such as macroalbuminuria, SGLT2 inhibitors offer additional benefits in their reduction of the worsening of estimated glomerular filtration rate, end-stage kidney disease, and renal death.⁵⁶

Obesity. Consider the effect of each drug class on weight when making initial treatment choices, taking special care to minimize weight gain and potentially promote weight loss.³⁴ The ADA prefers GLP-1 receptor agonists, but also suggests SGLT2 inhibitors in these patients. While all GLP-1 receptor agonists have an impact on weight, weekly subcutaneous semaglutide offers the most pronounced weight loss of 2 to 7 kg over 56 weeks.⁵⁹ SGLT2 inhibitors promote sustainable weight loss to a lesser degree, contributing to an average loss of 3 kg at 2 years.⁶⁰ Weight gain is common in patients taking sulfonylureas (2.01-2.3 kg)³¹ and insulin (3-9 kg weight gain in the first year)⁶¹ and should be avoided in patients with T2D and obesity.³⁴

Hypoglycemia risk. In addition to counseling patients on hypoglycemia management and prescribing glucagon rescue kits, offer medications with no or very low risk for hypoglycemia (eg, GLP-1 receptor agonists, SGLT2 inhibitors, dipeptidyl ­peptidase-4 inhibitors, and TZDs). Generally, avoid insulin and sulfonylureas in patients in whom hypoglycemia is a major concern (eg, older adults, individuals with labile blood glucose levels).³⁴ Patients with reduced renal function are at higher risk for hypoglycemia with insulin or sulfonylureas due to reduced drug clearance. However, insulin is often the only treatment for patients with advanced renal disease. Pay close attention to insulin dosing in patients with advanced renal disease, which may necessitate lower doses and smaller dose adjustments due to this risk.

Social determinants of health. Medication access and cost is a major burden in T2D management and should be considered for every patient. Compared with the period of 2005 to 2007, the annual cost of diabetes medications for an individual in 2015 to 2017 increased by 147%, rising from $1106 to $2727 per year.⁶² This increase is driven by the cost of insulin and newer medications without generic options.⁶² Identify local resources in your community, such as patient assistance programs and pharmacies with reduced-price generic prescription programs, which may be useful for patients who are underinsured or uninsured.

Continue to: Even if cost weren't an issue...

Even if cost weren’t an issue, many ­medications such as insulin and GLP-1 receptor agonists should be kept refrigerated and are only stable at room temperature for a limited time. Medications that are stable at room temperature should be prioritized in patients with limited or inconsistent access to refrigeration or unstable housing who may find it difficult to store their medications ­appropriately.

Do not delay insulin initiation in patients with high baseline A1C

Whenever possible, a GLP-1 receptor agonist is the preferred injectable medication to insulin. Starting insulin introduces numerous risks, including hypoglycemia, weight gain, and stigma. However, in the patient with newly diagnosed T2D, choose basal insulin when the baseline hyperglycemia is severe,³⁴ as indicated by:

• blood glucose > 300 mg/dL (16.7 mmol/L),
• A1C > 10% (86 mmol/mol),
• symptoms of hyperglycemia (polyuria or polydipsia), or
• evidence of catabolism (weight loss, hypertriglyceridemia, ketosis).

Basal insulin analogs are preferred over NPH given their reduced variability, dosing, and hypoglycemic risk.³⁵ Mixed insulins may be used if a patient is unable to afford an insulin analog, which can be quite costly. However, extensive counseling on dosing and management of hypoglycemia is crucial to patient safety with these agents. The ADA ­recommends initiating 0.1 to 0.2 units/kg of basal insulin daily or 10 units daily.³⁴ The AACE follows this recommendation for ­patients with baseline A1C < 8%, but it proposes a more aggressive initiation of 0.2 to 0.3 units/kg/d for patients with baseline A1C > 8%.³⁵ Titrate the dose by 2 units every 3 days to reach the target fasting blood glucose level. As hyperglycemia resolves, simplify the regimen and transition to noninsulin options per the previously discussed considerations.

It’s not just about glycemic control

In addition to the direct effects of hyperglycemia, a T2D diagnosis introduces an increased risk for ASCVD, a reduced ability to fight infection, and heightened risk for depression. Order a lipid panel at the time of T2D diagnosis and initiate lipid management as needed (TABLE 3^35,63,64). Both the ADA and the American Heart Association recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.⁶³ The AACE uses specific lipid targets and recommends moderate- to high-intensity statin therapy for patients with T2D.³⁵ All recommendations by professional organizations list high-intensity statins for patients with established ASCVD.

Both the ADA and the AHA recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.

It is also vital to recommend that patients with newly diagnosed T2D remain up to date on all indicated vaccinations. They should promptly receive the hepatitis B and pneumococcal vaccines if they have not already done so for a previous indication. COVID-19 and annual influenza vaccines also should be prioritized for these patients.⁶⁵

Finally, patients with diabetes are twice as likely to develop depression than patients without diabetes.⁶⁶ Individuals with T2D and depression exhibit poorer medication adherence, lifestyle choices, and glycemic control.⁶⁶ Screen for and treat these issues in all patients with T2D across the course of the disease.

Overall, work closely with patients to support them in managing their new diagnosis with evidence-based pharmacologic and nonpharmacologic approaches. The importance of lifestyle changes including high-fiber diets, regular exercise, and weight loss should not be overlooked. Do not delay starting pharmacotherapy after diagnosing T2D and consider medication-specific and patient-specific factors to individualize therapy, improve adherence, and prevent complications.

CORRESPONDENCE
Jennie B. Jarrett, PharmD, MMedEd, 833 South Wood Street (MC 886), Chicago, IL 60612; jarrett8@uic.edu

Nearly 40 antihyperglycemic agents have been approved by the US Food and Drug Administration (FDA) since the approval of human insulin in 1982.¹ In addition, existing antihyperglycemic medications are constantly gaining FDA approval for new indications for common type 2 diabetes (T2D) comorbidities. For example, in addition to their glycemic benefits, the sodium-glucose cotransporter-2 (SGLT2) inhibitors have been approved for use in patients with T2D and established atherosclerotic cardiovascular disease (ASCVD) to reduce the risk for major adverse cardiovascular events (MACE; canagliflozin), risk for hospitalization for heart failure (dapagliflozin), and cardiovascular death (empagliflozin).^2-4

The plethora of new agents and new data for existing agents, coupled with the annual release of guidelines from the American Diabetes Association (ADA) and practice recommendations from several other professional organizations,^5-7 make it challenging for family physicians to stay current and provide the most up-to-date, evidence-based care. In this article, we provide advice on how to approach the screening, diagnosis, and evaluation of T2D, and on how to manage newly diagnosed T2D.

Screening, Dx, and evaluation: A quick review

Screening

Screening recommendations vary among professional organizations (TABLE 1^5,6,8). The US Preventive Services Task Force (USPSTF) recommends screening adults ages 35 to 70 years who are overweight or obese. Clinicians also can consider screening patients with a higher risk for diabetes.⁵ The ADA suggests screening all adults starting at 35 years, regardless of risk factors.⁸ Asymptomatic adults of any age with overweight or obesity and 1 or more risk factors should be screened.⁸The American Association of Clinical Endocrinology (AACE) recommends screening adults of any age who have risk factors.⁶ If the screening result is normal, repeat testing in 3 years is appropriate, unless there is a change in symptoms or risks.^5,8 Annual testing can be considered in patients with ≥ 2 risk factors or with prediabetes (glycosylated hemoglobin [A1C] ≥ 5.7%).^6,8

Making the diagnosis

The initial diagnosis of diabetes can be made by a fasting plasma glucose level ≥ 126 mg/dL (7.0 mmol/L); a 2-hour plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) following an oral glucose tolerance test; or an A1C level ≥ 6.5%. Prioritize lab-drawn A1C measurements over point-of-care tests to diagnose T2D. In ­patients with classic symptoms of hyper­glycemia, a random plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) is also diagnostic. Generally, these tests are considered equally appropriate in screening for diabetes and may be used to detect prediabetes. In the absence of clear symptoms of hyperglycemia, the diagnosis of diabetes requires 2 abnormal screening test results, either via 1 blood sample (such as an abnormal A1C and glucose) or 2 separate blood samples of the same test. Further evaluation is advised if there is discordance between the 2 samples.⁸

Extended evaluations

Patients with newly diagnosed T2D require a thorough evaluation for comorbidities and complications of diabetes. Refer patients to an ophthalmologist for a dilated eye examination, with subsequent exams occurring every 1 to 2 years.^6,9 Additional referrals for diabetes education, family planning for women of reproductive age, and dental, social, or mental health services may be clinically appropriate.⁹

Setting goals for glycemic control

Glycemic control is commonly monitored by the A1C level and by blood glucose monitoring either through traditional point-of-care glucometers or continuous glucose monitors (CGMs).¹⁰ Generally, CGMs provide more glycemic data than traditional glucometers and may cue patients to choose healthier dietary options and engage in physical exercise.¹¹ Patients with T2D who use CGMs exhibit lower A1Cs, greater time in glycemic range, and reduced hypoglycemic episodes.¹¹ Generally, CGMs are reserved for patients with type 1 diabetes and patients with T2D who use multiple daily injections, subcutaneous insulin infusions, or basal insulin only.¹² Most professional organizations recommend that clinicians consider patient-specific factors to set individualized glycemic goals.^6,10,13,14 For example, more stringent glycemic goals could be pursued for patients with longer life expectancy, shorter disease duration, absence of complications (eg, nephropathy, neuropathy, or cardiovascular disease), fewer comorbid conditions, lower hypoglycemia risk, or higher cognitive function.⁶

With newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.

More specific A1C goals vary by professional organization. For nonpregnant adults, the ADA recommends an A1C goal of < 7% and a preprandial blood glucose level of 80 to 130 mg/dL (4.4-7.2 mmol/L).¹⁰ However, a lower A1C goal may be appropriate if it can be attained safely without causing hypoglycemia or other adverse effects.¹⁰ The AACE suggests an A1C goal of ≤ 6.5% and a fasting blood glucose level of < 110 mg/dL when it can be achieved safely.⁶ More stringent A1C goals may reduce long-term micro- and macrovascular complications—especially in patients with newly diagnosed T2D.¹⁰ While older studies such as the ACCORD trial found increased mortality in groups with more stringent glycemic targets, they did not include newer agents (SGLT2 inhibitors or glucagon-like peptide-1 [GLP-1] receptor agonists) that reduce cardiovascular events by mechanisms outside their glycemic-lowering effect. With these newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.¹⁰ Both the ADA and AACE recommend a less stringent A1C goal of 7% to 8% for patients with limited life expectancy or risks (eg, a history of hypoglycemia) that outweigh expected benefits.^6,10

Continue to: Lifestyle modifications

Nutrition

The energy-dense Western diet, combined with sedentary behavior, are thought to be a primary cause of T2D.¹⁵ Therefore, include lifestyle modifications in the initial management of newly diagnosed T2D. Diets that replace carbohydrates with saturated and trans fats are related to increased mortality in patients with T2D.¹⁶ Increased consumption of vegetables, fruits, legumes, nuts, fish, cereal, and oils reduces concentrations of saturated and trans fats and increases dietary intake of monounsaturated fatty acids, fiber, antioxidants, and polyphenols.¹⁷

Combined endurance and resistance training is superior for improving glycemic control, cardiorespiratory fitness, and body composition, compared with either type of training alone.

Increasing the intake of fiber, an undigestible carbohydrate, offers numerous benefits in T2D management. High-fiber diets can help regulate blood sugar and lipid levels, increase satiety, reduce inflammation, aid in weight management, and reduce premature mortality.¹⁸ Insoluble fiber, found in foods such as whole wheat flour, nuts, and cauliflower, helps food pass more quickly through the stomach and intestines and adds bulk to stool. Soluble fiber, found in foods such as chickpeas, lentils, and Brussels sprouts, absorbs water and forms a gel-like substance that protects nutrients from digestive enzymes and slows down digestion. The result is a more gradual rise in postprandial glucose levels and improved insulin sensitivity.¹⁹ Dietary fiber may produce short-chain fatty acids which in turn activate incretin secretion and stimulate a glucose-dependent release of insulin from the pancreas.²⁰

Simple dietary substitutions, such as whole grains and legumes for white rice, can reduce fasting blood glucose and A1C levels.²¹ In a randomized controlled trial (RCT), increasing whole grain oat intake improved measures of glycemic control, reducing A1C by 1% at 1-year follow-up.¹⁹ Encourage patients with T2D to increase consumption of high-fiber foods and replace animal fats and refined grains with vegetable fats (eg, nuts, avocados, olives).
Nutritional therapies should be individualized, taking into account personal preferences and cultural customs.²² Nutritional habits may be based on race/ethnicity, ­religion/spirituality, or even the city in which an individual resides. Nutrition recommendations should account for these differences as well as access to healthy foods. For instance, ethnic groups whose dietary patterns include tortillas could be counseled to choose high-fiber options such as corn instead of flour tortillas and to incorporate vegetables in place of high-fat foods. Additionally, ethnic groups who favor using animal fats in foods such as greens could be advised on ways to add flavor to vegetables without adding saturated fats. Taking this approach may lessen barriers to change and increase ability to make dietary modifications.²³

Exercise

Encourage all patients with T2D to exercise regularly. The atherosclerotic plaques found in patients with T2D have increased inflammatory properties and result in worse cardiovascular outcomes compared with plaques in individuals without T2D.²⁴ Regular exercise reduces levels of pro-inflammatory markers—C-reactive protein, interleukin (IL)-6, and tumor necrosis factor alpha—and increases levels of anti-­inflammatory markers (IL-4 and IL-10).²⁴ Regular exercise can improve body composition, physical fitness, lipid and glucose metabolism, and insulin sensitivity.^25,26

A meta-analysis of RCTs demonstrated that structured exercise > 150 minutes per week resulted in A1C reductions of 0.89%,²⁷ which is comparable to the effect of many oral antihyperglycemic medications.²⁶ The Health Benefits of Aerobic and Resistance Training in individuals with T2D (HART-D) and Diabetes Aerobic and Resistance Exercise (DARE) studies demonstrated that combining endurance and resistance training was superior for improving glycemic control, cardiorespiratory fitness, and body composition, than using either type of training alone.²⁵ Both the American College of Sports Medicine (ACSM) and the ADA recommend that adults engage in at least 150 total minutes of moderate-intensity aerobic activity per week and resistance training 2 to 3 times weekly.²⁶ ACSM defines moderate-intensity exercise as 65% to 75% of maximal heart rate, a rating of perceived exertion of 3 to 4, or a step rate of 100 steps per minute.²⁸

Continue to: Because of their longitudinal relationships...

Because of their longitudinal relationships with patients, family physicians are in an optimal position to assess a patient’s physical capacity level and provide individualized counseling. Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.²⁹ Encourage your patients with T2D to exercise regularly, considering each individual’s ability to engage in physical activity.

Weight loss

Include weight management in the initial treatment of patients with newly diagnosed T2D. Weight loss decreases hepatic glucose production and increases peripheral insulin sensitivity and insulin secretion.³⁰ Moderate decreases in weight (5%-10%) can reduce complications related to diabetes, and sustained significant weight loss (> 10%) can potentially cause T2D remission (A1C < 6.5% after stopping diabetes medications).^31,32

Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.

Diabetes self-management education supports patients by giving them tools for making and maintaining lifestyle changes. Understanding individual barriers to change and addressing these during motivational interviews is important. Through a qualitative interview study, participants in a diabetes self-management program revealed 4 factors that motivated them to maintain lifestyle changes: support from others, experiencing the impact of the changes they made, fear of T2D complications, and forming new habits.³³ Family physicians are key in helping patients acquire knowledge and support to make the lifestyle modifications needed to manage newly diagnosed T2D.

Individualized pharmacotherapy considerations

For decades, the initial pharmacotherapeutic regimen for patients with newly diagnosed T2D considered the patient’s baseline A1C as a major driver for therapy. Metformin has been the mainstay in T2D treatment due to its clinical efficacy, minimal risk for hypoglycemia, and low cost. Regardless of the regimen, pharmacotherapy should be initiated at the time of T2D diagnosis in conjunction with the aforementioned lifestyle modifications.³⁴

When selecting pharmacotherapy, practice guidelines recommend considering the efficacy and adverse effects of medications, patient-specific comorbidities, adherence, cost, and a patient’s lifestyle factors.³⁴ Drug classes with pertinent information are listed in TABLE 2.^34-54 After starting medication, monitor the A1C level every 3 months to determine whether therapy should be intensified. Patients should have their labs drawn ahead of the quarterly visit, or point-of-care measurements may be used to facilitate in-person patient–provider discussions.

Continue to: Consider patient-specific factors when starting pharmacotherapy

Consider patient-specific factors when starting pharmacotherapy

ASCVD. Regardless of baseline glycemic control, offer patients who have ASCVD, or who are at high risk for it, an SGLT2 inhibitor (canagliflozin, dapagliflozin, or empagliflozin) or a long-acting GLP-1 receptor agonist (dulaglutide, liraglutide, or semaglutide).^34,35 SGLT2 inhibitors reduced the risk for MACE by 11% in patients with established ASCVD.⁵⁵ They also reduced a composite outcome of cardiovascular death or hospitalization for heart failure by 23% in patients with or without ASCVD or heart failure at baseline.⁵⁵ GLP-1 receptor agonists offer a similar reduction in MACE to SGLT2 inhibitors, but they do not have significant effects in heart failure.⁵⁶ Thiazolidinediones (TZDs), saxagliptin, and alogliptin should be avoided in patients with heart failure.⁵⁷ TZDs may reduce the risk for recurrent stroke in patients with T2D.⁵⁸

Chronic kidney disease (CKD). As with ASCVD, prioritize SGLT2 inhibitors and ­GLP-1 receptor agonists in patients with CKD. While both classes reduced the risk for progression of kidney disease such as macroalbuminuria, SGLT2 inhibitors offer additional benefits in their reduction of the worsening of estimated glomerular filtration rate, end-stage kidney disease, and renal death.⁵⁶

Obesity. Consider the effect of each drug class on weight when making initial treatment choices, taking special care to minimize weight gain and potentially promote weight loss.³⁴ The ADA prefers GLP-1 receptor agonists, but also suggests SGLT2 inhibitors in these patients. While all GLP-1 receptor agonists have an impact on weight, weekly subcutaneous semaglutide offers the most pronounced weight loss of 2 to 7 kg over 56 weeks.⁵⁹ SGLT2 inhibitors promote sustainable weight loss to a lesser degree, contributing to an average loss of 3 kg at 2 years.⁶⁰ Weight gain is common in patients taking sulfonylureas (2.01-2.3 kg)³¹ and insulin (3-9 kg weight gain in the first year)⁶¹ and should be avoided in patients with T2D and obesity.³⁴

Hypoglycemia risk. In addition to counseling patients on hypoglycemia management and prescribing glucagon rescue kits, offer medications with no or very low risk for hypoglycemia (eg, GLP-1 receptor agonists, SGLT2 inhibitors, dipeptidyl ­peptidase-4 inhibitors, and TZDs). Generally, avoid insulin and sulfonylureas in patients in whom hypoglycemia is a major concern (eg, older adults, individuals with labile blood glucose levels).³⁴ Patients with reduced renal function are at higher risk for hypoglycemia with insulin or sulfonylureas due to reduced drug clearance. However, insulin is often the only treatment for patients with advanced renal disease. Pay close attention to insulin dosing in patients with advanced renal disease, which may necessitate lower doses and smaller dose adjustments due to this risk.

Social determinants of health. Medication access and cost is a major burden in T2D management and should be considered for every patient. Compared with the period of 2005 to 2007, the annual cost of diabetes medications for an individual in 2015 to 2017 increased by 147%, rising from $1106 to $2727 per year.⁶² This increase is driven by the cost of insulin and newer medications without generic options.⁶² Identify local resources in your community, such as patient assistance programs and pharmacies with reduced-price generic prescription programs, which may be useful for patients who are underinsured or uninsured.

Continue to: Even if cost weren't an issue...

Even if cost weren’t an issue, many ­medications such as insulin and GLP-1 receptor agonists should be kept refrigerated and are only stable at room temperature for a limited time. Medications that are stable at room temperature should be prioritized in patients with limited or inconsistent access to refrigeration or unstable housing who may find it difficult to store their medications ­appropriately.

Do not delay insulin initiation in patients with high baseline A1C

Whenever possible, a GLP-1 receptor agonist is the preferred injectable medication to insulin. Starting insulin introduces numerous risks, including hypoglycemia, weight gain, and stigma. However, in the patient with newly diagnosed T2D, choose basal insulin when the baseline hyperglycemia is severe,³⁴ as indicated by:

• blood glucose > 300 mg/dL (16.7 mmol/L),
• A1C > 10% (86 mmol/mol),
• symptoms of hyperglycemia (polyuria or polydipsia), or
• evidence of catabolism (weight loss, hypertriglyceridemia, ketosis).

Basal insulin analogs are preferred over NPH given their reduced variability, dosing, and hypoglycemic risk.³⁵ Mixed insulins may be used if a patient is unable to afford an insulin analog, which can be quite costly. However, extensive counseling on dosing and management of hypoglycemia is crucial to patient safety with these agents. The ADA ­recommends initiating 0.1 to 0.2 units/kg of basal insulin daily or 10 units daily.³⁴ The AACE follows this recommendation for ­patients with baseline A1C < 8%, but it proposes a more aggressive initiation of 0.2 to 0.3 units/kg/d for patients with baseline A1C > 8%.³⁵ Titrate the dose by 2 units every 3 days to reach the target fasting blood glucose level. As hyperglycemia resolves, simplify the regimen and transition to noninsulin options per the previously discussed considerations.

It’s not just about glycemic control

In addition to the direct effects of hyperglycemia, a T2D diagnosis introduces an increased risk for ASCVD, a reduced ability to fight infection, and heightened risk for depression. Order a lipid panel at the time of T2D diagnosis and initiate lipid management as needed (TABLE 3^35,63,64). Both the ADA and the American Heart Association recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.⁶³ The AACE uses specific lipid targets and recommends moderate- to high-intensity statin therapy for patients with T2D.³⁵ All recommendations by professional organizations list high-intensity statins for patients with established ASCVD.

Both the ADA and the AHA recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.

It is also vital to recommend that patients with newly diagnosed T2D remain up to date on all indicated vaccinations. They should promptly receive the hepatitis B and pneumococcal vaccines if they have not already done so for a previous indication. COVID-19 and annual influenza vaccines also should be prioritized for these patients.⁶⁵

Finally, patients with diabetes are twice as likely to develop depression than patients without diabetes.⁶⁶ Individuals with T2D and depression exhibit poorer medication adherence, lifestyle choices, and glycemic control.⁶⁶ Screen for and treat these issues in all patients with T2D across the course of the disease.

Overall, work closely with patients to support them in managing their new diagnosis with evidence-based pharmacologic and nonpharmacologic approaches. The importance of lifestyle changes including high-fiber diets, regular exercise, and weight loss should not be overlooked. Do not delay starting pharmacotherapy after diagnosing T2D and consider medication-specific and patient-specific factors to individualize therapy, improve adherence, and prevent complications.

CORRESPONDENCE
Jennie B. Jarrett, PharmD, MMedEd, 833 South Wood Street (MC 886), Chicago, IL 60612; jarrett8@uic.edu

Nearly 40 antihyperglycemic agents have been approved by the US Food and Drug Administration (FDA) since the approval of human insulin in 1982.¹ In addition, existing antihyperglycemic medications are constantly gaining FDA approval for new indications for common type 2 diabetes (T2D) comorbidities. For example, in addition to their glycemic benefits, the sodium-glucose cotransporter-2 (SGLT2) inhibitors have been approved for use in patients with T2D and established atherosclerotic cardiovascular disease (ASCVD) to reduce the risk for major adverse cardiovascular events (MACE; canagliflozin), risk for hospitalization for heart failure (dapagliflozin), and cardiovascular death (empagliflozin).^2-4

The plethora of new agents and new data for existing agents, coupled with the annual release of guidelines from the American Diabetes Association (ADA) and practice recommendations from several other professional organizations,^5-7 make it challenging for family physicians to stay current and provide the most up-to-date, evidence-based care. In this article, we provide advice on how to approach the screening, diagnosis, and evaluation of T2D, and on how to manage newly diagnosed T2D.

Screening, Dx, and evaluation: A quick review

Screening

Screening recommendations vary among professional organizations (TABLE 1^5,6,8). The US Preventive Services Task Force (USPSTF) recommends screening adults ages 35 to 70 years who are overweight or obese. Clinicians also can consider screening patients with a higher risk for diabetes.⁵ The ADA suggests screening all adults starting at 35 years, regardless of risk factors.⁸ Asymptomatic adults of any age with overweight or obesity and 1 or more risk factors should be screened.⁸The American Association of Clinical Endocrinology (AACE) recommends screening adults of any age who have risk factors.⁶ If the screening result is normal, repeat testing in 3 years is appropriate, unless there is a change in symptoms or risks.^5,8 Annual testing can be considered in patients with ≥ 2 risk factors or with prediabetes (glycosylated hemoglobin [A1C] ≥ 5.7%).^6,8

Making the diagnosis

The initial diagnosis of diabetes can be made by a fasting plasma glucose level ≥ 126 mg/dL (7.0 mmol/L); a 2-hour plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) following an oral glucose tolerance test; or an A1C level ≥ 6.5%. Prioritize lab-drawn A1C measurements over point-of-care tests to diagnose T2D. In ­patients with classic symptoms of hyper­glycemia, a random plasma glucose level ≥ 200 mg/dL (11.0 mmol/L) is also diagnostic. Generally, these tests are considered equally appropriate in screening for diabetes and may be used to detect prediabetes. In the absence of clear symptoms of hyperglycemia, the diagnosis of diabetes requires 2 abnormal screening test results, either via 1 blood sample (such as an abnormal A1C and glucose) or 2 separate blood samples of the same test. Further evaluation is advised if there is discordance between the 2 samples.⁸

Extended evaluations

Patients with newly diagnosed T2D require a thorough evaluation for comorbidities and complications of diabetes. Refer patients to an ophthalmologist for a dilated eye examination, with subsequent exams occurring every 1 to 2 years.^6,9 Additional referrals for diabetes education, family planning for women of reproductive age, and dental, social, or mental health services may be clinically appropriate.⁹

Setting goals for glycemic control

Glycemic control is commonly monitored by the A1C level and by blood glucose monitoring either through traditional point-of-care glucometers or continuous glucose monitors (CGMs).¹⁰ Generally, CGMs provide more glycemic data than traditional glucometers and may cue patients to choose healthier dietary options and engage in physical exercise.¹¹ Patients with T2D who use CGMs exhibit lower A1Cs, greater time in glycemic range, and reduced hypoglycemic episodes.¹¹ Generally, CGMs are reserved for patients with type 1 diabetes and patients with T2D who use multiple daily injections, subcutaneous insulin infusions, or basal insulin only.¹² Most professional organizations recommend that clinicians consider patient-specific factors to set individualized glycemic goals.^6,10,13,14 For example, more stringent glycemic goals could be pursued for patients with longer life expectancy, shorter disease duration, absence of complications (eg, nephropathy, neuropathy, or cardiovascular disease), fewer comorbid conditions, lower hypoglycemia risk, or higher cognitive function.⁶

With newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.

More specific A1C goals vary by professional organization. For nonpregnant adults, the ADA recommends an A1C goal of < 7% and a preprandial blood glucose level of 80 to 130 mg/dL (4.4-7.2 mmol/L).¹⁰ However, a lower A1C goal may be appropriate if it can be attained safely without causing hypoglycemia or other adverse effects.¹⁰ The AACE suggests an A1C goal of ≤ 6.5% and a fasting blood glucose level of < 110 mg/dL when it can be achieved safely.⁶ More stringent A1C goals may reduce long-term micro- and macrovascular complications—especially in patients with newly diagnosed T2D.¹⁰ While older studies such as the ACCORD trial found increased mortality in groups with more stringent glycemic targets, they did not include newer agents (SGLT2 inhibitors or glucagon-like peptide-1 [GLP-1] receptor agonists) that reduce cardiovascular events by mechanisms outside their glycemic-lowering effect. With these newer agents, more aggressive A1C goals can be targeted safely in select patients, particularly those with long life expectancy.¹⁰ Both the ADA and AACE recommend a less stringent A1C goal of 7% to 8% for patients with limited life expectancy or risks (eg, a history of hypoglycemia) that outweigh expected benefits.^6,10

Continue to: Lifestyle modifications

Nutrition

The energy-dense Western diet, combined with sedentary behavior, are thought to be a primary cause of T2D.¹⁵ Therefore, include lifestyle modifications in the initial management of newly diagnosed T2D. Diets that replace carbohydrates with saturated and trans fats are related to increased mortality in patients with T2D.¹⁶ Increased consumption of vegetables, fruits, legumes, nuts, fish, cereal, and oils reduces concentrations of saturated and trans fats and increases dietary intake of monounsaturated fatty acids, fiber, antioxidants, and polyphenols.¹⁷

Combined endurance and resistance training is superior for improving glycemic control, cardiorespiratory fitness, and body composition, compared with either type of training alone.

Increasing the intake of fiber, an undigestible carbohydrate, offers numerous benefits in T2D management. High-fiber diets can help regulate blood sugar and lipid levels, increase satiety, reduce inflammation, aid in weight management, and reduce premature mortality.¹⁸ Insoluble fiber, found in foods such as whole wheat flour, nuts, and cauliflower, helps food pass more quickly through the stomach and intestines and adds bulk to stool. Soluble fiber, found in foods such as chickpeas, lentils, and Brussels sprouts, absorbs water and forms a gel-like substance that protects nutrients from digestive enzymes and slows down digestion. The result is a more gradual rise in postprandial glucose levels and improved insulin sensitivity.¹⁹ Dietary fiber may produce short-chain fatty acids which in turn activate incretin secretion and stimulate a glucose-dependent release of insulin from the pancreas.²⁰

Simple dietary substitutions, such as whole grains and legumes for white rice, can reduce fasting blood glucose and A1C levels.²¹ In a randomized controlled trial (RCT), increasing whole grain oat intake improved measures of glycemic control, reducing A1C by 1% at 1-year follow-up.¹⁹ Encourage patients with T2D to increase consumption of high-fiber foods and replace animal fats and refined grains with vegetable fats (eg, nuts, avocados, olives).
Nutritional therapies should be individualized, taking into account personal preferences and cultural customs.²² Nutritional habits may be based on race/ethnicity, ­religion/spirituality, or even the city in which an individual resides. Nutrition recommendations should account for these differences as well as access to healthy foods. For instance, ethnic groups whose dietary patterns include tortillas could be counseled to choose high-fiber options such as corn instead of flour tortillas and to incorporate vegetables in place of high-fat foods. Additionally, ethnic groups who favor using animal fats in foods such as greens could be advised on ways to add flavor to vegetables without adding saturated fats. Taking this approach may lessen barriers to change and increase ability to make dietary modifications.²³

Exercise

Encourage all patients with T2D to exercise regularly. The atherosclerotic plaques found in patients with T2D have increased inflammatory properties and result in worse cardiovascular outcomes compared with plaques in individuals without T2D.²⁴ Regular exercise reduces levels of pro-inflammatory markers—C-reactive protein, interleukin (IL)-6, and tumor necrosis factor alpha—and increases levels of anti-­inflammatory markers (IL-4 and IL-10).²⁴ Regular exercise can improve body composition, physical fitness, lipid and glucose metabolism, and insulin sensitivity.^25,26

A meta-analysis of RCTs demonstrated that structured exercise > 150 minutes per week resulted in A1C reductions of 0.89%,²⁷ which is comparable to the effect of many oral antihyperglycemic medications.²⁶ The Health Benefits of Aerobic and Resistance Training in individuals with T2D (HART-D) and Diabetes Aerobic and Resistance Exercise (DARE) studies demonstrated that combining endurance and resistance training was superior for improving glycemic control, cardiorespiratory fitness, and body composition, than using either type of training alone.²⁵ Both the American College of Sports Medicine (ACSM) and the ADA recommend that adults engage in at least 150 total minutes of moderate-intensity aerobic activity per week and resistance training 2 to 3 times weekly.²⁶ ACSM defines moderate-intensity exercise as 65% to 75% of maximal heart rate, a rating of perceived exertion of 3 to 4, or a step rate of 100 steps per minute.²⁸

Continue to: Because of their longitudinal relationships...

Because of their longitudinal relationships with patients, family physicians are in an optimal position to assess a patient’s physical capacity level and provide individualized counseling. Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.²⁹ Encourage your patients with T2D to exercise regularly, considering each individual’s ability to engage in physical activity.

Weight loss

Include weight management in the initial treatment of patients with newly diagnosed T2D. Weight loss decreases hepatic glucose production and increases peripheral insulin sensitivity and insulin secretion.³⁰ Moderate decreases in weight (5%-10%) can reduce complications related to diabetes, and sustained significant weight loss (> 10%) can potentially cause T2D remission (A1C < 6.5% after stopping diabetes medications).^31,32

Several systematic reviews have demonstrated that counseling on exercise increases patients’ participation in physical activity.

Diabetes self-management education supports patients by giving them tools for making and maintaining lifestyle changes. Understanding individual barriers to change and addressing these during motivational interviews is important. Through a qualitative interview study, participants in a diabetes self-management program revealed 4 factors that motivated them to maintain lifestyle changes: support from others, experiencing the impact of the changes they made, fear of T2D complications, and forming new habits.³³ Family physicians are key in helping patients acquire knowledge and support to make the lifestyle modifications needed to manage newly diagnosed T2D.

Individualized pharmacotherapy considerations

For decades, the initial pharmacotherapeutic regimen for patients with newly diagnosed T2D considered the patient’s baseline A1C as a major driver for therapy. Metformin has been the mainstay in T2D treatment due to its clinical efficacy, minimal risk for hypoglycemia, and low cost. Regardless of the regimen, pharmacotherapy should be initiated at the time of T2D diagnosis in conjunction with the aforementioned lifestyle modifications.³⁴

When selecting pharmacotherapy, practice guidelines recommend considering the efficacy and adverse effects of medications, patient-specific comorbidities, adherence, cost, and a patient’s lifestyle factors.³⁴ Drug classes with pertinent information are listed in TABLE 2.^34-54 After starting medication, monitor the A1C level every 3 months to determine whether therapy should be intensified. Patients should have their labs drawn ahead of the quarterly visit, or point-of-care measurements may be used to facilitate in-person patient–provider discussions.

Continue to: Consider patient-specific factors when starting pharmacotherapy

Consider patient-specific factors when starting pharmacotherapy

ASCVD. Regardless of baseline glycemic control, offer patients who have ASCVD, or who are at high risk for it, an SGLT2 inhibitor (canagliflozin, dapagliflozin, or empagliflozin) or a long-acting GLP-1 receptor agonist (dulaglutide, liraglutide, or semaglutide).^34,35 SGLT2 inhibitors reduced the risk for MACE by 11% in patients with established ASCVD.⁵⁵ They also reduced a composite outcome of cardiovascular death or hospitalization for heart failure by 23% in patients with or without ASCVD or heart failure at baseline.⁵⁵ GLP-1 receptor agonists offer a similar reduction in MACE to SGLT2 inhibitors, but they do not have significant effects in heart failure.⁵⁶ Thiazolidinediones (TZDs), saxagliptin, and alogliptin should be avoided in patients with heart failure.⁵⁷ TZDs may reduce the risk for recurrent stroke in patients with T2D.⁵⁸

Chronic kidney disease (CKD). As with ASCVD, prioritize SGLT2 inhibitors and ­GLP-1 receptor agonists in patients with CKD. While both classes reduced the risk for progression of kidney disease such as macroalbuminuria, SGLT2 inhibitors offer additional benefits in their reduction of the worsening of estimated glomerular filtration rate, end-stage kidney disease, and renal death.⁵⁶

Obesity. Consider the effect of each drug class on weight when making initial treatment choices, taking special care to minimize weight gain and potentially promote weight loss.³⁴ The ADA prefers GLP-1 receptor agonists, but also suggests SGLT2 inhibitors in these patients. While all GLP-1 receptor agonists have an impact on weight, weekly subcutaneous semaglutide offers the most pronounced weight loss of 2 to 7 kg over 56 weeks.⁵⁹ SGLT2 inhibitors promote sustainable weight loss to a lesser degree, contributing to an average loss of 3 kg at 2 years.⁶⁰ Weight gain is common in patients taking sulfonylureas (2.01-2.3 kg)³¹ and insulin (3-9 kg weight gain in the first year)⁶¹ and should be avoided in patients with T2D and obesity.³⁴

Hypoglycemia risk. In addition to counseling patients on hypoglycemia management and prescribing glucagon rescue kits, offer medications with no or very low risk for hypoglycemia (eg, GLP-1 receptor agonists, SGLT2 inhibitors, dipeptidyl ­peptidase-4 inhibitors, and TZDs). Generally, avoid insulin and sulfonylureas in patients in whom hypoglycemia is a major concern (eg, older adults, individuals with labile blood glucose levels).³⁴ Patients with reduced renal function are at higher risk for hypoglycemia with insulin or sulfonylureas due to reduced drug clearance. However, insulin is often the only treatment for patients with advanced renal disease. Pay close attention to insulin dosing in patients with advanced renal disease, which may necessitate lower doses and smaller dose adjustments due to this risk.

Social determinants of health. Medication access and cost is a major burden in T2D management and should be considered for every patient. Compared with the period of 2005 to 2007, the annual cost of diabetes medications for an individual in 2015 to 2017 increased by 147%, rising from $1106 to $2727 per year.⁶² This increase is driven by the cost of insulin and newer medications without generic options.⁶² Identify local resources in your community, such as patient assistance programs and pharmacies with reduced-price generic prescription programs, which may be useful for patients who are underinsured or uninsured.

Continue to: Even if cost weren't an issue...

Even if cost weren’t an issue, many ­medications such as insulin and GLP-1 receptor agonists should be kept refrigerated and are only stable at room temperature for a limited time. Medications that are stable at room temperature should be prioritized in patients with limited or inconsistent access to refrigeration or unstable housing who may find it difficult to store their medications ­appropriately.

Do not delay insulin initiation in patients with high baseline A1C

Whenever possible, a GLP-1 receptor agonist is the preferred injectable medication to insulin. Starting insulin introduces numerous risks, including hypoglycemia, weight gain, and stigma. However, in the patient with newly diagnosed T2D, choose basal insulin when the baseline hyperglycemia is severe,³⁴ as indicated by:

• blood glucose > 300 mg/dL (16.7 mmol/L),
• A1C > 10% (86 mmol/mol),
• symptoms of hyperglycemia (polyuria or polydipsia), or
• evidence of catabolism (weight loss, hypertriglyceridemia, ketosis).

Basal insulin analogs are preferred over NPH given their reduced variability, dosing, and hypoglycemic risk.³⁵ Mixed insulins may be used if a patient is unable to afford an insulin analog, which can be quite costly. However, extensive counseling on dosing and management of hypoglycemia is crucial to patient safety with these agents. The ADA ­recommends initiating 0.1 to 0.2 units/kg of basal insulin daily or 10 units daily.³⁴ The AACE follows this recommendation for ­patients with baseline A1C < 8%, but it proposes a more aggressive initiation of 0.2 to 0.3 units/kg/d for patients with baseline A1C > 8%.³⁵ Titrate the dose by 2 units every 3 days to reach the target fasting blood glucose level. As hyperglycemia resolves, simplify the regimen and transition to noninsulin options per the previously discussed considerations.

It’s not just about glycemic control

In addition to the direct effects of hyperglycemia, a T2D diagnosis introduces an increased risk for ASCVD, a reduced ability to fight infection, and heightened risk for depression. Order a lipid panel at the time of T2D diagnosis and initiate lipid management as needed (TABLE 3^35,63,64). Both the ADA and the American Heart Association recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.⁶³ The AACE uses specific lipid targets and recommends moderate- to high-intensity statin therapy for patients with T2D.³⁵ All recommendations by professional organizations list high-intensity statins for patients with established ASCVD.

Both the ADA and the AHA recommend starting a moderate-intensity statin as primary prevention for all patients with T2D between 40 and 75 years of age regardless of the 10-year ASCVD risk.

It is also vital to recommend that patients with newly diagnosed T2D remain up to date on all indicated vaccinations. They should promptly receive the hepatitis B and pneumococcal vaccines if they have not already done so for a previous indication. COVID-19 and annual influenza vaccines also should be prioritized for these patients.⁶⁵

Finally, patients with diabetes are twice as likely to develop depression than patients without diabetes.⁶⁶ Individuals with T2D and depression exhibit poorer medication adherence, lifestyle choices, and glycemic control.⁶⁶ Screen for and treat these issues in all patients with T2D across the course of the disease.

Overall, work closely with patients to support them in managing their new diagnosis with evidence-based pharmacologic and nonpharmacologic approaches. The importance of lifestyle changes including high-fiber diets, regular exercise, and weight loss should not be overlooked. Do not delay starting pharmacotherapy after diagnosing T2D and consider medication-specific and patient-specific factors to individualize therapy, improve adherence, and prevent complications.

CORRESPONDENCE
Jennie B. Jarrett, PharmD, MMedEd, 833 South Wood Street (MC 886), Chicago, IL 60612; jarrett8@uic.edu

Platelet-rich plasma (PRP) injections have become a popular treatment option in a variety of specialties including sports medicine, maxillofacial surgery, dermatology, cosmetology, and reproductive medicine.¹ PRP is an autologous blood product derived from whole blood, using a centrifuge to isolate a concentrated layer of platelets. The ­a-granules in platelets release transforming growth factor b 1, vascular endothelial growth factor, platelet-derived growth factor, basic fibroblast growth factor, epidermal growth factor, insulin-like growth factor 1, and other mediatorsthat enhance the natural healing process.²

When patients ask. Familiarity with the use of PRP to treat specific musculoskeletal (MSK) conditions is essential for family physicians who frequently are asked by patients about whether PRP is right for them. These patients may have experienced failure of medication therapy or declined surgical intervention, or may not be surgical candidates. This review details the evidence surrounding common intra-articular and extra-articular applications of PRP. But first, a word about how PRP is prepared, its contraindications, and costs.

Preparation and types of PRP

Although there are many commercial systems for preparing PRP, there is no consensus on the optimal formulation.² Other terms for PRP, such as autologous concentrated platelets and super-concentrated platelets, are based on concentration of red blood cells, leukocytes, and fibrin.³ PRP therapies usually are categorized as leukocyte-rich PRP (LR-PRP) or leukocyte-poor PRP (LP-PRP), based on neutrophil concentrations that are above and below baseline.² Leukocyte concentration is one of the most debated topics in PRP therapy.⁴

Common commercially available preparation systems produce platelet concentrations between 3 to 6 times the baseline platelet count.⁵ Although there is no universally agreed upon PRP formulation, studies have shown 2 centrifugation cycles (“double-spun” or “dual centrifugation”) that yield platelet concentrations between 1.8 to 1.9 times the baseline values significantly improve MSK conditions.^6-8

Familiarity with the use of platelet-rich plasma to treat specific musculoskeletal conditions is essential for FPs who frequently are asked by patients about whether it is right for them.

For MSK purposes, PRP may be injected into intratendinous, peritendinous, and intra-articular spaces. Currently, there is no consensus regarding injection frequency. Many studies have incorporated single-­injection protocols, while some have used 2 to 3 injections repeated over several weeks to months. PRP commonly is injected at point-of-care without requiring storage.

Contraindications. PRP has been shown to be safe, with most adverse effects attributed to local injection site pain, bleeding, swelling, and bruising.⁹

Contraindications to PRP include active malignancy or recent remission from malignancy with the exception of nonmetastatic skin tumors.¹⁰ PRP is not recommended for patients with an allergy to manufacturing components (eg, dimethyl sulfoxide), thrombocytopenia, nonsteroidal anti-­inflammatory drug use within 2 weeks, active infection causing fever, and local infection at the injection site.¹⁰ Since local anesthetics may impair platelet function, they should not be given at the same injection site as PRP.¹⁰

Continue to: Cost

Cost. PRP is not covered by most insurance plans.^11,12 The cost for PRP may range from $500 to $2500 for a single injection.¹²

Evidence-based summary by condition

Knee osteoarthritis

❯❯❯ Consider using PRP

Knee osteoarthritis (OA) is a common cause of pain and disability. Treatment options include physical therapy, pharmacotherapy, and surgery. PRP has gained popularity as a nonsurgical option. A recent meta-analysis by Costa et al¹³ of 40 studies with 3035 participants comparing intra-articular PRP with hyaluronic acid (HA), corticosteroid, and saline injections, found that PRP appears to be more effective or as effective as other nonsurgical modalities. However, due to study heterogeneity and high risk for bias, the authors could not recommend PRP for knee OA in clinical practice.¹³

Despite Costa et al’s findings, reproducible data have demonstrated the superiority of PRP over other nonsurgical treatment options for knee OA. A 2021 systematic review and meta-analysis of 18 randomized controlled trials (RCTs; N = 811) by Belk et al⁶ comparing PRP to HA injections showed a higher mean improvement in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores in the PRP group compared to the HA group (44.7% vs 12.6%, respectively; P < .01).⁶ Six of 11 studies using the visual analog scale (VAS) for pain reported significantly less pain in the PRP group compared to the HA group (P < .05).⁶ The mean follow-up time was 11.1 months.⁶ Three of 6 studies reported improved subjective International Knee Documentation Committee (IKDC) scores (range from 0-100, with higher scores representing higher levels of function and lower levels of symptoms) in the PRP group compared to the HA group: 75.7 ± 15.1 vs 65.6 ± 16.9 (P = .004); 65.5 ± 3.6 vs 55.8 ± 3.8 (P = .01); and 60.8 ± 9.8 vs 48.4 ± 6.2 (P < .05).⁶ There was concern for moderate-to-high heterogeneity.⁶

PRP has been shown to be safe, with most adverse effects attributed to local injectionsite pain, bleeding, swelling, and bruising.

Other systematic reviews and meta-­analyses found similar efficacy of PRP for knee OA, including improved WOMAC scores and patient-reported outcomes (eg, pain, physical function, stiffness) compared to other injectable options.^14,15 A systematic review of 14 RCTs (N = 1423) by Shen et al¹⁵ showed improved WOMAC scores at 3 months (mean differences [MD] = –14.53; 95% CI, –29.97 to –7.09; P < .001), 6 months (MD = –18.21; 95% CI, –27.84 to –8.95; P < .001), and 12 months (MD = –19.45; 95% CI, –26.90 to –12.82; P < .001) in favor of PRP vs controls (saline placebo, ozone, corticosteroids, HA).¹⁵

Despite a lack of consensus regarding the optimal preparation of PRP for knee OA, another recent RCT (N = 192) found significant improvement in mean subjective IKDC scores in the LR-PRP group (45.5 ± 15.5 to 60.7 ± 21.1; P < .0005) and the LP-PRP group (46.8 ± 15.8 to 62.9 ± 19.9; P < .0005), indicating efficacy regardless of PRP type.⁴

Continue to: Ankle osteoarthritis

❯ ❯ ❯   Additional research is needed

Ankle OA affects 3.4% of all adults and is more common in the younger population than knee or hip OA.¹⁶ An RCT (N = 100) investigating PRP vs placebo (saline) injections showed no statistically significant difference in American Orthopedic Foot and Ankle Society scores evaluating pain and function over 26 weeks (–2 points; 95% CI, –5 to 1; P = .16).¹⁶ Limitations to this study include its small sample size and the PRP formulation used. (The intervention group received 2 injections of 2 mL of PRP, and the platelet concentration was not reported.)¹⁶

A 2020 systematic review and meta-­analysis of 4 RCTs and 5 case series by Evans et al¹⁷ concluded that PRP improves pain and function in small-joint OA compared to controls of saline, corticosteroids, and HA.¹⁷ One of the case series (N = 20) included in the study demonstrated improvement in ankle OA pain and function scores at 24 weeks posttreatment (P = .04), although improvement in pain and function peaked at 12 weeks.¹⁷ In addition, a 2017 retrospective study (N = 20) from the review reported improved VAS scores and function at 17 months following 4 injections of PRP over 4 weeks (P < .001).¹⁷ Given that RCT data found no benefit with PRP in treating small-joint OA, additional research is indicated.

Hip osteoarthritis

❯ ❯ ❯   Additional research is needed

Symptomatic hip OA occurs in 40% of adults older than 65 years, with a higher prevalence in women.¹⁸ Currently, corticosteroid injections are the only intra-articular therapy recommended by international guidelines for hip OA.¹⁹ A systematic review and meta-analysis comparing PRP to HA injections that included 4 RCTs (N = 303) showed a statistically significant reduction in VAS scores at 2 months in the PRP group compared to the HA group (weighted mean difference [WMD] = –0.376; 95% CI, –0.614 to –0.138; P = .002).¹⁸ However, there were no significant differences in VAS scores between the PRP and HA groups at 6 months (WMD = –0.141; 95% CI, –0.401 to 0.119; P = .289) and 12 months (WMD = –0.083; 95% CI, –0.343 to 0.117; P = .534). Likewise, no significant differences were found in WOMAC scores at 6 months (WMD = –2.841; 95% CI, –6.248 to 0.565; P = .102) and 12 months (WMD = –3.134; 95% CI, –6.624 to 0.356; P = .078) and Harris Hip Scores (HHS) at 6 months (WMD = 2.782; 95% CI, –6.639 to 12.203; P =.563) and 12 months (WMD = 0.706; 95% CI, –6.333 to 7.745; P = .844).¹⁸

A systematic review of 6 RCTs (N = 408) by Belk et al²⁰ comparing PRP to HA for hip OA found similar short-term improvements in WOMAC scores (standardized mean differences [SMD] = 0.27; 95% CI, –0.05 to 0.59; P = .09), VAS scores (MD = 0.59; 95% CI, –0.741 to 1.92; P = .39), and HHS (MD = -0.81; 95% CI, –10.06 to 8.43; P = .93).The average follow-up time was 12.2 and 11.9 months for the PRP and HA groups, respectively.²⁰

LR-PRP, which was used in 1 of the 6 RCTs, showed improvement in VAS scores and HHS from baseline, but no significant difference compared to HA at the latest follow-­up.²⁰ A pooled subanalysis of the 3 studies that used LP-PRP found no difference in WOMAC scores between the PRP and HA groups (SMD = 0.42; 95% CI, –0.01 to 0.86; P = .06).²⁰ Future studies comparing the efficacy of intra-articular steroid vs PRP for hip OA would be beneficial.¹⁸

Continue to: Rotator cuff tendinopathy

Rotator cuff tendinopathy

❯ ❯ ❯   Consider PRP for short-term pain relief

Painful conditions of the rotator cuff include impingement syndrome, tendonitis, and partial and complete tears. A 2021 RCT (N = 58) by Dadgostar et al²¹ comparing PRP injection to corticosteroid therapy (methylprednisolone and lidocaine) for the treatment of rotator cuff tendinopathy showed significant improvement in VAS scores at 3 months in the PRP group compared to the corticosteroid group (6.66 ± 2.26 to 3.08 ± 2.14 vs 5.53 ± 1.80 to 3.88 ± 1.99, respectively; P = .023). There also were more significant improvements in adduction in the PRP group compared to the corticosteroid group (20.50° ± 8.23° to 28° ± 3.61° vs 23.21° ± 7.09° to 28.46° ± 4.18°, respectively; P = .011), and external rotation (59.66° ± 23.81° to 76.66° ± 18.30° vs 57.14°± 24.69° to 65.57° ± 26.39° for the PRP and corticosteroid groups, respectively; P = .036).²¹

Another RCT (N = 99) by Kwong et al²² comparing PRP to corticosteroids found similar short-term advantages of LP-PRP with an improved VAS score (–13.6 vs 0.4; P = .03), American Shoulder and Elbow Surgeons score (13.0 vs 2.9; P = .02), and Western Ontario Rotator Cuff Index score (16.8 vs 5.8; P = .03).However, there was no long-term benefit of PRP over corticosteroids found at 12 months.²²

A 2021 systematic review and meta-­analysis by Hamid et al²³ that included 8 RCTs (N = 976) favored PRP over control (no injection, saline injections, and/or shoulder rehabilitation) with improved VAS scores at 12 months (SMD = –0.5; 95% CI, –0.7 to –0.2; P < .001).The evidence on functional outcome was mixed. Data pooled from 2 studies (n = 228) found better Shoulder Pain and Disability Index (SPADI) scores compared to controls at 3- and 6-month follow-ups. However, there were no significant differences in Disabilities of the Arm, Shoulder and Hand (DASH) scores between the 2 groups.²³

Patellar tendinopathy

❯ ❯ ❯   Consider using PRP for return to sport

Patellar tendinopathy, a common MSK condition encountered in the primary care setting, has an overall prevalence of 22% in elite athletes at some point in their career.^{24 ­}Nonsurgical management options include rest, ice, eccentric and isometric exercises, anti-­inflammatory drugs, extracorporeal shock wave therapy (ESWT), and dry needling (DN).

Currently, corticosteroid injections are the only intraarticular therapy recommended by international guidelines for hip OA.

A 2014 RCT (N = 23) evaluating DN vs PRP for patellar tendinopathy favored PRP with improved VAS scores (mean ± SD = 25.4 ± 23.2 points; P = .01 vs 5.2 ± 12.5 points; P = .20) at 12 weeks (P = .02). However, at ≥ 26 weeks, the improvement in pain and function scores was similar between the DN and PRP groups (33.2 ± 14.0 points; P = .001 vs 28.9 ± 25.2 points; P = .01). Notably, there was significantly more improvement in the PRP group at 12 weeks (P = .02) but not at 26 weeks (P = .66).²⁵

Continue to: Another perspective study...

Another prospective study (N = 31) comparing PRP to physiotherapy showed a greater improvement in sport activity level reflected by the Tegner score in the PRP group (percentage improvement, 39 ± 22%) compared to control (20 ± 27%; P = .048) at 6 months.⁷

A recent RCT (N = 20) revealed improved VAS scores at 6 months with rehabilitation paired with either bone marrow mesenchymal stem cells (BM-MSC) or LP-PRP when compared with baseline (BM-MSC group: 4.23 ± 2.13 to 2.52 ± 2.37; P = .0621; LP-PRP group: 3.10 ± 1.20 to 1.13 ± 1.25; P = .0083). Pain was significantly reduced during sport play in both groups at 6 months when compared with baseline (BM-MSC group: 6.91 ± 1.11 to 3.06 ± 2.89, P = .0049; PRP group: 7.03 ± 1.42 to 1.94 ± 1.24, P = .0001).²⁶

A 2019 systematic review and meta-analysis (N = 2530) demonstrated greater improvements in Victorian Institute of Sport Assessment scale for patellar tendinopathy (VISA-P) with multiple injections of PRP (38.7 points; 95% CI, 26.3-51.2 points) compared to single injections of PRP (24.3 points; 95% CI, 18.2-30.5 points), eccentric exercise (28.3 points; 95% CI, 18.9-37.8 points) and ESWT (27.4 points; 95% CI, 10.0-39.8 points) after 6 months.²⁷ In contrast, an RCT (n = 57) comparing a single injection of LR-PRP or LP-PRP was no more effective than a single injection of saline for improvement in mean VISA-P scores (P > .05) at 1 year.²⁸

Lateral epicondylitis

❯ ❯ ❯   Consider using PRP

Lateral epicondylitis (“tennis elbow”) is caused by overuse of the elbow extensors at the site of the lateral epicondyle. Chronic lateral epicondylosis involves tissue degeneration and microtrauma.Most cases of epicondylar tendinopathies are treated nonoperatively, with corticosteroid injections being a mainstay of treatment despite their short-term benefit²⁹ and potential to deteriorate connective tissue over time. Recent studies suggest PRP therapy for epicondylitis and epicondylosis may increase long-term pain relief and improve function.

The evidence on functional outcome of platelet-rich plasma for rotator cuff tendinopathy is mixed.

A 2017 systematic review and meta-­analysis of 16 RCTs (N = 1018) concluded PRP was more efficacious than control injections (bupivacaine) for pain reduction in tendinopathies (effect size = 0.47; 95% CI, 0.22-0.72).³⁰ In the review, lateral epicondylitis was evaluated in 12 studies and was most responsive to PRP (effect size = 0.57) when compared to control injection.³⁰ In another systematic review (5 RCTs; 250 patients), corticosteroid injections improved pain within the first 6 weeks of treatment. However, PRP outperformed corticosteroid in VAS scores (21.3 ± 28.1 vs 42.4 ± 26.8) and DASH scores (17.6 ± 24.0 vs 36.5 ± 23.8) (P < .001) at 2 years.³¹

Continue to: A 2022 systematic review...

A 2022 systematic review and meta-­analysis (26 studies; N = 1040) comparing scores at baseline vs 2 years post-PRP showed improvement in VAS scores (7.4 ± 1.30 vs 3.71 ± 2.35; P < .001), DASH scores (60.8 ± 12.5 vs 13.0 ± 18.5; P < .001), Patient-Rated Tennis Elbow Evaluation (55.6 ± 14.7 vs 48.8 ± 4.1; P < .001), and Mayo Clinic Performance Index (55.5 ± 6.1 vs 93.0 ± 6.7; P < .001).³²

Regarding the therapeutic effects of different PRP types in lateral epicondylitis, a 2022 systematic review of 33 studies (N = 2420) found improved function and pain relief with LR-PRP and LP-PRP with no significant differences.³³ Pretreatment VAS scores in the LR-PRP group, which ranged from 6.1 to 8.0, improved to 1.5 to 4.0 at 3 months and 0.6 to 3.3 after 1 year.³³ Similarly, pretreatment VAS scores in the LP-PRP group, which ranged from 4.2 to 8.4, improved to 1.6 to 5.9 at 3 months and 0.7 to 2.7 after 1 year.³⁴ DASH scores also improved in the LR-PRP and LP-PRP groups, with pretreatment scores (LR-PRP, 47.0 to 54.3; LP-PRP, 30.0 to 67.7) improving to 20.0 to 22.0 and 5.5 to 19.0, respectively, at 1 year.³³

Achilles tendinopathy

❯ ❯ ❯   Do not use PRP; evidence is lacking

Achilles tendinopathy, caused by chronic overuse and overload resulting in microtrauma and poor tissue healing, typically occurs in the most poorly vascularized portion of the tendon and is common in runners. First-line treatments for Achilles tendinopathy include eccentric strength training and anti-­inflammatory drugs.^34,35 Corticosteroid injections are not recommended, given concern for degraded tendon tissue over time and worse function.³⁴

A 2020 systematic review of 11 randomized and nonrandomized clinical trials (N = 406) found PRP improved Victorian Institute of Sports Assessment—Achilles (VISA-A) scores at 24 weeks compared to other nonsurgical treatment options (41.2 vs 70.12; P < .018).³⁴ However, a higher-quality 2021 systematic review and meta-analysis of 4 RCTs (N = 170) comparing PRP injections with placebo showed no significant difference in VISA-A scores at 3 months (0.23; 95% CI, –0.45 to 0.91), 6 months (0.83; 95% CI, –0.26 to 1.92), and 12 months (0.83; 95% CI, –0.77 to 2.44).³⁶ Therefore, further studies are warranted to evaluate the benefit of PRP injections for Achilles tendinopathy.

Conclusions

While high-quality studies support the use of PRP for knee OA and lateral epicondylitis, they have a moderate-to-high risk for bias. Several RCTs show that PRP provides superior short-term pain relief and range of motion compared to corticosteroids for rotator cuff tendinopathy. Multiple injections of PRP for patellar tendinopathy may accelerate return to sport and improve symptoms over the long term. However, current evidence does not support PRP therapy for Achilles tendinopathy. Given variability in PRP preparation, an accurate interpretation of the literature regarding its use in MSK conditions is recommended (TABLE^{4,6,7,14-18,20-23,25-28,30-34,36}).

Continue to: Concerning the effectiveness of PRP...

Concerning the effectiveness of PRP, it is important to consider early publication bias. Although recent studies have shown its benefits,^6,14,15,37 additional studies comparing PRP to placebo will help demonstrate its efficacy. Interestingly, a literature search by Bar-Or et al³⁸ found intra-articular saline may have a therapeutic effect on knee OA and confound findings when used as a placebo.

RCT data showed no benefit with platelet-rich plasma in treating small-joint osteoarthritis.

Recognizing the presence or lack of clinically significant improvement in the literature is important. For example, while some recent studies have shown PRP exceeds the minimal clinically significant difference for knee OA and lateral epicondylitis, others have not.^32,37 A 2021 systematic review of 11 clinical practice guidelines for the use of PRP in knee OA found that 9 were “uncertain or unable to make a recommendation” and 2 recommended against it.³⁹

In its 2021 position statement for the responsible use of regenerative medicine, the American Medical Society for Sports Medicine includes guidance on integrating orthobiologics into clinical practice. The guideline emphasizes informed consent and provides an evidence-based rationale for using PRP in certain patient populations (lateral epicondylitis and younger patients with mild-to-­moderate knee OA), recommending its use only after exhausting other conservative options.⁴⁰ Patients should be referred to physicians with experience using PRP and image-guided procedures.

CORRESPONDENCE
Gregory D. Bentz Jr, MD, 3640 High Street Suite 3B, Portsmouth, VA 23707; bentzgd@evms.edu

Platelet-rich plasma (PRP) injections have become a popular treatment option in a variety of specialties including sports medicine, maxillofacial surgery, dermatology, cosmetology, and reproductive medicine.¹ PRP is an autologous blood product derived from whole blood, using a centrifuge to isolate a concentrated layer of platelets. The ­a-granules in platelets release transforming growth factor b 1, vascular endothelial growth factor, platelet-derived growth factor, basic fibroblast growth factor, epidermal growth factor, insulin-like growth factor 1, and other mediatorsthat enhance the natural healing process.²

When patients ask. Familiarity with the use of PRP to treat specific musculoskeletal (MSK) conditions is essential for family physicians who frequently are asked by patients about whether PRP is right for them. These patients may have experienced failure of medication therapy or declined surgical intervention, or may not be surgical candidates. This review details the evidence surrounding common intra-articular and extra-articular applications of PRP. But first, a word about how PRP is prepared, its contraindications, and costs.

Preparation and types of PRP

Although there are many commercial systems for preparing PRP, there is no consensus on the optimal formulation.² Other terms for PRP, such as autologous concentrated platelets and super-concentrated platelets, are based on concentration of red blood cells, leukocytes, and fibrin.³ PRP therapies usually are categorized as leukocyte-rich PRP (LR-PRP) or leukocyte-poor PRP (LP-PRP), based on neutrophil concentrations that are above and below baseline.² Leukocyte concentration is one of the most debated topics in PRP therapy.⁴

Common commercially available preparation systems produce platelet concentrations between 3 to 6 times the baseline platelet count.⁵ Although there is no universally agreed upon PRP formulation, studies have shown 2 centrifugation cycles (“double-spun” or “dual centrifugation”) that yield platelet concentrations between 1.8 to 1.9 times the baseline values significantly improve MSK conditions.^6-8

Familiarity with the use of platelet-rich plasma to treat specific musculoskeletal conditions is essential for FPs who frequently are asked by patients about whether it is right for them.

For MSK purposes, PRP may be injected into intratendinous, peritendinous, and intra-articular spaces. Currently, there is no consensus regarding injection frequency. Many studies have incorporated single-­injection protocols, while some have used 2 to 3 injections repeated over several weeks to months. PRP commonly is injected at point-of-care without requiring storage.

Contraindications. PRP has been shown to be safe, with most adverse effects attributed to local injection site pain, bleeding, swelling, and bruising.⁹

Contraindications to PRP include active malignancy or recent remission from malignancy with the exception of nonmetastatic skin tumors.¹⁰ PRP is not recommended for patients with an allergy to manufacturing components (eg, dimethyl sulfoxide), thrombocytopenia, nonsteroidal anti-­inflammatory drug use within 2 weeks, active infection causing fever, and local infection at the injection site.¹⁰ Since local anesthetics may impair platelet function, they should not be given at the same injection site as PRP.¹⁰

Continue to: Cost

Cost. PRP is not covered by most insurance plans.^11,12 The cost for PRP may range from $500 to $2500 for a single injection.¹²

Evidence-based summary by condition

Knee osteoarthritis

❯❯❯ Consider using PRP

Knee osteoarthritis (OA) is a common cause of pain and disability. Treatment options include physical therapy, pharmacotherapy, and surgery. PRP has gained popularity as a nonsurgical option. A recent meta-analysis by Costa et al¹³ of 40 studies with 3035 participants comparing intra-articular PRP with hyaluronic acid (HA), corticosteroid, and saline injections, found that PRP appears to be more effective or as effective as other nonsurgical modalities. However, due to study heterogeneity and high risk for bias, the authors could not recommend PRP for knee OA in clinical practice.¹³

Despite Costa et al’s findings, reproducible data have demonstrated the superiority of PRP over other nonsurgical treatment options for knee OA. A 2021 systematic review and meta-analysis of 18 randomized controlled trials (RCTs; N = 811) by Belk et al⁶ comparing PRP to HA injections showed a higher mean improvement in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores in the PRP group compared to the HA group (44.7% vs 12.6%, respectively; P < .01).⁶ Six of 11 studies using the visual analog scale (VAS) for pain reported significantly less pain in the PRP group compared to the HA group (P < .05).⁶ The mean follow-up time was 11.1 months.⁶ Three of 6 studies reported improved subjective International Knee Documentation Committee (IKDC) scores (range from 0-100, with higher scores representing higher levels of function and lower levels of symptoms) in the PRP group compared to the HA group: 75.7 ± 15.1 vs 65.6 ± 16.9 (P = .004); 65.5 ± 3.6 vs 55.8 ± 3.8 (P = .01); and 60.8 ± 9.8 vs 48.4 ± 6.2 (P < .05).⁶ There was concern for moderate-to-high heterogeneity.⁶

PRP has been shown to be safe, with most adverse effects attributed to local injectionsite pain, bleeding, swelling, and bruising.

Other systematic reviews and meta-­analyses found similar efficacy of PRP for knee OA, including improved WOMAC scores and patient-reported outcomes (eg, pain, physical function, stiffness) compared to other injectable options.^14,15 A systematic review of 14 RCTs (N = 1423) by Shen et al¹⁵ showed improved WOMAC scores at 3 months (mean differences [MD] = –14.53; 95% CI, –29.97 to –7.09; P < .001), 6 months (MD = –18.21; 95% CI, –27.84 to –8.95; P < .001), and 12 months (MD = –19.45; 95% CI, –26.90 to –12.82; P < .001) in favor of PRP vs controls (saline placebo, ozone, corticosteroids, HA).¹⁵

Despite a lack of consensus regarding the optimal preparation of PRP for knee OA, another recent RCT (N = 192) found significant improvement in mean subjective IKDC scores in the LR-PRP group (45.5 ± 15.5 to 60.7 ± 21.1; P < .0005) and the LP-PRP group (46.8 ± 15.8 to 62.9 ± 19.9; P < .0005), indicating efficacy regardless of PRP type.⁴

Continue to: Ankle osteoarthritis

❯ ❯ ❯   Additional research is needed

Ankle OA affects 3.4% of all adults and is more common in the younger population than knee or hip OA.¹⁶ An RCT (N = 100) investigating PRP vs placebo (saline) injections showed no statistically significant difference in American Orthopedic Foot and Ankle Society scores evaluating pain and function over 26 weeks (–2 points; 95% CI, –5 to 1; P = .16).¹⁶ Limitations to this study include its small sample size and the PRP formulation used. (The intervention group received 2 injections of 2 mL of PRP, and the platelet concentration was not reported.)¹⁶

A 2020 systematic review and meta-­analysis of 4 RCTs and 5 case series by Evans et al¹⁷ concluded that PRP improves pain and function in small-joint OA compared to controls of saline, corticosteroids, and HA.¹⁷ One of the case series (N = 20) included in the study demonstrated improvement in ankle OA pain and function scores at 24 weeks posttreatment (P = .04), although improvement in pain and function peaked at 12 weeks.¹⁷ In addition, a 2017 retrospective study (N = 20) from the review reported improved VAS scores and function at 17 months following 4 injections of PRP over 4 weeks (P < .001).¹⁷ Given that RCT data found no benefit with PRP in treating small-joint OA, additional research is indicated.

Hip osteoarthritis

❯ ❯ ❯   Additional research is needed

Symptomatic hip OA occurs in 40% of adults older than 65 years, with a higher prevalence in women.¹⁸ Currently, corticosteroid injections are the only intra-articular therapy recommended by international guidelines for hip OA.¹⁹ A systematic review and meta-analysis comparing PRP to HA injections that included 4 RCTs (N = 303) showed a statistically significant reduction in VAS scores at 2 months in the PRP group compared to the HA group (weighted mean difference [WMD] = –0.376; 95% CI, –0.614 to –0.138; P = .002).¹⁸ However, there were no significant differences in VAS scores between the PRP and HA groups at 6 months (WMD = –0.141; 95% CI, –0.401 to 0.119; P = .289) and 12 months (WMD = –0.083; 95% CI, –0.343 to 0.117; P = .534). Likewise, no significant differences were found in WOMAC scores at 6 months (WMD = –2.841; 95% CI, –6.248 to 0.565; P = .102) and 12 months (WMD = –3.134; 95% CI, –6.624 to 0.356; P = .078) and Harris Hip Scores (HHS) at 6 months (WMD = 2.782; 95% CI, –6.639 to 12.203; P =.563) and 12 months (WMD = 0.706; 95% CI, –6.333 to 7.745; P = .844).¹⁸

A systematic review of 6 RCTs (N = 408) by Belk et al²⁰ comparing PRP to HA for hip OA found similar short-term improvements in WOMAC scores (standardized mean differences [SMD] = 0.27; 95% CI, –0.05 to 0.59; P = .09), VAS scores (MD = 0.59; 95% CI, –0.741 to 1.92; P = .39), and HHS (MD = -0.81; 95% CI, –10.06 to 8.43; P = .93).The average follow-up time was 12.2 and 11.9 months for the PRP and HA groups, respectively.²⁰

LR-PRP, which was used in 1 of the 6 RCTs, showed improvement in VAS scores and HHS from baseline, but no significant difference compared to HA at the latest follow-­up.²⁰ A pooled subanalysis of the 3 studies that used LP-PRP found no difference in WOMAC scores between the PRP and HA groups (SMD = 0.42; 95% CI, –0.01 to 0.86; P = .06).²⁰ Future studies comparing the efficacy of intra-articular steroid vs PRP for hip OA would be beneficial.¹⁸

Continue to: Rotator cuff tendinopathy

Rotator cuff tendinopathy

❯ ❯ ❯   Consider PRP for short-term pain relief

Painful conditions of the rotator cuff include impingement syndrome, tendonitis, and partial and complete tears. A 2021 RCT (N = 58) by Dadgostar et al²¹ comparing PRP injection to corticosteroid therapy (methylprednisolone and lidocaine) for the treatment of rotator cuff tendinopathy showed significant improvement in VAS scores at 3 months in the PRP group compared to the corticosteroid group (6.66 ± 2.26 to 3.08 ± 2.14 vs 5.53 ± 1.80 to 3.88 ± 1.99, respectively; P = .023). There also were more significant improvements in adduction in the PRP group compared to the corticosteroid group (20.50° ± 8.23° to 28° ± 3.61° vs 23.21° ± 7.09° to 28.46° ± 4.18°, respectively; P = .011), and external rotation (59.66° ± 23.81° to 76.66° ± 18.30° vs 57.14°± 24.69° to 65.57° ± 26.39° for the PRP and corticosteroid groups, respectively; P = .036).²¹

Another RCT (N = 99) by Kwong et al²² comparing PRP to corticosteroids found similar short-term advantages of LP-PRP with an improved VAS score (–13.6 vs 0.4; P = .03), American Shoulder and Elbow Surgeons score (13.0 vs 2.9; P = .02), and Western Ontario Rotator Cuff Index score (16.8 vs 5.8; P = .03).However, there was no long-term benefit of PRP over corticosteroids found at 12 months.²²

A 2021 systematic review and meta-­analysis by Hamid et al²³ that included 8 RCTs (N = 976) favored PRP over control (no injection, saline injections, and/or shoulder rehabilitation) with improved VAS scores at 12 months (SMD = –0.5; 95% CI, –0.7 to –0.2; P < .001).The evidence on functional outcome was mixed. Data pooled from 2 studies (n = 228) found better Shoulder Pain and Disability Index (SPADI) scores compared to controls at 3- and 6-month follow-ups. However, there were no significant differences in Disabilities of the Arm, Shoulder and Hand (DASH) scores between the 2 groups.²³

Patellar tendinopathy

❯ ❯ ❯   Consider using PRP for return to sport

Patellar tendinopathy, a common MSK condition encountered in the primary care setting, has an overall prevalence of 22% in elite athletes at some point in their career.^{24 ­}Nonsurgical management options include rest, ice, eccentric and isometric exercises, anti-­inflammatory drugs, extracorporeal shock wave therapy (ESWT), and dry needling (DN).

Currently, corticosteroid injections are the only intraarticular therapy recommended by international guidelines for hip OA.

A 2014 RCT (N = 23) evaluating DN vs PRP for patellar tendinopathy favored PRP with improved VAS scores (mean ± SD = 25.4 ± 23.2 points; P = .01 vs 5.2 ± 12.5 points; P = .20) at 12 weeks (P = .02). However, at ≥ 26 weeks, the improvement in pain and function scores was similar between the DN and PRP groups (33.2 ± 14.0 points; P = .001 vs 28.9 ± 25.2 points; P = .01). Notably, there was significantly more improvement in the PRP group at 12 weeks (P = .02) but not at 26 weeks (P = .66).²⁵

Continue to: Another perspective study...

Another prospective study (N = 31) comparing PRP to physiotherapy showed a greater improvement in sport activity level reflected by the Tegner score in the PRP group (percentage improvement, 39 ± 22%) compared to control (20 ± 27%; P = .048) at 6 months.⁷

A recent RCT (N = 20) revealed improved VAS scores at 6 months with rehabilitation paired with either bone marrow mesenchymal stem cells (BM-MSC) or LP-PRP when compared with baseline (BM-MSC group: 4.23 ± 2.13 to 2.52 ± 2.37; P = .0621; LP-PRP group: 3.10 ± 1.20 to 1.13 ± 1.25; P = .0083). Pain was significantly reduced during sport play in both groups at 6 months when compared with baseline (BM-MSC group: 6.91 ± 1.11 to 3.06 ± 2.89, P = .0049; PRP group: 7.03 ± 1.42 to 1.94 ± 1.24, P = .0001).²⁶

A 2019 systematic review and meta-analysis (N = 2530) demonstrated greater improvements in Victorian Institute of Sport Assessment scale for patellar tendinopathy (VISA-P) with multiple injections of PRP (38.7 points; 95% CI, 26.3-51.2 points) compared to single injections of PRP (24.3 points; 95% CI, 18.2-30.5 points), eccentric exercise (28.3 points; 95% CI, 18.9-37.8 points) and ESWT (27.4 points; 95% CI, 10.0-39.8 points) after 6 months.²⁷ In contrast, an RCT (n = 57) comparing a single injection of LR-PRP or LP-PRP was no more effective than a single injection of saline for improvement in mean VISA-P scores (P > .05) at 1 year.²⁸

Lateral epicondylitis

❯ ❯ ❯   Consider using PRP

Lateral epicondylitis (“tennis elbow”) is caused by overuse of the elbow extensors at the site of the lateral epicondyle. Chronic lateral epicondylosis involves tissue degeneration and microtrauma.Most cases of epicondylar tendinopathies are treated nonoperatively, with corticosteroid injections being a mainstay of treatment despite their short-term benefit²⁹ and potential to deteriorate connective tissue over time. Recent studies suggest PRP therapy for epicondylitis and epicondylosis may increase long-term pain relief and improve function.

The evidence on functional outcome of platelet-rich plasma for rotator cuff tendinopathy is mixed.

A 2017 systematic review and meta-­analysis of 16 RCTs (N = 1018) concluded PRP was more efficacious than control injections (bupivacaine) for pain reduction in tendinopathies (effect size = 0.47; 95% CI, 0.22-0.72).³⁰ In the review, lateral epicondylitis was evaluated in 12 studies and was most responsive to PRP (effect size = 0.57) when compared to control injection.³⁰ In another systematic review (5 RCTs; 250 patients), corticosteroid injections improved pain within the first 6 weeks of treatment. However, PRP outperformed corticosteroid in VAS scores (21.3 ± 28.1 vs 42.4 ± 26.8) and DASH scores (17.6 ± 24.0 vs 36.5 ± 23.8) (P < .001) at 2 years.³¹

Continue to: A 2022 systematic review...

A 2022 systematic review and meta-­analysis (26 studies; N = 1040) comparing scores at baseline vs 2 years post-PRP showed improvement in VAS scores (7.4 ± 1.30 vs 3.71 ± 2.35; P < .001), DASH scores (60.8 ± 12.5 vs 13.0 ± 18.5; P < .001), Patient-Rated Tennis Elbow Evaluation (55.6 ± 14.7 vs 48.8 ± 4.1; P < .001), and Mayo Clinic Performance Index (55.5 ± 6.1 vs 93.0 ± 6.7; P < .001).³²

Regarding the therapeutic effects of different PRP types in lateral epicondylitis, a 2022 systematic review of 33 studies (N = 2420) found improved function and pain relief with LR-PRP and LP-PRP with no significant differences.³³ Pretreatment VAS scores in the LR-PRP group, which ranged from 6.1 to 8.0, improved to 1.5 to 4.0 at 3 months and 0.6 to 3.3 after 1 year.³³ Similarly, pretreatment VAS scores in the LP-PRP group, which ranged from 4.2 to 8.4, improved to 1.6 to 5.9 at 3 months and 0.7 to 2.7 after 1 year.³⁴ DASH scores also improved in the LR-PRP and LP-PRP groups, with pretreatment scores (LR-PRP, 47.0 to 54.3; LP-PRP, 30.0 to 67.7) improving to 20.0 to 22.0 and 5.5 to 19.0, respectively, at 1 year.³³

Achilles tendinopathy

❯ ❯ ❯   Do not use PRP; evidence is lacking

Achilles tendinopathy, caused by chronic overuse and overload resulting in microtrauma and poor tissue healing, typically occurs in the most poorly vascularized portion of the tendon and is common in runners. First-line treatments for Achilles tendinopathy include eccentric strength training and anti-­inflammatory drugs.^34,35 Corticosteroid injections are not recommended, given concern for degraded tendon tissue over time and worse function.³⁴

A 2020 systematic review of 11 randomized and nonrandomized clinical trials (N = 406) found PRP improved Victorian Institute of Sports Assessment—Achilles (VISA-A) scores at 24 weeks compared to other nonsurgical treatment options (41.2 vs 70.12; P < .018).³⁴ However, a higher-quality 2021 systematic review and meta-analysis of 4 RCTs (N = 170) comparing PRP injections with placebo showed no significant difference in VISA-A scores at 3 months (0.23; 95% CI, –0.45 to 0.91), 6 months (0.83; 95% CI, –0.26 to 1.92), and 12 months (0.83; 95% CI, –0.77 to 2.44).³⁶ Therefore, further studies are warranted to evaluate the benefit of PRP injections for Achilles tendinopathy.

Conclusions

While high-quality studies support the use of PRP for knee OA and lateral epicondylitis, they have a moderate-to-high risk for bias. Several RCTs show that PRP provides superior short-term pain relief and range of motion compared to corticosteroids for rotator cuff tendinopathy. Multiple injections of PRP for patellar tendinopathy may accelerate return to sport and improve symptoms over the long term. However, current evidence does not support PRP therapy for Achilles tendinopathy. Given variability in PRP preparation, an accurate interpretation of the literature regarding its use in MSK conditions is recommended (TABLE^{4,6,7,14-18,20-23,25-28,30-34,36}).

Continue to: Concerning the effectiveness of PRP...

Concerning the effectiveness of PRP, it is important to consider early publication bias. Although recent studies have shown its benefits,^6,14,15,37 additional studies comparing PRP to placebo will help demonstrate its efficacy. Interestingly, a literature search by Bar-Or et al³⁸ found intra-articular saline may have a therapeutic effect on knee OA and confound findings when used as a placebo.

RCT data showed no benefit with platelet-rich plasma in treating small-joint osteoarthritis.

Recognizing the presence or lack of clinically significant improvement in the literature is important. For example, while some recent studies have shown PRP exceeds the minimal clinically significant difference for knee OA and lateral epicondylitis, others have not.^32,37 A 2021 systematic review of 11 clinical practice guidelines for the use of PRP in knee OA found that 9 were “uncertain or unable to make a recommendation” and 2 recommended against it.³⁹

In its 2021 position statement for the responsible use of regenerative medicine, the American Medical Society for Sports Medicine includes guidance on integrating orthobiologics into clinical practice. The guideline emphasizes informed consent and provides an evidence-based rationale for using PRP in certain patient populations (lateral epicondylitis and younger patients with mild-to-­moderate knee OA), recommending its use only after exhausting other conservative options.⁴⁰ Patients should be referred to physicians with experience using PRP and image-guided procedures.

CORRESPONDENCE
Gregory D. Bentz Jr, MD, 3640 High Street Suite 3B, Portsmouth, VA 23707; bentzgd@evms.edu

Platelet-rich plasma (PRP) injections have become a popular treatment option in a variety of specialties including sports medicine, maxillofacial surgery, dermatology, cosmetology, and reproductive medicine.¹ PRP is an autologous blood product derived from whole blood, using a centrifuge to isolate a concentrated layer of platelets. The ­a-granules in platelets release transforming growth factor b 1, vascular endothelial growth factor, platelet-derived growth factor, basic fibroblast growth factor, epidermal growth factor, insulin-like growth factor 1, and other mediatorsthat enhance the natural healing process.²

When patients ask. Familiarity with the use of PRP to treat specific musculoskeletal (MSK) conditions is essential for family physicians who frequently are asked by patients about whether PRP is right for them. These patients may have experienced failure of medication therapy or declined surgical intervention, or may not be surgical candidates. This review details the evidence surrounding common intra-articular and extra-articular applications of PRP. But first, a word about how PRP is prepared, its contraindications, and costs.

Preparation and types of PRP

Although there are many commercial systems for preparing PRP, there is no consensus on the optimal formulation.² Other terms for PRP, such as autologous concentrated platelets and super-concentrated platelets, are based on concentration of red blood cells, leukocytes, and fibrin.³ PRP therapies usually are categorized as leukocyte-rich PRP (LR-PRP) or leukocyte-poor PRP (LP-PRP), based on neutrophil concentrations that are above and below baseline.² Leukocyte concentration is one of the most debated topics in PRP therapy.⁴

Common commercially available preparation systems produce platelet concentrations between 3 to 6 times the baseline platelet count.⁵ Although there is no universally agreed upon PRP formulation, studies have shown 2 centrifugation cycles (“double-spun” or “dual centrifugation”) that yield platelet concentrations between 1.8 to 1.9 times the baseline values significantly improve MSK conditions.^6-8

Familiarity with the use of platelet-rich plasma to treat specific musculoskeletal conditions is essential for FPs who frequently are asked by patients about whether it is right for them.

For MSK purposes, PRP may be injected into intratendinous, peritendinous, and intra-articular spaces. Currently, there is no consensus regarding injection frequency. Many studies have incorporated single-­injection protocols, while some have used 2 to 3 injections repeated over several weeks to months. PRP commonly is injected at point-of-care without requiring storage.

Contraindications. PRP has been shown to be safe, with most adverse effects attributed to local injection site pain, bleeding, swelling, and bruising.⁹

Contraindications to PRP include active malignancy or recent remission from malignancy with the exception of nonmetastatic skin tumors.¹⁰ PRP is not recommended for patients with an allergy to manufacturing components (eg, dimethyl sulfoxide), thrombocytopenia, nonsteroidal anti-­inflammatory drug use within 2 weeks, active infection causing fever, and local infection at the injection site.¹⁰ Since local anesthetics may impair platelet function, they should not be given at the same injection site as PRP.¹⁰

Continue to: Cost

Cost. PRP is not covered by most insurance plans.^11,12 The cost for PRP may range from $500 to $2500 for a single injection.¹²

Evidence-based summary by condition

Knee osteoarthritis

❯❯❯ Consider using PRP

Knee osteoarthritis (OA) is a common cause of pain and disability. Treatment options include physical therapy, pharmacotherapy, and surgery. PRP has gained popularity as a nonsurgical option. A recent meta-analysis by Costa et al¹³ of 40 studies with 3035 participants comparing intra-articular PRP with hyaluronic acid (HA), corticosteroid, and saline injections, found that PRP appears to be more effective or as effective as other nonsurgical modalities. However, due to study heterogeneity and high risk for bias, the authors could not recommend PRP for knee OA in clinical practice.¹³

Despite Costa et al’s findings, reproducible data have demonstrated the superiority of PRP over other nonsurgical treatment options for knee OA. A 2021 systematic review and meta-analysis of 18 randomized controlled trials (RCTs; N = 811) by Belk et al⁶ comparing PRP to HA injections showed a higher mean improvement in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores in the PRP group compared to the HA group (44.7% vs 12.6%, respectively; P < .01).⁶ Six of 11 studies using the visual analog scale (VAS) for pain reported significantly less pain in the PRP group compared to the HA group (P < .05).⁶ The mean follow-up time was 11.1 months.⁶ Three of 6 studies reported improved subjective International Knee Documentation Committee (IKDC) scores (range from 0-100, with higher scores representing higher levels of function and lower levels of symptoms) in the PRP group compared to the HA group: 75.7 ± 15.1 vs 65.6 ± 16.9 (P = .004); 65.5 ± 3.6 vs 55.8 ± 3.8 (P = .01); and 60.8 ± 9.8 vs 48.4 ± 6.2 (P < .05).⁶ There was concern for moderate-to-high heterogeneity.⁶

PRP has been shown to be safe, with most adverse effects attributed to local injectionsite pain, bleeding, swelling, and bruising.

Other systematic reviews and meta-­analyses found similar efficacy of PRP for knee OA, including improved WOMAC scores and patient-reported outcomes (eg, pain, physical function, stiffness) compared to other injectable options.^14,15 A systematic review of 14 RCTs (N = 1423) by Shen et al¹⁵ showed improved WOMAC scores at 3 months (mean differences [MD] = –14.53; 95% CI, –29.97 to –7.09; P < .001), 6 months (MD = –18.21; 95% CI, –27.84 to –8.95; P < .001), and 12 months (MD = –19.45; 95% CI, –26.90 to –12.82; P < .001) in favor of PRP vs controls (saline placebo, ozone, corticosteroids, HA).¹⁵

Despite a lack of consensus regarding the optimal preparation of PRP for knee OA, another recent RCT (N = 192) found significant improvement in mean subjective IKDC scores in the LR-PRP group (45.5 ± 15.5 to 60.7 ± 21.1; P < .0005) and the LP-PRP group (46.8 ± 15.8 to 62.9 ± 19.9; P < .0005), indicating efficacy regardless of PRP type.⁴

Continue to: Ankle osteoarthritis

❯ ❯ ❯   Additional research is needed

Ankle OA affects 3.4% of all adults and is more common in the younger population than knee or hip OA.¹⁶ An RCT (N = 100) investigating PRP vs placebo (saline) injections showed no statistically significant difference in American Orthopedic Foot and Ankle Society scores evaluating pain and function over 26 weeks (–2 points; 95% CI, –5 to 1; P = .16).¹⁶ Limitations to this study include its small sample size and the PRP formulation used. (The intervention group received 2 injections of 2 mL of PRP, and the platelet concentration was not reported.)¹⁶

A 2020 systematic review and meta-­analysis of 4 RCTs and 5 case series by Evans et al¹⁷ concluded that PRP improves pain and function in small-joint OA compared to controls of saline, corticosteroids, and HA.¹⁷ One of the case series (N = 20) included in the study demonstrated improvement in ankle OA pain and function scores at 24 weeks posttreatment (P = .04), although improvement in pain and function peaked at 12 weeks.¹⁷ In addition, a 2017 retrospective study (N = 20) from the review reported improved VAS scores and function at 17 months following 4 injections of PRP over 4 weeks (P < .001).¹⁷ Given that RCT data found no benefit with PRP in treating small-joint OA, additional research is indicated.

Hip osteoarthritis

❯ ❯ ❯   Additional research is needed

Symptomatic hip OA occurs in 40% of adults older than 65 years, with a higher prevalence in women.¹⁸ Currently, corticosteroid injections are the only intra-articular therapy recommended by international guidelines for hip OA.¹⁹ A systematic review and meta-analysis comparing PRP to HA injections that included 4 RCTs (N = 303) showed a statistically significant reduction in VAS scores at 2 months in the PRP group compared to the HA group (weighted mean difference [WMD] = –0.376; 95% CI, –0.614 to –0.138; P = .002).¹⁸ However, there were no significant differences in VAS scores between the PRP and HA groups at 6 months (WMD = –0.141; 95% CI, –0.401 to 0.119; P = .289) and 12 months (WMD = –0.083; 95% CI, –0.343 to 0.117; P = .534). Likewise, no significant differences were found in WOMAC scores at 6 months (WMD = –2.841; 95% CI, –6.248 to 0.565; P = .102) and 12 months (WMD = –3.134; 95% CI, –6.624 to 0.356; P = .078) and Harris Hip Scores (HHS) at 6 months (WMD = 2.782; 95% CI, –6.639 to 12.203; P =.563) and 12 months (WMD = 0.706; 95% CI, –6.333 to 7.745; P = .844).¹⁸

A systematic review of 6 RCTs (N = 408) by Belk et al²⁰ comparing PRP to HA for hip OA found similar short-term improvements in WOMAC scores (standardized mean differences [SMD] = 0.27; 95% CI, –0.05 to 0.59; P = .09), VAS scores (MD = 0.59; 95% CI, –0.741 to 1.92; P = .39), and HHS (MD = -0.81; 95% CI, –10.06 to 8.43; P = .93).The average follow-up time was 12.2 and 11.9 months for the PRP and HA groups, respectively.²⁰

LR-PRP, which was used in 1 of the 6 RCTs, showed improvement in VAS scores and HHS from baseline, but no significant difference compared to HA at the latest follow-­up.²⁰ A pooled subanalysis of the 3 studies that used LP-PRP found no difference in WOMAC scores between the PRP and HA groups (SMD = 0.42; 95% CI, –0.01 to 0.86; P = .06).²⁰ Future studies comparing the efficacy of intra-articular steroid vs PRP for hip OA would be beneficial.¹⁸

Continue to: Rotator cuff tendinopathy

Rotator cuff tendinopathy

❯ ❯ ❯   Consider PRP for short-term pain relief

Painful conditions of the rotator cuff include impingement syndrome, tendonitis, and partial and complete tears. A 2021 RCT (N = 58) by Dadgostar et al²¹ comparing PRP injection to corticosteroid therapy (methylprednisolone and lidocaine) for the treatment of rotator cuff tendinopathy showed significant improvement in VAS scores at 3 months in the PRP group compared to the corticosteroid group (6.66 ± 2.26 to 3.08 ± 2.14 vs 5.53 ± 1.80 to 3.88 ± 1.99, respectively; P = .023). There also were more significant improvements in adduction in the PRP group compared to the corticosteroid group (20.50° ± 8.23° to 28° ± 3.61° vs 23.21° ± 7.09° to 28.46° ± 4.18°, respectively; P = .011), and external rotation (59.66° ± 23.81° to 76.66° ± 18.30° vs 57.14°± 24.69° to 65.57° ± 26.39° for the PRP and corticosteroid groups, respectively; P = .036).²¹

Another RCT (N = 99) by Kwong et al²² comparing PRP to corticosteroids found similar short-term advantages of LP-PRP with an improved VAS score (–13.6 vs 0.4; P = .03), American Shoulder and Elbow Surgeons score (13.0 vs 2.9; P = .02), and Western Ontario Rotator Cuff Index score (16.8 vs 5.8; P = .03).However, there was no long-term benefit of PRP over corticosteroids found at 12 months.²²

A 2021 systematic review and meta-­analysis by Hamid et al²³ that included 8 RCTs (N = 976) favored PRP over control (no injection, saline injections, and/or shoulder rehabilitation) with improved VAS scores at 12 months (SMD = –0.5; 95% CI, –0.7 to –0.2; P < .001).The evidence on functional outcome was mixed. Data pooled from 2 studies (n = 228) found better Shoulder Pain and Disability Index (SPADI) scores compared to controls at 3- and 6-month follow-ups. However, there were no significant differences in Disabilities of the Arm, Shoulder and Hand (DASH) scores between the 2 groups.²³

Patellar tendinopathy

❯ ❯ ❯   Consider using PRP for return to sport

Patellar tendinopathy, a common MSK condition encountered in the primary care setting, has an overall prevalence of 22% in elite athletes at some point in their career.^{24 ­}Nonsurgical management options include rest, ice, eccentric and isometric exercises, anti-­inflammatory drugs, extracorporeal shock wave therapy (ESWT), and dry needling (DN).

Currently, corticosteroid injections are the only intraarticular therapy recommended by international guidelines for hip OA.

A 2014 RCT (N = 23) evaluating DN vs PRP for patellar tendinopathy favored PRP with improved VAS scores (mean ± SD = 25.4 ± 23.2 points; P = .01 vs 5.2 ± 12.5 points; P = .20) at 12 weeks (P = .02). However, at ≥ 26 weeks, the improvement in pain and function scores was similar between the DN and PRP groups (33.2 ± 14.0 points; P = .001 vs 28.9 ± 25.2 points; P = .01). Notably, there was significantly more improvement in the PRP group at 12 weeks (P = .02) but not at 26 weeks (P = .66).²⁵

Continue to: Another perspective study...

Another prospective study (N = 31) comparing PRP to physiotherapy showed a greater improvement in sport activity level reflected by the Tegner score in the PRP group (percentage improvement, 39 ± 22%) compared to control (20 ± 27%; P = .048) at 6 months.⁷

A recent RCT (N = 20) revealed improved VAS scores at 6 months with rehabilitation paired with either bone marrow mesenchymal stem cells (BM-MSC) or LP-PRP when compared with baseline (BM-MSC group: 4.23 ± 2.13 to 2.52 ± 2.37; P = .0621; LP-PRP group: 3.10 ± 1.20 to 1.13 ± 1.25; P = .0083). Pain was significantly reduced during sport play in both groups at 6 months when compared with baseline (BM-MSC group: 6.91 ± 1.11 to 3.06 ± 2.89, P = .0049; PRP group: 7.03 ± 1.42 to 1.94 ± 1.24, P = .0001).²⁶

A 2019 systematic review and meta-analysis (N = 2530) demonstrated greater improvements in Victorian Institute of Sport Assessment scale for patellar tendinopathy (VISA-P) with multiple injections of PRP (38.7 points; 95% CI, 26.3-51.2 points) compared to single injections of PRP (24.3 points; 95% CI, 18.2-30.5 points), eccentric exercise (28.3 points; 95% CI, 18.9-37.8 points) and ESWT (27.4 points; 95% CI, 10.0-39.8 points) after 6 months.²⁷ In contrast, an RCT (n = 57) comparing a single injection of LR-PRP or LP-PRP was no more effective than a single injection of saline for improvement in mean VISA-P scores (P > .05) at 1 year.²⁸

Lateral epicondylitis

❯ ❯ ❯   Consider using PRP

Lateral epicondylitis (“tennis elbow”) is caused by overuse of the elbow extensors at the site of the lateral epicondyle. Chronic lateral epicondylosis involves tissue degeneration and microtrauma.Most cases of epicondylar tendinopathies are treated nonoperatively, with corticosteroid injections being a mainstay of treatment despite their short-term benefit²⁹ and potential to deteriorate connective tissue over time. Recent studies suggest PRP therapy for epicondylitis and epicondylosis may increase long-term pain relief and improve function.

The evidence on functional outcome of platelet-rich plasma for rotator cuff tendinopathy is mixed.

A 2017 systematic review and meta-­analysis of 16 RCTs (N = 1018) concluded PRP was more efficacious than control injections (bupivacaine) for pain reduction in tendinopathies (effect size = 0.47; 95% CI, 0.22-0.72).³⁰ In the review, lateral epicondylitis was evaluated in 12 studies and was most responsive to PRP (effect size = 0.57) when compared to control injection.³⁰ In another systematic review (5 RCTs; 250 patients), corticosteroid injections improved pain within the first 6 weeks of treatment. However, PRP outperformed corticosteroid in VAS scores (21.3 ± 28.1 vs 42.4 ± 26.8) and DASH scores (17.6 ± 24.0 vs 36.5 ± 23.8) (P < .001) at 2 years.³¹

Continue to: A 2022 systematic review...

A 2022 systematic review and meta-­analysis (26 studies; N = 1040) comparing scores at baseline vs 2 years post-PRP showed improvement in VAS scores (7.4 ± 1.30 vs 3.71 ± 2.35; P < .001), DASH scores (60.8 ± 12.5 vs 13.0 ± 18.5; P < .001), Patient-Rated Tennis Elbow Evaluation (55.6 ± 14.7 vs 48.8 ± 4.1; P < .001), and Mayo Clinic Performance Index (55.5 ± 6.1 vs 93.0 ± 6.7; P < .001).³²

Regarding the therapeutic effects of different PRP types in lateral epicondylitis, a 2022 systematic review of 33 studies (N = 2420) found improved function and pain relief with LR-PRP and LP-PRP with no significant differences.³³ Pretreatment VAS scores in the LR-PRP group, which ranged from 6.1 to 8.0, improved to 1.5 to 4.0 at 3 months and 0.6 to 3.3 after 1 year.³³ Similarly, pretreatment VAS scores in the LP-PRP group, which ranged from 4.2 to 8.4, improved to 1.6 to 5.9 at 3 months and 0.7 to 2.7 after 1 year.³⁴ DASH scores also improved in the LR-PRP and LP-PRP groups, with pretreatment scores (LR-PRP, 47.0 to 54.3; LP-PRP, 30.0 to 67.7) improving to 20.0 to 22.0 and 5.5 to 19.0, respectively, at 1 year.³³

Achilles tendinopathy

❯ ❯ ❯   Do not use PRP; evidence is lacking

Achilles tendinopathy, caused by chronic overuse and overload resulting in microtrauma and poor tissue healing, typically occurs in the most poorly vascularized portion of the tendon and is common in runners. First-line treatments for Achilles tendinopathy include eccentric strength training and anti-­inflammatory drugs.^34,35 Corticosteroid injections are not recommended, given concern for degraded tendon tissue over time and worse function.³⁴

A 2020 systematic review of 11 randomized and nonrandomized clinical trials (N = 406) found PRP improved Victorian Institute of Sports Assessment—Achilles (VISA-A) scores at 24 weeks compared to other nonsurgical treatment options (41.2 vs 70.12; P < .018).³⁴ However, a higher-quality 2021 systematic review and meta-analysis of 4 RCTs (N = 170) comparing PRP injections with placebo showed no significant difference in VISA-A scores at 3 months (0.23; 95% CI, –0.45 to 0.91), 6 months (0.83; 95% CI, –0.26 to 1.92), and 12 months (0.83; 95% CI, –0.77 to 2.44).³⁶ Therefore, further studies are warranted to evaluate the benefit of PRP injections for Achilles tendinopathy.

Conclusions

While high-quality studies support the use of PRP for knee OA and lateral epicondylitis, they have a moderate-to-high risk for bias. Several RCTs show that PRP provides superior short-term pain relief and range of motion compared to corticosteroids for rotator cuff tendinopathy. Multiple injections of PRP for patellar tendinopathy may accelerate return to sport and improve symptoms over the long term. However, current evidence does not support PRP therapy for Achilles tendinopathy. Given variability in PRP preparation, an accurate interpretation of the literature regarding its use in MSK conditions is recommended (TABLE^{4,6,7,14-18,20-23,25-28,30-34,36}).

Continue to: Concerning the effectiveness of PRP...

Concerning the effectiveness of PRP, it is important to consider early publication bias. Although recent studies have shown its benefits,^6,14,15,37 additional studies comparing PRP to placebo will help demonstrate its efficacy. Interestingly, a literature search by Bar-Or et al³⁸ found intra-articular saline may have a therapeutic effect on knee OA and confound findings when used as a placebo.

RCT data showed no benefit with platelet-rich plasma in treating small-joint osteoarthritis.

Recognizing the presence or lack of clinically significant improvement in the literature is important. For example, while some recent studies have shown PRP exceeds the minimal clinically significant difference for knee OA and lateral epicondylitis, others have not.^32,37 A 2021 systematic review of 11 clinical practice guidelines for the use of PRP in knee OA found that 9 were “uncertain or unable to make a recommendation” and 2 recommended against it.³⁹

In its 2021 position statement for the responsible use of regenerative medicine, the American Medical Society for Sports Medicine includes guidance on integrating orthobiologics into clinical practice. The guideline emphasizes informed consent and provides an evidence-based rationale for using PRP in certain patient populations (lateral epicondylitis and younger patients with mild-to-­moderate knee OA), recommending its use only after exhausting other conservative options.⁴⁰ Patients should be referred to physicians with experience using PRP and image-guided procedures.

CORRESPONDENCE
Gregory D. Bentz Jr, MD, 3640 High Street Suite 3B, Portsmouth, VA 23707; bentzgd@evms.edu

Approximately 1 in 4 people ages 18 years and older and 1 in 3 people ages 18 to 25 years had a mental illness in the past year, according to the 2021 National Survey of Drug Use and Health.¹ The survey also found that adults ages 18 to 25 years had the highest rate of serious mental illness but the lowest treatment rate compared to other adult age groups.¹ Unfortunately, more than 60% of patients receiving mental health treatment fail to benefit to a clinically meaningful degree.²

However, there is growing evidence that referring patients to behavioral health practitioners (BHPs) with outcome-measured skills that meet the patient’s specific needs can have a dramatic and positive impact. There are 2 main steps to pairing patients with an appropriate BHP: (1) use of measurement-based care data that can be analyzed at the patient and therapist level, and (2) data-driven referrals that pair patients with BHPs based on such routine outcome monitoring data (paired-on outcome data).

Psychotherapy’s slow road toward measurement-based care

Routine outcome monitoring is the systematic measurement of symptoms and functioning during treatment. It serves multiple functions, including program evaluation and benchmarking of patient improvement rates. Moreover, routine outcome monitoring–­derived feedback (based on repeated patient outcome measurements) can inform personalized and responsive care decisions throughout treatment.

For all intents and purposes, routine outcome monitoring plus feedback is synonymous with measurement-based care, which is becoming the preferred term in psychotherapy.³ Although measurement-based care is often the standard of practice for treating physical health conditions, the adoption of measurement-based care practices for treating mental health conditions remains low.³ The implementation of routine outcome monitoring is the lynchpin of measurement-based care, which in psychotherapy includes³:

• routinely administered symptom/functioning measure, ideally before each clinical encounter,
• practitioner review of these patient-level data,
• patient review of these data with their practitioner, and
• collaborative reevaluation of the person-specific treatment plan informed by these data.

CASE SCENARIO

Violeta W is a 33-year-old woman who presented to her family physician for her annual wellness exam. Prior to the exam, the medical assistant administered a Patient Health ­Questionnaire-9 (PHQ-9) to screen for depressive symptoms. Ms. W’s score was 20 out of 27, suggestive of depression. To further assess the severity of depressive symptoms and their effect on daily function, the physician reviewed responses to the questionnaire with her and discussed treatment options. Ms. W was most interested in trying a low-dose selective serotonin reuptake inhibitor (SSRI).

At her follow-up visit 4 weeks later, the medical assistant re-administered the PHQ-9. The physician then reviewed Ms. W’s responses­ with her and, based on Ms. W’s subjective report and objective symptoms (still a score of 20 out of 27 on the PHQ-9), increased her SSRI dose. At each subsequent visit, Ms. W completed a PHQ-9 and reviewed responses and depressive symptoms with her physician.

The value of measurement-based care in mental health care

A narrative review by Lewis et al³ of 21 randomized controlled clinical trials (RCTs) across a range of age groups (eg, adolescents, young adults, adults), disorders (eg, anxiety, mood), and settings (eg, outpatient, inpatient) found that in at least 9 review articles, measurement-based care was associated with significantly improved outcomes vs usual care (ie, treatment without routine outcome monitoring plus feedback). The average increase in treatment effect size was about 30% when treatment was accompanied by measurement-based care.³

Continue to: Moreover, a recent within-patient meta-analysis...

Routine outcome monitoring– derived feedback can inform personalized and responsive care decisions throughout treatment.

Moreover, a recent within-patient meta-analysis by de Jong et al⁴ shows that ­measurement-based care yields a small but significant increase in therapeutic outcomes (d = .15). Use of measurement-based care also is associated with improved communication between the patient and therapist.⁵ In pharmacotherapy practice, measurement-based care has been shown to predict rapid dose increases and changes in medication, when necessary; faster recovery rates; higher response rates to treatment³; and fewer ­dropouts.⁴

Perhaps one of the best-studied benefits of measurement-based mental health care is the ability to predict deterioration in care (ie, patients who are off-track in a way that practitioners often miss without the help of routine outcome monitoring data).^6,7 Studies show that without a data-informed approach to care, some forms of psychotherapy or therapy with BHPs who are not sufficiently skilled in treating a given diagnosis increase symptoms or create significant harmful and iatrogenic effects.^8-10 Conversely, the meta-analysis by de Jong et al⁴ found a lower percentage of deterioration in patients receiving ­measurement-based care. The difference in deterioration was significant: An average of 5.4% of patients in control conditions deteriorated compared to an average of 4.6% in feedback (measurement-based care) groups. There were even larger effect sizes when therapists received training in the feedback system.⁴

Routine outcome monitoring without a dialogue between patient and practitioner about the assessments (eg, ignoring complete measurement-based care requirements) may be inadequate. A recent review by Muir et al⁶ found no differences in patient outcomes when data were used solely for aggregate quality improvement activities, suggesting the need for practitioners to review results of routine outcome monitoring assessments with patients and use data to alter care when necessary.

Measurement-based care is believed to deliver benefits and reduce harm by enhancing and encouraging active patient involvement, improving patient understanding of symptoms, promoting better communication, and facilitating better care coordination.³ The benefits of measurement-based care can be enhanced with a comprehensive core routine outcome monitoring tool and the level of monitoring-generated information delivered for multiple stakeholders (eg, patient, therapist, clinic).¹¹

A look at multidimensional assessment

The features of routine outcome monitoring tools vary significantly.¹² Some measures assess single-symptom or problem domains (eg, PHQ-9 for depression or Generalized Anxiety Disorder-7 [GAD-7] scale for anxiety) or multiple dimensions (multidimensional routine outcome monitoring).Multidimensional routine outcome monitoring may have benefits over single-domain measures. Single­-domain measures and the subscales or factors of more comprehensive multidimensional routine outcome monitoring assessments should possess adequate specificity and sensitivity.

Continue to: Some recent research findings...

Some recent research findings question the construct validity of brief single-domain measures of common presenting problems, such as depression and anxiety. For example, results from a factor analysis of the PHQ-9 and GAD-7 scale in patients with traumatic brain injury suggest these tools measure 1 psychological construct that includes depression and the cognitive components of anxiety (eg, worry)¹³—a finding consistent with those of other tools.¹⁴ Similarly, a larger study of 7763 BH patients found that a single factor accounted for most of the variance of the 2 combined measures, with no set of factors meeting the exacting standards used to develop multidimensional routine outcome monitoring.¹⁵ These findings suggest that the PHQ-9 and GAD-7 largely overlap and are not measuring different aspects of health as most practitioners believe (eg, depression and anxiety).

In commonly used assessments, ­multiple-factor analytic studies with high standards have supported the construct validity of domain-specific subscales, indicating that the various questions tap into different constructs of psychological health.^14,16,17

Perhaps one of the best-studied benefits of mental health measurementbased care is the ability to predict deterioration in care.

Beyond multiple domain–specific indicators, multidimensional routine outcome measurements provide a global total score that minimizes Type I (false-positive conclusion) and Type II (false-negative conclusion) errors in tracking patient improvement or deterioration.¹⁸ As one would expect, multidimensional routine outcome monitoring generally includes more items than single-domain measures; however, this comes with a trade-off. If there are specificity and sensitivity concerns with an ultra-brief single-domain measure, an alternative to a core multidimensional routine outcome measurement is to aggregate a series of single-domain measures into a battery of patient self-reports. However, this approach may take longer for patients to complete since they would have to shift among the varying response sets and wording across the unique single-domain measures.

In addition, the standardization/­normalization of multidimensional routine outcome monitoring likely makes interpretation easier than referring to norms and clinical severity cutoffs for many distinct measures. Furthermore, increased specificity enhances predictive power and allows BHPs to screen and track other conditions besides depression and anxiety. (It is worth noting that there are no known studies that have looked at the difference in time to administer or ease of interpretation of multidimensional routine outcome monitoring tools vs multiple single-domain measures.)

Two multidimensional routine outcome monitoring tools that cover a comprehensive series of discrete symptom and functional domains are the Treatment Outcome Package¹² and Counseling Center ­Assessment of Psychological Symptoms.¹⁶ These tools, which include subscales beyond general depression and anxiety (eg, sleep, substance misuse, social conflict), take 7 to 10 minutes to complete and provide outcome results across 12 symptom and 8 functional dimensions. As an example, the Treatment Outcome Package has good psychometric qualities (eg, reliability, construct and concurrent validity) for adults,¹² children,^14,19 and adolescents,¹⁹ and can be administered through a secure online data collection portal. The Counseling Center Assessment of Psychological Symptoms has demonstrated high construct validity and good convergent validity.¹⁶ These assessments can be administered in paper or digital (eg, electronic medical record portal, smartphone) format.²⁰

Continue to: CASE SCENARIO

CASE SCENARIO

Ms. W’s physician asked her to go online using her phone and answer the questions in the Treatment Outcome Package. Her results, which she viewed with her physician, were displayed in graph form (FIGURE). Her scores were represented in Z scores normalized to the general population, with “0” representing the general, nontreatment-seeking population average and positive scores representing the number of standard deviations (SDs) more severe than the general population average.

Although this assessment scored Ms. W’s clinically elevated depression as mild, it revealed abnormalities in 3 other domains. Sexual functioning issues represented the most abnormal domain at greater than 3 SDs (more severe than the general population), followed by poor life quality and school/work functioning.

After reviewing Ms. W’s report, her physician decided that pharmacologic management alone (for depression) was not the most appropriate treatment course. Therefore, her physician recommended psychotherapy in addition to the SSRI she was taking. Ms. W agreed to a customized referral for psychotherapy.

Data-driven referrals

When psychotherapy is chosen as a treatment, the individual BHP is an active component of that treatment. Consequently, it is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains. It is rare for a BHP to be skilled in treating every mental health domain.⁹ Multiple studies have shown that BHPs have identifiable treatment skills in specific domains, which physicians should consider when making referrals.^9,21,22 These studies demonstrate the utility of aggregating patient-level routine outcome monitoring data to better understand therapist-level (and ultimately clinic- and system-level) outcomes.

It is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains.

Additionally, recent research has tested this idea prospectively. An RCT funded by the Patient-Centered Outcome Research Institute and published in JAMA Psychiatry showed a significant and positive effect on patient outcomes (ie, reductions in general impairment, impairment involving a patient’s most elevated domain, and global distress) using paired-on outcome data matching vs as-usual matching protocols (eg, therapist self-defined areas of specialty).²² In the RCT, the most effective matching protocol was a combination of eliminating harm and matching the patient on their 3 most problematic domains (the highest match level). These patients ended care as healthy as the general population after 16 weeks of treatment. A random 1-year follow-up assessment from the original RCT showed that most patients who had been matched had maintained their improvement.²³

Continue to: Therefore, a multidimensional routine outcome...

Therefore, a multidimensional routine outcome monitoring tool can be used to identify a BHP’s relative strengths and weaknesses across multiple outcome domains. Within a system of care, a sample of BHPs will possess varying outcome-domain profiles. When a new patient is seeking a referral to a BHP, these profiles (or domain-specific outcome track records) can be used to support paired-on outcome data matching. Specifically, a new patient completes the multidimensional routine outcome monitoring tool at pretreatment, and the results reveal the outcome domains on which the patient is most clinically severe. This pattern of domain-specific severity then can be used to pair the new patient with a BHP who has demonstrated success in addressing the same outcome domain(s). This approach matches a new patient to a BHP with established expertise based on routine outcome monitoring.

Retrospective and prospective studies have found that most BHPs have stable performance in their strengths and weaknesses.^11,21 One study found that assessing BHP performance with their most recent 30 patients can reliably predict future performance with their next 30 patients.²⁴ This predictability in a practitioner’s outcomes suggests report cards that are updated frequently can be utilized to make case assignments within BH or referrals to a specific BHP from primary care.

Making a paired-on outcome data–matched referral

Making customized BH referrals requires access to information about a practitioner’s previous routine outcome monitoring data per clinical domain (eg, suicidality, violence, quality of life) from their most recent patients. Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses (ie, effectiveness “report card”) per clinical ­domain.²⁴

Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses.

Few, if any, physicians have access to this level of updated outcome data from their referral network. To facilitate widespread use of paired-on outcome data matching, a new Web system (MatchedTherapists.com) will allow the general public and PCPs to access these grades. As a public service option, this site currently allows for a self-assessment using the Treatment Outcome Package. Pending versions will generate paired-on outcome data grades, and users will receive a list of local therapists available for in-person appointments as well as therapists available for virtual appointments. The paired-on outcome data grades are delivered in school-based letter grades. An “A+,” for example, represents the best matching grade. Users also will be able to sort and filter results for other criteria such as telemedicine, insurance, age, gender, and appointment availability. Currently, there are more than 77,000 therapists listed on the site nationwide. A basic listing is free.

CASE SCENARIO

After Ms. W took the multidimensional routine outcome assessment online, she received a list of therapists rank-ordered by paired-on outcome data grade, with the “A+” matches listed first. Three of the best-matched referrals accepted her insurance and were willing to see her through telemedicine. Therapists with available in-person appointments had a “B” grade. After discussing the options with her physician, Ms. W opted for telehealth counseling with the therapist whose profile she liked best. The therapist and PCP tracked her progress through routine outcome monitoring reporting until all her symptoms became subclinical.

Continue to: The future of a "referral bridge"

In this article, we present a solution to a common issue faced by mental health care patients: failure to benefit meaningfully from mental health treatment. Matching patients to specific BHPs based on effectiveness data regarding the therapist’s strengths and skills can improve patient outcomes and reduce harm. In addition, patients appear to value this approach. A Robert Wood Johnson Foundation–­funded study demonstrated that patients value seeing practitioners who have a track record of successfully treating previous patients with similar issues.^25,26 In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience or the same demographic characteristics as the patient.^25,26

In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience.

These findings may represent a new area in the science of health care. Over the past century, major advances in diagnosis and treatment—the 2 primary pillars of health care—have turned the art of medicine into a science. However, the art of making referrals has not advanced commensurately, as there has been little attention focused on the “referral bridge” between these 2 pillars. As the studies reviewed in this paper demonstrate, a referral bridge deserves exploration in all fields of medicine.

CORRESPONDENCE
David R. Kraus, PhD, 1 Speen Street, Framingham, MA 01701; dkraus@outcomereferrals.com

Approximately 1 in 4 people ages 18 years and older and 1 in 3 people ages 18 to 25 years had a mental illness in the past year, according to the 2021 National Survey of Drug Use and Health.¹ The survey also found that adults ages 18 to 25 years had the highest rate of serious mental illness but the lowest treatment rate compared to other adult age groups.¹ Unfortunately, more than 60% of patients receiving mental health treatment fail to benefit to a clinically meaningful degree.²

However, there is growing evidence that referring patients to behavioral health practitioners (BHPs) with outcome-measured skills that meet the patient’s specific needs can have a dramatic and positive impact. There are 2 main steps to pairing patients with an appropriate BHP: (1) use of measurement-based care data that can be analyzed at the patient and therapist level, and (2) data-driven referrals that pair patients with BHPs based on such routine outcome monitoring data (paired-on outcome data).

Psychotherapy’s slow road toward measurement-based care

Routine outcome monitoring is the systematic measurement of symptoms and functioning during treatment. It serves multiple functions, including program evaluation and benchmarking of patient improvement rates. Moreover, routine outcome monitoring–­derived feedback (based on repeated patient outcome measurements) can inform personalized and responsive care decisions throughout treatment.

For all intents and purposes, routine outcome monitoring plus feedback is synonymous with measurement-based care, which is becoming the preferred term in psychotherapy.³ Although measurement-based care is often the standard of practice for treating physical health conditions, the adoption of measurement-based care practices for treating mental health conditions remains low.³ The implementation of routine outcome monitoring is the lynchpin of measurement-based care, which in psychotherapy includes³:

• routinely administered symptom/functioning measure, ideally before each clinical encounter,
• practitioner review of these patient-level data,
• patient review of these data with their practitioner, and
• collaborative reevaluation of the person-specific treatment plan informed by these data.

CASE SCENARIO

Violeta W is a 33-year-old woman who presented to her family physician for her annual wellness exam. Prior to the exam, the medical assistant administered a Patient Health ­Questionnaire-9 (PHQ-9) to screen for depressive symptoms. Ms. W’s score was 20 out of 27, suggestive of depression. To further assess the severity of depressive symptoms and their effect on daily function, the physician reviewed responses to the questionnaire with her and discussed treatment options. Ms. W was most interested in trying a low-dose selective serotonin reuptake inhibitor (SSRI).

At her follow-up visit 4 weeks later, the medical assistant re-administered the PHQ-9. The physician then reviewed Ms. W’s responses­ with her and, based on Ms. W’s subjective report and objective symptoms (still a score of 20 out of 27 on the PHQ-9), increased her SSRI dose. At each subsequent visit, Ms. W completed a PHQ-9 and reviewed responses and depressive symptoms with her physician.

The value of measurement-based care in mental health care

A narrative review by Lewis et al³ of 21 randomized controlled clinical trials (RCTs) across a range of age groups (eg, adolescents, young adults, adults), disorders (eg, anxiety, mood), and settings (eg, outpatient, inpatient) found that in at least 9 review articles, measurement-based care was associated with significantly improved outcomes vs usual care (ie, treatment without routine outcome monitoring plus feedback). The average increase in treatment effect size was about 30% when treatment was accompanied by measurement-based care.³

Continue to: Moreover, a recent within-patient meta-analysis...

Routine outcome monitoring– derived feedback can inform personalized and responsive care decisions throughout treatment.

Moreover, a recent within-patient meta-analysis by de Jong et al⁴ shows that ­measurement-based care yields a small but significant increase in therapeutic outcomes (d = .15). Use of measurement-based care also is associated with improved communication between the patient and therapist.⁵ In pharmacotherapy practice, measurement-based care has been shown to predict rapid dose increases and changes in medication, when necessary; faster recovery rates; higher response rates to treatment³; and fewer ­dropouts.⁴

Perhaps one of the best-studied benefits of measurement-based mental health care is the ability to predict deterioration in care (ie, patients who are off-track in a way that practitioners often miss without the help of routine outcome monitoring data).^6,7 Studies show that without a data-informed approach to care, some forms of psychotherapy or therapy with BHPs who are not sufficiently skilled in treating a given diagnosis increase symptoms or create significant harmful and iatrogenic effects.^8-10 Conversely, the meta-analysis by de Jong et al⁴ found a lower percentage of deterioration in patients receiving ­measurement-based care. The difference in deterioration was significant: An average of 5.4% of patients in control conditions deteriorated compared to an average of 4.6% in feedback (measurement-based care) groups. There were even larger effect sizes when therapists received training in the feedback system.⁴

Routine outcome monitoring without a dialogue between patient and practitioner about the assessments (eg, ignoring complete measurement-based care requirements) may be inadequate. A recent review by Muir et al⁶ found no differences in patient outcomes when data were used solely for aggregate quality improvement activities, suggesting the need for practitioners to review results of routine outcome monitoring assessments with patients and use data to alter care when necessary.

Measurement-based care is believed to deliver benefits and reduce harm by enhancing and encouraging active patient involvement, improving patient understanding of symptoms, promoting better communication, and facilitating better care coordination.³ The benefits of measurement-based care can be enhanced with a comprehensive core routine outcome monitoring tool and the level of monitoring-generated information delivered for multiple stakeholders (eg, patient, therapist, clinic).¹¹

A look at multidimensional assessment

The features of routine outcome monitoring tools vary significantly.¹² Some measures assess single-symptom or problem domains (eg, PHQ-9 for depression or Generalized Anxiety Disorder-7 [GAD-7] scale for anxiety) or multiple dimensions (multidimensional routine outcome monitoring).Multidimensional routine outcome monitoring may have benefits over single-domain measures. Single­-domain measures and the subscales or factors of more comprehensive multidimensional routine outcome monitoring assessments should possess adequate specificity and sensitivity.

Continue to: Some recent research findings...

Some recent research findings question the construct validity of brief single-domain measures of common presenting problems, such as depression and anxiety. For example, results from a factor analysis of the PHQ-9 and GAD-7 scale in patients with traumatic brain injury suggest these tools measure 1 psychological construct that includes depression and the cognitive components of anxiety (eg, worry)¹³—a finding consistent with those of other tools.¹⁴ Similarly, a larger study of 7763 BH patients found that a single factor accounted for most of the variance of the 2 combined measures, with no set of factors meeting the exacting standards used to develop multidimensional routine outcome monitoring.¹⁵ These findings suggest that the PHQ-9 and GAD-7 largely overlap and are not measuring different aspects of health as most practitioners believe (eg, depression and anxiety).

In commonly used assessments, ­multiple-factor analytic studies with high standards have supported the construct validity of domain-specific subscales, indicating that the various questions tap into different constructs of psychological health.^14,16,17

Perhaps one of the best-studied benefits of mental health measurementbased care is the ability to predict deterioration in care.

Beyond multiple domain–specific indicators, multidimensional routine outcome measurements provide a global total score that minimizes Type I (false-positive conclusion) and Type II (false-negative conclusion) errors in tracking patient improvement or deterioration.¹⁸ As one would expect, multidimensional routine outcome monitoring generally includes more items than single-domain measures; however, this comes with a trade-off. If there are specificity and sensitivity concerns with an ultra-brief single-domain measure, an alternative to a core multidimensional routine outcome measurement is to aggregate a series of single-domain measures into a battery of patient self-reports. However, this approach may take longer for patients to complete since they would have to shift among the varying response sets and wording across the unique single-domain measures.

In addition, the standardization/­normalization of multidimensional routine outcome monitoring likely makes interpretation easier than referring to norms and clinical severity cutoffs for many distinct measures. Furthermore, increased specificity enhances predictive power and allows BHPs to screen and track other conditions besides depression and anxiety. (It is worth noting that there are no known studies that have looked at the difference in time to administer or ease of interpretation of multidimensional routine outcome monitoring tools vs multiple single-domain measures.)

Two multidimensional routine outcome monitoring tools that cover a comprehensive series of discrete symptom and functional domains are the Treatment Outcome Package¹² and Counseling Center ­Assessment of Psychological Symptoms.¹⁶ These tools, which include subscales beyond general depression and anxiety (eg, sleep, substance misuse, social conflict), take 7 to 10 minutes to complete and provide outcome results across 12 symptom and 8 functional dimensions. As an example, the Treatment Outcome Package has good psychometric qualities (eg, reliability, construct and concurrent validity) for adults,¹² children,^14,19 and adolescents,¹⁹ and can be administered through a secure online data collection portal. The Counseling Center Assessment of Psychological Symptoms has demonstrated high construct validity and good convergent validity.¹⁶ These assessments can be administered in paper or digital (eg, electronic medical record portal, smartphone) format.²⁰

Continue to: CASE SCENARIO

CASE SCENARIO

Ms. W’s physician asked her to go online using her phone and answer the questions in the Treatment Outcome Package. Her results, which she viewed with her physician, were displayed in graph form (FIGURE). Her scores were represented in Z scores normalized to the general population, with “0” representing the general, nontreatment-seeking population average and positive scores representing the number of standard deviations (SDs) more severe than the general population average.

Although this assessment scored Ms. W’s clinically elevated depression as mild, it revealed abnormalities in 3 other domains. Sexual functioning issues represented the most abnormal domain at greater than 3 SDs (more severe than the general population), followed by poor life quality and school/work functioning.

After reviewing Ms. W’s report, her physician decided that pharmacologic management alone (for depression) was not the most appropriate treatment course. Therefore, her physician recommended psychotherapy in addition to the SSRI she was taking. Ms. W agreed to a customized referral for psychotherapy.

Data-driven referrals

When psychotherapy is chosen as a treatment, the individual BHP is an active component of that treatment. Consequently, it is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains. It is rare for a BHP to be skilled in treating every mental health domain.⁹ Multiple studies have shown that BHPs have identifiable treatment skills in specific domains, which physicians should consider when making referrals.^9,21,22 These studies demonstrate the utility of aggregating patient-level routine outcome monitoring data to better understand therapist-level (and ultimately clinic- and system-level) outcomes.

It is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains.

Additionally, recent research has tested this idea prospectively. An RCT funded by the Patient-Centered Outcome Research Institute and published in JAMA Psychiatry showed a significant and positive effect on patient outcomes (ie, reductions in general impairment, impairment involving a patient’s most elevated domain, and global distress) using paired-on outcome data matching vs as-usual matching protocols (eg, therapist self-defined areas of specialty).²² In the RCT, the most effective matching protocol was a combination of eliminating harm and matching the patient on their 3 most problematic domains (the highest match level). These patients ended care as healthy as the general population after 16 weeks of treatment. A random 1-year follow-up assessment from the original RCT showed that most patients who had been matched had maintained their improvement.²³

Continue to: Therefore, a multidimensional routine outcome...

Therefore, a multidimensional routine outcome monitoring tool can be used to identify a BHP’s relative strengths and weaknesses across multiple outcome domains. Within a system of care, a sample of BHPs will possess varying outcome-domain profiles. When a new patient is seeking a referral to a BHP, these profiles (or domain-specific outcome track records) can be used to support paired-on outcome data matching. Specifically, a new patient completes the multidimensional routine outcome monitoring tool at pretreatment, and the results reveal the outcome domains on which the patient is most clinically severe. This pattern of domain-specific severity then can be used to pair the new patient with a BHP who has demonstrated success in addressing the same outcome domain(s). This approach matches a new patient to a BHP with established expertise based on routine outcome monitoring.

Retrospective and prospective studies have found that most BHPs have stable performance in their strengths and weaknesses.^11,21 One study found that assessing BHP performance with their most recent 30 patients can reliably predict future performance with their next 30 patients.²⁴ This predictability in a practitioner’s outcomes suggests report cards that are updated frequently can be utilized to make case assignments within BH or referrals to a specific BHP from primary care.

Making a paired-on outcome data–matched referral

Making customized BH referrals requires access to information about a practitioner’s previous routine outcome monitoring data per clinical domain (eg, suicidality, violence, quality of life) from their most recent patients. Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses (ie, effectiveness “report card”) per clinical ­domain.²⁴

Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses.

Few, if any, physicians have access to this level of updated outcome data from their referral network. To facilitate widespread use of paired-on outcome data matching, a new Web system (MatchedTherapists.com) will allow the general public and PCPs to access these grades. As a public service option, this site currently allows for a self-assessment using the Treatment Outcome Package. Pending versions will generate paired-on outcome data grades, and users will receive a list of local therapists available for in-person appointments as well as therapists available for virtual appointments. The paired-on outcome data grades are delivered in school-based letter grades. An “A+,” for example, represents the best matching grade. Users also will be able to sort and filter results for other criteria such as telemedicine, insurance, age, gender, and appointment availability. Currently, there are more than 77,000 therapists listed on the site nationwide. A basic listing is free.

CASE SCENARIO

After Ms. W took the multidimensional routine outcome assessment online, she received a list of therapists rank-ordered by paired-on outcome data grade, with the “A+” matches listed first. Three of the best-matched referrals accepted her insurance and were willing to see her through telemedicine. Therapists with available in-person appointments had a “B” grade. After discussing the options with her physician, Ms. W opted for telehealth counseling with the therapist whose profile she liked best. The therapist and PCP tracked her progress through routine outcome monitoring reporting until all her symptoms became subclinical.

Continue to: The future of a "referral bridge"

In this article, we present a solution to a common issue faced by mental health care patients: failure to benefit meaningfully from mental health treatment. Matching patients to specific BHPs based on effectiveness data regarding the therapist’s strengths and skills can improve patient outcomes and reduce harm. In addition, patients appear to value this approach. A Robert Wood Johnson Foundation–­funded study demonstrated that patients value seeing practitioners who have a track record of successfully treating previous patients with similar issues.^25,26 In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience or the same demographic characteristics as the patient.^25,26

In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience.

These findings may represent a new area in the science of health care. Over the past century, major advances in diagnosis and treatment—the 2 primary pillars of health care—have turned the art of medicine into a science. However, the art of making referrals has not advanced commensurately, as there has been little attention focused on the “referral bridge” between these 2 pillars. As the studies reviewed in this paper demonstrate, a referral bridge deserves exploration in all fields of medicine.

CORRESPONDENCE
David R. Kraus, PhD, 1 Speen Street, Framingham, MA 01701; dkraus@outcomereferrals.com

Approximately 1 in 4 people ages 18 years and older and 1 in 3 people ages 18 to 25 years had a mental illness in the past year, according to the 2021 National Survey of Drug Use and Health.¹ The survey also found that adults ages 18 to 25 years had the highest rate of serious mental illness but the lowest treatment rate compared to other adult age groups.¹ Unfortunately, more than 60% of patients receiving mental health treatment fail to benefit to a clinically meaningful degree.²

However, there is growing evidence that referring patients to behavioral health practitioners (BHPs) with outcome-measured skills that meet the patient’s specific needs can have a dramatic and positive impact. There are 2 main steps to pairing patients with an appropriate BHP: (1) use of measurement-based care data that can be analyzed at the patient and therapist level, and (2) data-driven referrals that pair patients with BHPs based on such routine outcome monitoring data (paired-on outcome data).

Psychotherapy’s slow road toward measurement-based care

Routine outcome monitoring is the systematic measurement of symptoms and functioning during treatment. It serves multiple functions, including program evaluation and benchmarking of patient improvement rates. Moreover, routine outcome monitoring–­derived feedback (based on repeated patient outcome measurements) can inform personalized and responsive care decisions throughout treatment.

For all intents and purposes, routine outcome monitoring plus feedback is synonymous with measurement-based care, which is becoming the preferred term in psychotherapy.³ Although measurement-based care is often the standard of practice for treating physical health conditions, the adoption of measurement-based care practices for treating mental health conditions remains low.³ The implementation of routine outcome monitoring is the lynchpin of measurement-based care, which in psychotherapy includes³:

• routinely administered symptom/functioning measure, ideally before each clinical encounter,
• practitioner review of these patient-level data,
• patient review of these data with their practitioner, and
• collaborative reevaluation of the person-specific treatment plan informed by these data.

CASE SCENARIO

Violeta W is a 33-year-old woman who presented to her family physician for her annual wellness exam. Prior to the exam, the medical assistant administered a Patient Health ­Questionnaire-9 (PHQ-9) to screen for depressive symptoms. Ms. W’s score was 20 out of 27, suggestive of depression. To further assess the severity of depressive symptoms and their effect on daily function, the physician reviewed responses to the questionnaire with her and discussed treatment options. Ms. W was most interested in trying a low-dose selective serotonin reuptake inhibitor (SSRI).

At her follow-up visit 4 weeks later, the medical assistant re-administered the PHQ-9. The physician then reviewed Ms. W’s responses­ with her and, based on Ms. W’s subjective report and objective symptoms (still a score of 20 out of 27 on the PHQ-9), increased her SSRI dose. At each subsequent visit, Ms. W completed a PHQ-9 and reviewed responses and depressive symptoms with her physician.

The value of measurement-based care in mental health care

A narrative review by Lewis et al³ of 21 randomized controlled clinical trials (RCTs) across a range of age groups (eg, adolescents, young adults, adults), disorders (eg, anxiety, mood), and settings (eg, outpatient, inpatient) found that in at least 9 review articles, measurement-based care was associated with significantly improved outcomes vs usual care (ie, treatment without routine outcome monitoring plus feedback). The average increase in treatment effect size was about 30% when treatment was accompanied by measurement-based care.³

Continue to: Moreover, a recent within-patient meta-analysis...

Routine outcome monitoring– derived feedback can inform personalized and responsive care decisions throughout treatment.

Moreover, a recent within-patient meta-analysis by de Jong et al⁴ shows that ­measurement-based care yields a small but significant increase in therapeutic outcomes (d = .15). Use of measurement-based care also is associated with improved communication between the patient and therapist.⁵ In pharmacotherapy practice, measurement-based care has been shown to predict rapid dose increases and changes in medication, when necessary; faster recovery rates; higher response rates to treatment³; and fewer ­dropouts.⁴

Perhaps one of the best-studied benefits of measurement-based mental health care is the ability to predict deterioration in care (ie, patients who are off-track in a way that practitioners often miss without the help of routine outcome monitoring data).^6,7 Studies show that without a data-informed approach to care, some forms of psychotherapy or therapy with BHPs who are not sufficiently skilled in treating a given diagnosis increase symptoms or create significant harmful and iatrogenic effects.^8-10 Conversely, the meta-analysis by de Jong et al⁴ found a lower percentage of deterioration in patients receiving ­measurement-based care. The difference in deterioration was significant: An average of 5.4% of patients in control conditions deteriorated compared to an average of 4.6% in feedback (measurement-based care) groups. There were even larger effect sizes when therapists received training in the feedback system.⁴

Routine outcome monitoring without a dialogue between patient and practitioner about the assessments (eg, ignoring complete measurement-based care requirements) may be inadequate. A recent review by Muir et al⁶ found no differences in patient outcomes when data were used solely for aggregate quality improvement activities, suggesting the need for practitioners to review results of routine outcome monitoring assessments with patients and use data to alter care when necessary.

Measurement-based care is believed to deliver benefits and reduce harm by enhancing and encouraging active patient involvement, improving patient understanding of symptoms, promoting better communication, and facilitating better care coordination.³ The benefits of measurement-based care can be enhanced with a comprehensive core routine outcome monitoring tool and the level of monitoring-generated information delivered for multiple stakeholders (eg, patient, therapist, clinic).¹¹

A look at multidimensional assessment

The features of routine outcome monitoring tools vary significantly.¹² Some measures assess single-symptom or problem domains (eg, PHQ-9 for depression or Generalized Anxiety Disorder-7 [GAD-7] scale for anxiety) or multiple dimensions (multidimensional routine outcome monitoring).Multidimensional routine outcome monitoring may have benefits over single-domain measures. Single­-domain measures and the subscales or factors of more comprehensive multidimensional routine outcome monitoring assessments should possess adequate specificity and sensitivity.

Continue to: Some recent research findings...

Some recent research findings question the construct validity of brief single-domain measures of common presenting problems, such as depression and anxiety. For example, results from a factor analysis of the PHQ-9 and GAD-7 scale in patients with traumatic brain injury suggest these tools measure 1 psychological construct that includes depression and the cognitive components of anxiety (eg, worry)¹³—a finding consistent with those of other tools.¹⁴ Similarly, a larger study of 7763 BH patients found that a single factor accounted for most of the variance of the 2 combined measures, with no set of factors meeting the exacting standards used to develop multidimensional routine outcome monitoring.¹⁵ These findings suggest that the PHQ-9 and GAD-7 largely overlap and are not measuring different aspects of health as most practitioners believe (eg, depression and anxiety).

In commonly used assessments, ­multiple-factor analytic studies with high standards have supported the construct validity of domain-specific subscales, indicating that the various questions tap into different constructs of psychological health.^14,16,17

Perhaps one of the best-studied benefits of mental health measurementbased care is the ability to predict deterioration in care.

Beyond multiple domain–specific indicators, multidimensional routine outcome measurements provide a global total score that minimizes Type I (false-positive conclusion) and Type II (false-negative conclusion) errors in tracking patient improvement or deterioration.¹⁸ As one would expect, multidimensional routine outcome monitoring generally includes more items than single-domain measures; however, this comes with a trade-off. If there are specificity and sensitivity concerns with an ultra-brief single-domain measure, an alternative to a core multidimensional routine outcome measurement is to aggregate a series of single-domain measures into a battery of patient self-reports. However, this approach may take longer for patients to complete since they would have to shift among the varying response sets and wording across the unique single-domain measures.

In addition, the standardization/­normalization of multidimensional routine outcome monitoring likely makes interpretation easier than referring to norms and clinical severity cutoffs for many distinct measures. Furthermore, increased specificity enhances predictive power and allows BHPs to screen and track other conditions besides depression and anxiety. (It is worth noting that there are no known studies that have looked at the difference in time to administer or ease of interpretation of multidimensional routine outcome monitoring tools vs multiple single-domain measures.)

Two multidimensional routine outcome monitoring tools that cover a comprehensive series of discrete symptom and functional domains are the Treatment Outcome Package¹² and Counseling Center ­Assessment of Psychological Symptoms.¹⁶ These tools, which include subscales beyond general depression and anxiety (eg, sleep, substance misuse, social conflict), take 7 to 10 minutes to complete and provide outcome results across 12 symptom and 8 functional dimensions. As an example, the Treatment Outcome Package has good psychometric qualities (eg, reliability, construct and concurrent validity) for adults,¹² children,^14,19 and adolescents,¹⁹ and can be administered through a secure online data collection portal. The Counseling Center Assessment of Psychological Symptoms has demonstrated high construct validity and good convergent validity.¹⁶ These assessments can be administered in paper or digital (eg, electronic medical record portal, smartphone) format.²⁰

Continue to: CASE SCENARIO

CASE SCENARIO

Ms. W’s physician asked her to go online using her phone and answer the questions in the Treatment Outcome Package. Her results, which she viewed with her physician, were displayed in graph form (FIGURE). Her scores were represented in Z scores normalized to the general population, with “0” representing the general, nontreatment-seeking population average and positive scores representing the number of standard deviations (SDs) more severe than the general population average.

Although this assessment scored Ms. W’s clinically elevated depression as mild, it revealed abnormalities in 3 other domains. Sexual functioning issues represented the most abnormal domain at greater than 3 SDs (more severe than the general population), followed by poor life quality and school/work functioning.

After reviewing Ms. W’s report, her physician decided that pharmacologic management alone (for depression) was not the most appropriate treatment course. Therefore, her physician recommended psychotherapy in addition to the SSRI she was taking. Ms. W agreed to a customized referral for psychotherapy.

Data-driven referrals

When psychotherapy is chosen as a treatment, the individual BHP is an active component of that treatment. Consequently, it is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains. It is rare for a BHP to be skilled in treating every mental health domain.⁹ Multiple studies have shown that BHPs have identifiable treatment skills in specific domains, which physicians should consider when making referrals.^9,21,22 These studies demonstrate the utility of aggregating patient-level routine outcome monitoring data to better understand therapist-level (and ultimately clinic- and system-level) outcomes.

It is essential to customize referrals to match a patient’s most elevated mental health concerns with a therapist who will most effectively treat those domains.

Additionally, recent research has tested this idea prospectively. An RCT funded by the Patient-Centered Outcome Research Institute and published in JAMA Psychiatry showed a significant and positive effect on patient outcomes (ie, reductions in general impairment, impairment involving a patient’s most elevated domain, and global distress) using paired-on outcome data matching vs as-usual matching protocols (eg, therapist self-defined areas of specialty).²² In the RCT, the most effective matching protocol was a combination of eliminating harm and matching the patient on their 3 most problematic domains (the highest match level). These patients ended care as healthy as the general population after 16 weeks of treatment. A random 1-year follow-up assessment from the original RCT showed that most patients who had been matched had maintained their improvement.²³

Continue to: Therefore, a multidimensional routine outcome...

Therefore, a multidimensional routine outcome monitoring tool can be used to identify a BHP’s relative strengths and weaknesses across multiple outcome domains. Within a system of care, a sample of BHPs will possess varying outcome-domain profiles. When a new patient is seeking a referral to a BHP, these profiles (or domain-specific outcome track records) can be used to support paired-on outcome data matching. Specifically, a new patient completes the multidimensional routine outcome monitoring tool at pretreatment, and the results reveal the outcome domains on which the patient is most clinically severe. This pattern of domain-specific severity then can be used to pair the new patient with a BHP who has demonstrated success in addressing the same outcome domain(s). This approach matches a new patient to a BHP with established expertise based on routine outcome monitoring.

Retrospective and prospective studies have found that most BHPs have stable performance in their strengths and weaknesses.^11,21 One study found that assessing BHP performance with their most recent 30 patients can reliably predict future performance with their next 30 patients.²⁴ This predictability in a practitioner’s outcomes suggests report cards that are updated frequently can be utilized to make case assignments within BH or referrals to a specific BHP from primary care.

Making a paired-on outcome data–matched referral

Making customized BH referrals requires access to information about a practitioner’s previous routine outcome monitoring data per clinical domain (eg, suicidality, violence, quality of life) from their most recent patients. Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses (ie, effectiveness “report card”) per clinical ­domain.²⁴

Previous research suggests that follow-up data from a minimum of 15 patients is necessary to make a reliable evaluation of a practitioner’s strengths and weaknesses.

Few, if any, physicians have access to this level of updated outcome data from their referral network. To facilitate widespread use of paired-on outcome data matching, a new Web system (MatchedTherapists.com) will allow the general public and PCPs to access these grades. As a public service option, this site currently allows for a self-assessment using the Treatment Outcome Package. Pending versions will generate paired-on outcome data grades, and users will receive a list of local therapists available for in-person appointments as well as therapists available for virtual appointments. The paired-on outcome data grades are delivered in school-based letter grades. An “A+,” for example, represents the best matching grade. Users also will be able to sort and filter results for other criteria such as telemedicine, insurance, age, gender, and appointment availability. Currently, there are more than 77,000 therapists listed on the site nationwide. A basic listing is free.

CASE SCENARIO

After Ms. W took the multidimensional routine outcome assessment online, she received a list of therapists rank-ordered by paired-on outcome data grade, with the “A+” matches listed first. Three of the best-matched referrals accepted her insurance and were willing to see her through telemedicine. Therapists with available in-person appointments had a “B” grade. After discussing the options with her physician, Ms. W opted for telehealth counseling with the therapist whose profile she liked best. The therapist and PCP tracked her progress through routine outcome monitoring reporting until all her symptoms became subclinical.

Continue to: The future of a "referral bridge"

In this article, we present a solution to a common issue faced by mental health care patients: failure to benefit meaningfully from mental health treatment. Matching patients to specific BHPs based on effectiveness data regarding the therapist’s strengths and skills can improve patient outcomes and reduce harm. In addition, patients appear to value this approach. A Robert Wood Johnson Foundation–­funded study demonstrated that patients value seeing practitioners who have a track record of successfully treating previous patients with similar issues.^25,26 In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience or the same demographic characteristics as the patient.^25,26

In many cases, patients indicated they would prioritize this matching process over other factors such as practitioners with a higher number of years of experience.

These findings may represent a new area in the science of health care. Over the past century, major advances in diagnosis and treatment—the 2 primary pillars of health care—have turned the art of medicine into a science. However, the art of making referrals has not advanced commensurately, as there has been little attention focused on the “referral bridge” between these 2 pillars. As the studies reviewed in this paper demonstrate, a referral bridge deserves exploration in all fields of medicine.

CORRESPONDENCE
David R. Kraus, PhD, 1 Speen Street, Framingham, MA 01701; dkraus@outcomereferrals.com

Subclinical hypothyroidism (SCH) is a biochemical state in which the thyroid-stimulating hormone (TSH) is elevated while the free thyroxine (T4) level is normal. Overt hypothyroidism is not diagnosed until the free T4 level is decreased, regardless of the degree of TSH elevation.

The overall prevalence of SCH in iodine-rich areas is 4% to 10%, with a risk for progression to overt hypothyroidism of between 2% and 6% annually.¹ The prevalence of SCH varies depending on the TSH reference range used.¹ The normal reference range for TSH varies depending on the laboratory and/or the reference population surveyed, with the range likely widening with increasing age.

SCH is most common among women, the elderly, and White individuals.² The discovery of SCH is often incidental, given that usually it is detected by laboratory findings alone without associated symptoms of overt hypothyroidism.³

The not-so-significant role of symptoms in subclinical hypothyroidism

Symptoms associated with overt hypothyroidism include constipation, dry skin, fatigue, slow thinking, poor memory, muscle cramps, weakness, and cold intolerance. In SCH, these symptoms are inconsistent, with around 1 in 3 patients having no symptoms at all.⁴

One study reported that roughly 18% of euthyroid individuals, 22% of SCH patients, and 26% of those with overt hypothyroidism reported 4 or more symptoms classically thought to be related to hypothyroidism.⁴ A large Danish cohort study found that hypothyroid symptoms were no more common in patients with SCH than in euthyroid individuals in the general population.⁵ These studies question the validity of attributing symptoms to SCH.

Adverse health associations

Observational data suggest that SCH is associated with an increased risk for dyslipidemia, coronary heart disease, heart failure, and cardiovascular mortality, particularly in those with TSH levels ≥ 10 mIU/L.^6,7 Such associations were not found for most adults with TSH levels between 5 and 10 mIU/L.⁸ There are also potential associations of SCH with obesity, nonalcoholic fatty liver disease, and nonalcoholic steatohepatitis.^9,10 Despite thyroid studies being commonly ordered as part of a mental health evaluation, SCH has not been statistically associated with depressive symptoms.^11,12

Caveats with laboratory testing

There are several issues to consider when performing a laboratory assessment of thyroid function. TSH levels fluctuate considerably during the day, as TSH secretion has a circadian rhythm. TSH values are 50% higher at night and in the early morning than during the rest of the day.¹³ TSH values also may rise in response to current illness or stress. Due to this biologic variability, repeat testing to confirm TSH levels is recommended if an initial test result is abnormal.¹⁴

Continue to: An exact reference range...

Supplements containing biotin should be withheld for several days before assessing thyroid function.

An exact reference range for TSH is not widely agreed upon—although most laboratories regard 4.0 to 5.0 mIU/L as the high-end cutoff for normal. Also, “normal” TSH levels appear to differ by age. Accordingly, some experts have recommended an age-based reference range for TSH levels,¹⁵ although this is not implemented widely by laboratories. A TSH level of 6.0 mIU/L (or even higher) may be more appropriate for adults older than 65 years.¹

Biotin supplementation has been shown to cause spurious thyroid testing results (TSH, T3, T4) depending on the type of assay used. Therefore, supplements containing biotin should be withheld for several days before assessing thyroid function.¹⁶Patients with SCH are often categorized as having TSH levels between 4.5 and 10 mIU/L (around 90% of patients) or levels ≥ 10 mIU/L.^8,17 If followed for 5 years, approximately 60% of patients with SCH and TSH levels between 4 and 10 mIU/L will normalize without intervention.¹⁸ Normalization is less common in patients with a TSH level greater than 10 mIU/L.¹⁸

The risk for progression to overt hypothyroidism also appears to be higher for those with certain risk factors. These include higher baseline TSH levels, presence of thyroid peroxidase antibodies (TPOAbs), or history of neck irradiation or radioactive iodine uptake.¹ Other risk factors for eventual thyroid dysfunction include female sex, older age, goiter, and high iodine intake.¹³

Evidence for treatment varies

Guidelines for the treatment of SCH (TABLE 1^8,14,19,20) are founded on the condition’s risk for progression to overt hypothyroidism and its association with health consequences such as cardiovascular disease. Guidelines of the American Thyroid Association (ATA) and European Thyroid Association (ETA), and those of the United Kingdom–based National Institute for Health and Care Excellence (NICE), prioritize treatment for individuals with a TSH level > 10 mIU/L^a and for those with TSH values < 10 mIU/L but still elevated and apparent symptoms of hypothyroidism.^14,19,20 The strength of evidence behind this guidance is challenged by a lack of data from prospective randomized controlled trials (RCTs) demonstrating health benefits following treatment of SCH. The British Medical Journal (BMJ) Guideline cites this lack of evidence and recommends against treating SCH at any TSH level, regardless of symptoms.⁸

There are few large RCTs of treatment outcomes for SCH. A 2017 RCT (the Thyroid Hormone Replacement for Untreated Older Adults with Subclinical Hypothyroidism, or TRUST, trial) of 737 adults older than 65 years with SCH evaluated the ability of levothyroxine to normalize TSH values compared with placebo. At 1 year, there was no difference in hypothyroid symptoms or tiredness scale scores with levothyroxine treatment compared with placebo.²¹ This finding was consistent even in the subgroup with a higher baseline symptom burden.²²

Continue to: Two small RCTs evaluated...

Two small RCTs evaluated treatment of SCH with depressive symptoms and cognitive function, neither finding benefit compared with placebo.^12,23 A 2018 systematic review and meta-analysis of 21 studies and 2192 adults did not show a benefit to quality of life or thyroid-specific symptoms in those treated for SCH compared with controls.²⁴

RCT support also is lacking for a reduction in cardiovascular mortality following treatment for SCH. A large population-level retrospective cohort from Denmark showed no difference in cardiovascular mortality or myocardial infarction in those treated for SCH compared with controls.²⁵ Pooled results from 2 RCTs (for patients older than 65 years, and those older than 80 years) showed no change in risk for cardiovascular outcomes in older adults treated for SCH.²⁶ Older adults treated for SCH in the TRUST trial showed no improvements in systolic or diastolic­ function on echocardiography.^27­ Two trials showed no difference in carotid intima-media thickness with treatment of SCH compared with placebo.^28,29

While most of the RCT data come from older adults, a retrospective cohort study in the United Kingdom of younger (ages 40-70 years; n = 3093) and older (age > 70 years; n = 1642) patients showed a reduction in cardiovascular mortality among treated patients who were younger (hazard ratio [HR] = 0.61; 4.2% vs. 6.6%) but not those who were older (HR = 0.99; 12.7% vs. 10.7%).³⁰ There is also evidence that thyroid size in those with goiter can be reduced with treatment of SCH.³¹

A measured approach to treating subclinical hypothyroidism

Consider several factors when deciding whether to treat SCH. For instance, RCT data suggest a lack of treatment benefit in relieving depression, improving cognition, or reducing general hypothyroid symptoms. Treatment of SCH in older adults does not appear to improve cardiovascular outcomes. The question of whether long-term treatment of SCH in younger patients reduces cardiovascular morbidity or mortality lacks answers from RCTs. Before diagnosing SCH or starting treatment, always confirm SCH with repeat testing in 2 to 3 months, as a high percentage of those with untreated SCH will have normal thyroid function on repeat testing.

Before diagnosing subclinical hypothyroidism (SCH) or starting treatment, always confirm SCH with repeat testing in 2 to 3 months.

In the event you and your patient elect to treat SCH, guidelines and trials generally support a low initial daily dose of 25 to 50 mcg of levothyroxine (T4), followed with dose changes­ every 4 to 8 weeks and a goal of normalizing TSH to within the lower half of the reference range (0.4-2.5 mIU/L).¹⁴ This is generally similar to published treatment goals for primary hypothyroidism and is based on studies suggesting the lower half of the reference range is normal for young, healthy, euthyroid individuals.³² Though full replacement doses (1.6-1.8 mcg/kg of ideal body weight) can be started for those who are elderly or who have ischemic heart disease or angina, this approach should be avoided in favor of low-dose initial therapy.³³ Thyroid supplements are best absorbed when taken apart from food, calcium, or iron supplements. The ATA suggests taking thyroid medication 60 minutes before breakfast or at bedtime (3 or more hours after the evening meal).³³

Continue to: Screening guidelines differ

Lacking population-level screening data from RCTs, most organizations do not recommend screening for thyroid dysfunction or they note insufficient evidence to make a screening recommendation (TABLE 2^17,19,20,34). In their most recent recommendation statement on the subject in 2015, the US Preventive Services Task Force (USPSTF) concluded the current evidence was insufficient to recommend for or against thyroid dysfunction screening in nonpregnant, asymptomatic adults.¹⁷ This differs from the ATA and the American Association of Clinical Endocrinology (AACE; formerly known as the American Association of Clinical Endocrinologists), which both recommend targeted screening for thyroid dysfunction based on symptoms or risk factors.²⁰

What about subclinical hypothyroidism in pregnancy?

Overt hypothyroidism is associated with adverse events during pregnancy and with subsequent neurodevelopmental complications in children, although the effects of SCH during pregnancy remain less certain. Concerns have been raised over the potential association of SCH with pregnancy loss, placental abruption, premature rupture of membranes, and neonatal death.³⁵ Historically, the prevalence of SCH during pregnancy has ranged from 2% to 2.5%, but using lower trimester-based TSH reference ranges, the prevalence of SCH in pregnancy may be as high as 15%.³⁵

Guided by a large RCT that failed to find benefit (pregnancy outcomes, neurodevelopmental outcomes in children) following treatment of SCH in pregnancy,³⁶ the American College of Obstetricians and Gynecologists (ACOG) recommends against routine screening for thyroid disease in pregnancy.³⁴ The ATA notes insufficient evidence to rec-ommend universal screening for thyroid dysfunction in pregnancy but recommends targeted screening of those with risk factors.³⁷ Data are conflicting on the benefit of treating known or recently detected SCH on pregnancy outcomes including pregnancy loss.^35,38 As such, the American Society of Reproductive Medicine and the ATA both generally recommend treatment of SCH in pregnant patients, particularly when the TSH is ≥ 4.0 mIU/L and TPOAbs are present.^37,39

^a The ATA, ETA, and NICE have slightly different recommendations when a TSH level = 10 mIU/L. ETA and NICE recommend prioritizing treatment for individuals with this level, while ATA recommends treatment when individual factors are also considered.

ACKNOWLEDGEMENT
The authors thank Family Medicine Medical Librarian Gwen Wilson, MLS, AHIP, for her assistance with literature searches.

CORRESPONDENCE
Nicholas LeFevre, MD, Family and Community Medicine, University of Missouri–Columbia School of Medicine, One Hospital Drive, M224 Medical Science Building, Columbia, MO 65212; nlefevre@health.missouri.edu

Subclinical hypothyroidism (SCH) is a biochemical state in which the thyroid-stimulating hormone (TSH) is elevated while the free thyroxine (T4) level is normal. Overt hypothyroidism is not diagnosed until the free T4 level is decreased, regardless of the degree of TSH elevation.

The overall prevalence of SCH in iodine-rich areas is 4% to 10%, with a risk for progression to overt hypothyroidism of between 2% and 6% annually.¹ The prevalence of SCH varies depending on the TSH reference range used.¹ The normal reference range for TSH varies depending on the laboratory and/or the reference population surveyed, with the range likely widening with increasing age.

SCH is most common among women, the elderly, and White individuals.² The discovery of SCH is often incidental, given that usually it is detected by laboratory findings alone without associated symptoms of overt hypothyroidism.³

The not-so-significant role of symptoms in subclinical hypothyroidism

Symptoms associated with overt hypothyroidism include constipation, dry skin, fatigue, slow thinking, poor memory, muscle cramps, weakness, and cold intolerance. In SCH, these symptoms are inconsistent, with around 1 in 3 patients having no symptoms at all.⁴

One study reported that roughly 18% of euthyroid individuals, 22% of SCH patients, and 26% of those with overt hypothyroidism reported 4 or more symptoms classically thought to be related to hypothyroidism.⁴ A large Danish cohort study found that hypothyroid symptoms were no more common in patients with SCH than in euthyroid individuals in the general population.⁵ These studies question the validity of attributing symptoms to SCH.

Adverse health associations

Observational data suggest that SCH is associated with an increased risk for dyslipidemia, coronary heart disease, heart failure, and cardiovascular mortality, particularly in those with TSH levels ≥ 10 mIU/L.^6,7 Such associations were not found for most adults with TSH levels between 5 and 10 mIU/L.⁸ There are also potential associations of SCH with obesity, nonalcoholic fatty liver disease, and nonalcoholic steatohepatitis.^9,10 Despite thyroid studies being commonly ordered as part of a mental health evaluation, SCH has not been statistically associated with depressive symptoms.^11,12

Caveats with laboratory testing

There are several issues to consider when performing a laboratory assessment of thyroid function. TSH levels fluctuate considerably during the day, as TSH secretion has a circadian rhythm. TSH values are 50% higher at night and in the early morning than during the rest of the day.¹³ TSH values also may rise in response to current illness or stress. Due to this biologic variability, repeat testing to confirm TSH levels is recommended if an initial test result is abnormal.¹⁴

Continue to: An exact reference range...

Supplements containing biotin should be withheld for several days before assessing thyroid function.

An exact reference range for TSH is not widely agreed upon—although most laboratories regard 4.0 to 5.0 mIU/L as the high-end cutoff for normal. Also, “normal” TSH levels appear to differ by age. Accordingly, some experts have recommended an age-based reference range for TSH levels,¹⁵ although this is not implemented widely by laboratories. A TSH level of 6.0 mIU/L (or even higher) may be more appropriate for adults older than 65 years.¹

Biotin supplementation has been shown to cause spurious thyroid testing results (TSH, T3, T4) depending on the type of assay used. Therefore, supplements containing biotin should be withheld for several days before assessing thyroid function.¹⁶Patients with SCH are often categorized as having TSH levels between 4.5 and 10 mIU/L (around 90% of patients) or levels ≥ 10 mIU/L.^8,17 If followed for 5 years, approximately 60% of patients with SCH and TSH levels between 4 and 10 mIU/L will normalize without intervention.¹⁸ Normalization is less common in patients with a TSH level greater than 10 mIU/L.¹⁸

The risk for progression to overt hypothyroidism also appears to be higher for those with certain risk factors. These include higher baseline TSH levels, presence of thyroid peroxidase antibodies (TPOAbs), or history of neck irradiation or radioactive iodine uptake.¹ Other risk factors for eventual thyroid dysfunction include female sex, older age, goiter, and high iodine intake.¹³

Evidence for treatment varies

Guidelines for the treatment of SCH (TABLE 1^8,14,19,20) are founded on the condition’s risk for progression to overt hypothyroidism and its association with health consequences such as cardiovascular disease. Guidelines of the American Thyroid Association (ATA) and European Thyroid Association (ETA), and those of the United Kingdom–based National Institute for Health and Care Excellence (NICE), prioritize treatment for individuals with a TSH level > 10 mIU/L^a and for those with TSH values < 10 mIU/L but still elevated and apparent symptoms of hypothyroidism.^14,19,20 The strength of evidence behind this guidance is challenged by a lack of data from prospective randomized controlled trials (RCTs) demonstrating health benefits following treatment of SCH. The British Medical Journal (BMJ) Guideline cites this lack of evidence and recommends against treating SCH at any TSH level, regardless of symptoms.⁸

There are few large RCTs of treatment outcomes for SCH. A 2017 RCT (the Thyroid Hormone Replacement for Untreated Older Adults with Subclinical Hypothyroidism, or TRUST, trial) of 737 adults older than 65 years with SCH evaluated the ability of levothyroxine to normalize TSH values compared with placebo. At 1 year, there was no difference in hypothyroid symptoms or tiredness scale scores with levothyroxine treatment compared with placebo.²¹ This finding was consistent even in the subgroup with a higher baseline symptom burden.²²

Continue to: Two small RCTs evaluated...

Two small RCTs evaluated treatment of SCH with depressive symptoms and cognitive function, neither finding benefit compared with placebo.^12,23 A 2018 systematic review and meta-analysis of 21 studies and 2192 adults did not show a benefit to quality of life or thyroid-specific symptoms in those treated for SCH compared with controls.²⁴

RCT support also is lacking for a reduction in cardiovascular mortality following treatment for SCH. A large population-level retrospective cohort from Denmark showed no difference in cardiovascular mortality or myocardial infarction in those treated for SCH compared with controls.²⁵ Pooled results from 2 RCTs (for patients older than 65 years, and those older than 80 years) showed no change in risk for cardiovascular outcomes in older adults treated for SCH.²⁶ Older adults treated for SCH in the TRUST trial showed no improvements in systolic or diastolic­ function on echocardiography.^27­ Two trials showed no difference in carotid intima-media thickness with treatment of SCH compared with placebo.^28,29

While most of the RCT data come from older adults, a retrospective cohort study in the United Kingdom of younger (ages 40-70 years; n = 3093) and older (age > 70 years; n = 1642) patients showed a reduction in cardiovascular mortality among treated patients who were younger (hazard ratio [HR] = 0.61; 4.2% vs. 6.6%) but not those who were older (HR = 0.99; 12.7% vs. 10.7%).³⁰ There is also evidence that thyroid size in those with goiter can be reduced with treatment of SCH.³¹

A measured approach to treating subclinical hypothyroidism

Consider several factors when deciding whether to treat SCH. For instance, RCT data suggest a lack of treatment benefit in relieving depression, improving cognition, or reducing general hypothyroid symptoms. Treatment of SCH in older adults does not appear to improve cardiovascular outcomes. The question of whether long-term treatment of SCH in younger patients reduces cardiovascular morbidity or mortality lacks answers from RCTs. Before diagnosing SCH or starting treatment, always confirm SCH with repeat testing in 2 to 3 months, as a high percentage of those with untreated SCH will have normal thyroid function on repeat testing.

Before diagnosing subclinical hypothyroidism (SCH) or starting treatment, always confirm SCH with repeat testing in 2 to 3 months.

In the event you and your patient elect to treat SCH, guidelines and trials generally support a low initial daily dose of 25 to 50 mcg of levothyroxine (T4), followed with dose changes­ every 4 to 8 weeks and a goal of normalizing TSH to within the lower half of the reference range (0.4-2.5 mIU/L).¹⁴ This is generally similar to published treatment goals for primary hypothyroidism and is based on studies suggesting the lower half of the reference range is normal for young, healthy, euthyroid individuals.³² Though full replacement doses (1.6-1.8 mcg/kg of ideal body weight) can be started for those who are elderly or who have ischemic heart disease or angina, this approach should be avoided in favor of low-dose initial therapy.³³ Thyroid supplements are best absorbed when taken apart from food, calcium, or iron supplements. The ATA suggests taking thyroid medication 60 minutes before breakfast or at bedtime (3 or more hours after the evening meal).³³

Continue to: Screening guidelines differ

Lacking population-level screening data from RCTs, most organizations do not recommend screening for thyroid dysfunction or they note insufficient evidence to make a screening recommendation (TABLE 2^17,19,20,34). In their most recent recommendation statement on the subject in 2015, the US Preventive Services Task Force (USPSTF) concluded the current evidence was insufficient to recommend for or against thyroid dysfunction screening in nonpregnant, asymptomatic adults.¹⁷ This differs from the ATA and the American Association of Clinical Endocrinology (AACE; formerly known as the American Association of Clinical Endocrinologists), which both recommend targeted screening for thyroid dysfunction based on symptoms or risk factors.²⁰

What about subclinical hypothyroidism in pregnancy?

Overt hypothyroidism is associated with adverse events during pregnancy and with subsequent neurodevelopmental complications in children, although the effects of SCH during pregnancy remain less certain. Concerns have been raised over the potential association of SCH with pregnancy loss, placental abruption, premature rupture of membranes, and neonatal death.³⁵ Historically, the prevalence of SCH during pregnancy has ranged from 2% to 2.5%, but using lower trimester-based TSH reference ranges, the prevalence of SCH in pregnancy may be as high as 15%.³⁵

Guided by a large RCT that failed to find benefit (pregnancy outcomes, neurodevelopmental outcomes in children) following treatment of SCH in pregnancy,³⁶ the American College of Obstetricians and Gynecologists (ACOG) recommends against routine screening for thyroid disease in pregnancy.³⁴ The ATA notes insufficient evidence to rec-ommend universal screening for thyroid dysfunction in pregnancy but recommends targeted screening of those with risk factors.³⁷ Data are conflicting on the benefit of treating known or recently detected SCH on pregnancy outcomes including pregnancy loss.^35,38 As such, the American Society of Reproductive Medicine and the ATA both generally recommend treatment of SCH in pregnant patients, particularly when the TSH is ≥ 4.0 mIU/L and TPOAbs are present.^37,39

^a The ATA, ETA, and NICE have slightly different recommendations when a TSH level = 10 mIU/L. ETA and NICE recommend prioritizing treatment for individuals with this level, while ATA recommends treatment when individual factors are also considered.

ACKNOWLEDGEMENT
The authors thank Family Medicine Medical Librarian Gwen Wilson, MLS, AHIP, for her assistance with literature searches.

CORRESPONDENCE
Nicholas LeFevre, MD, Family and Community Medicine, University of Missouri–Columbia School of Medicine, One Hospital Drive, M224 Medical Science Building, Columbia, MO 65212; nlefevre@health.missouri.edu

Subclinical hypothyroidism (SCH) is a biochemical state in which the thyroid-stimulating hormone (TSH) is elevated while the free thyroxine (T4) level is normal. Overt hypothyroidism is not diagnosed until the free T4 level is decreased, regardless of the degree of TSH elevation.

The overall prevalence of SCH in iodine-rich areas is 4% to 10%, with a risk for progression to overt hypothyroidism of between 2% and 6% annually.¹ The prevalence of SCH varies depending on the TSH reference range used.¹ The normal reference range for TSH varies depending on the laboratory and/or the reference population surveyed, with the range likely widening with increasing age.

SCH is most common among women, the elderly, and White individuals.² The discovery of SCH is often incidental, given that usually it is detected by laboratory findings alone without associated symptoms of overt hypothyroidism.³

The not-so-significant role of symptoms in subclinical hypothyroidism

Symptoms associated with overt hypothyroidism include constipation, dry skin, fatigue, slow thinking, poor memory, muscle cramps, weakness, and cold intolerance. In SCH, these symptoms are inconsistent, with around 1 in 3 patients having no symptoms at all.⁴

One study reported that roughly 18% of euthyroid individuals, 22% of SCH patients, and 26% of those with overt hypothyroidism reported 4 or more symptoms classically thought to be related to hypothyroidism.⁴ A large Danish cohort study found that hypothyroid symptoms were no more common in patients with SCH than in euthyroid individuals in the general population.⁵ These studies question the validity of attributing symptoms to SCH.

Adverse health associations

Observational data suggest that SCH is associated with an increased risk for dyslipidemia, coronary heart disease, heart failure, and cardiovascular mortality, particularly in those with TSH levels ≥ 10 mIU/L.^6,7 Such associations were not found for most adults with TSH levels between 5 and 10 mIU/L.⁸ There are also potential associations of SCH with obesity, nonalcoholic fatty liver disease, and nonalcoholic steatohepatitis.^9,10 Despite thyroid studies being commonly ordered as part of a mental health evaluation, SCH has not been statistically associated with depressive symptoms.^11,12

Caveats with laboratory testing

There are several issues to consider when performing a laboratory assessment of thyroid function. TSH levels fluctuate considerably during the day, as TSH secretion has a circadian rhythm. TSH values are 50% higher at night and in the early morning than during the rest of the day.¹³ TSH values also may rise in response to current illness or stress. Due to this biologic variability, repeat testing to confirm TSH levels is recommended if an initial test result is abnormal.¹⁴

Continue to: An exact reference range...

Supplements containing biotin should be withheld for several days before assessing thyroid function.

An exact reference range for TSH is not widely agreed upon—although most laboratories regard 4.0 to 5.0 mIU/L as the high-end cutoff for normal. Also, “normal” TSH levels appear to differ by age. Accordingly, some experts have recommended an age-based reference range for TSH levels,¹⁵ although this is not implemented widely by laboratories. A TSH level of 6.0 mIU/L (or even higher) may be more appropriate for adults older than 65 years.¹

Biotin supplementation has been shown to cause spurious thyroid testing results (TSH, T3, T4) depending on the type of assay used. Therefore, supplements containing biotin should be withheld for several days before assessing thyroid function.¹⁶Patients with SCH are often categorized as having TSH levels between 4.5 and 10 mIU/L (around 90% of patients) or levels ≥ 10 mIU/L.^8,17 If followed for 5 years, approximately 60% of patients with SCH and TSH levels between 4 and 10 mIU/L will normalize without intervention.¹⁸ Normalization is less common in patients with a TSH level greater than 10 mIU/L.¹⁸

The risk for progression to overt hypothyroidism also appears to be higher for those with certain risk factors. These include higher baseline TSH levels, presence of thyroid peroxidase antibodies (TPOAbs), or history of neck irradiation or radioactive iodine uptake.¹ Other risk factors for eventual thyroid dysfunction include female sex, older age, goiter, and high iodine intake.¹³

Evidence for treatment varies

Guidelines for the treatment of SCH (TABLE 1^8,14,19,20) are founded on the condition’s risk for progression to overt hypothyroidism and its association with health consequences such as cardiovascular disease. Guidelines of the American Thyroid Association (ATA) and European Thyroid Association (ETA), and those of the United Kingdom–based National Institute for Health and Care Excellence (NICE), prioritize treatment for individuals with a TSH level > 10 mIU/L^a and for those with TSH values < 10 mIU/L but still elevated and apparent symptoms of hypothyroidism.^14,19,20 The strength of evidence behind this guidance is challenged by a lack of data from prospective randomized controlled trials (RCTs) demonstrating health benefits following treatment of SCH. The British Medical Journal (BMJ) Guideline cites this lack of evidence and recommends against treating SCH at any TSH level, regardless of symptoms.⁸

There are few large RCTs of treatment outcomes for SCH. A 2017 RCT (the Thyroid Hormone Replacement for Untreated Older Adults with Subclinical Hypothyroidism, or TRUST, trial) of 737 adults older than 65 years with SCH evaluated the ability of levothyroxine to normalize TSH values compared with placebo. At 1 year, there was no difference in hypothyroid symptoms or tiredness scale scores with levothyroxine treatment compared with placebo.²¹ This finding was consistent even in the subgroup with a higher baseline symptom burden.²²

Continue to: Two small RCTs evaluated...

Two small RCTs evaluated treatment of SCH with depressive symptoms and cognitive function, neither finding benefit compared with placebo.^12,23 A 2018 systematic review and meta-analysis of 21 studies and 2192 adults did not show a benefit to quality of life or thyroid-specific symptoms in those treated for SCH compared with controls.²⁴

RCT support also is lacking for a reduction in cardiovascular mortality following treatment for SCH. A large population-level retrospective cohort from Denmark showed no difference in cardiovascular mortality or myocardial infarction in those treated for SCH compared with controls.²⁵ Pooled results from 2 RCTs (for patients older than 65 years, and those older than 80 years) showed no change in risk for cardiovascular outcomes in older adults treated for SCH.²⁶ Older adults treated for SCH in the TRUST trial showed no improvements in systolic or diastolic­ function on echocardiography.^27­ Two trials showed no difference in carotid intima-media thickness with treatment of SCH compared with placebo.^28,29

While most of the RCT data come from older adults, a retrospective cohort study in the United Kingdom of younger (ages 40-70 years; n = 3093) and older (age > 70 years; n = 1642) patients showed a reduction in cardiovascular mortality among treated patients who were younger (hazard ratio [HR] = 0.61; 4.2% vs. 6.6%) but not those who were older (HR = 0.99; 12.7% vs. 10.7%).³⁰ There is also evidence that thyroid size in those with goiter can be reduced with treatment of SCH.³¹

A measured approach to treating subclinical hypothyroidism

Consider several factors when deciding whether to treat SCH. For instance, RCT data suggest a lack of treatment benefit in relieving depression, improving cognition, or reducing general hypothyroid symptoms. Treatment of SCH in older adults does not appear to improve cardiovascular outcomes. The question of whether long-term treatment of SCH in younger patients reduces cardiovascular morbidity or mortality lacks answers from RCTs. Before diagnosing SCH or starting treatment, always confirm SCH with repeat testing in 2 to 3 months, as a high percentage of those with untreated SCH will have normal thyroid function on repeat testing.

Before diagnosing subclinical hypothyroidism (SCH) or starting treatment, always confirm SCH with repeat testing in 2 to 3 months.

In the event you and your patient elect to treat SCH, guidelines and trials generally support a low initial daily dose of 25 to 50 mcg of levothyroxine (T4), followed with dose changes­ every 4 to 8 weeks and a goal of normalizing TSH to within the lower half of the reference range (0.4-2.5 mIU/L).¹⁴ This is generally similar to published treatment goals for primary hypothyroidism and is based on studies suggesting the lower half of the reference range is normal for young, healthy, euthyroid individuals.³² Though full replacement doses (1.6-1.8 mcg/kg of ideal body weight) can be started for those who are elderly or who have ischemic heart disease or angina, this approach should be avoided in favor of low-dose initial therapy.³³ Thyroid supplements are best absorbed when taken apart from food, calcium, or iron supplements. The ATA suggests taking thyroid medication 60 minutes before breakfast or at bedtime (3 or more hours after the evening meal).³³

Continue to: Screening guidelines differ

Lacking population-level screening data from RCTs, most organizations do not recommend screening for thyroid dysfunction or they note insufficient evidence to make a screening recommendation (TABLE 2^17,19,20,34). In their most recent recommendation statement on the subject in 2015, the US Preventive Services Task Force (USPSTF) concluded the current evidence was insufficient to recommend for or against thyroid dysfunction screening in nonpregnant, asymptomatic adults.¹⁷ This differs from the ATA and the American Association of Clinical Endocrinology (AACE; formerly known as the American Association of Clinical Endocrinologists), which both recommend targeted screening for thyroid dysfunction based on symptoms or risk factors.²⁰

What about subclinical hypothyroidism in pregnancy?

Overt hypothyroidism is associated with adverse events during pregnancy and with subsequent neurodevelopmental complications in children, although the effects of SCH during pregnancy remain less certain. Concerns have been raised over the potential association of SCH with pregnancy loss, placental abruption, premature rupture of membranes, and neonatal death.³⁵ Historically, the prevalence of SCH during pregnancy has ranged from 2% to 2.5%, but using lower trimester-based TSH reference ranges, the prevalence of SCH in pregnancy may be as high as 15%.³⁵

Guided by a large RCT that failed to find benefit (pregnancy outcomes, neurodevelopmental outcomes in children) following treatment of SCH in pregnancy,³⁶ the American College of Obstetricians and Gynecologists (ACOG) recommends against routine screening for thyroid disease in pregnancy.³⁴ The ATA notes insufficient evidence to rec-ommend universal screening for thyroid dysfunction in pregnancy but recommends targeted screening of those with risk factors.³⁷ Data are conflicting on the benefit of treating known or recently detected SCH on pregnancy outcomes including pregnancy loss.^35,38 As such, the American Society of Reproductive Medicine and the ATA both generally recommend treatment of SCH in pregnant patients, particularly when the TSH is ≥ 4.0 mIU/L and TPOAbs are present.^37,39

^a The ATA, ETA, and NICE have slightly different recommendations when a TSH level = 10 mIU/L. ETA and NICE recommend prioritizing treatment for individuals with this level, while ATA recommends treatment when individual factors are also considered.

ACKNOWLEDGEMENT
The authors thank Family Medicine Medical Librarian Gwen Wilson, MLS, AHIP, for her assistance with literature searches.

CORRESPONDENCE
Nicholas LeFevre, MD, Family and Community Medicine, University of Missouri–Columbia School of Medicine, One Hospital Drive, M224 Medical Science Building, Columbia, MO 65212; nlefevre@health.missouri.edu

Meet Dr. A and Dr. M

Dr. A is a 50-year-old family physician who provides prenatal care in a busy practice. She sees patients in eight 4-hour clinic sessions per week and is on inpatient call 1 week out of every 2 months. Dr. A has become disillusioned with her practice. She typically works until 7 pm and arrives home exhausted, with little energy to interact with her family. She spends hours in the evenings and on weekends completing charts and answering phone calls. Dr. A is concerned because she recently gained weight and lacks an established fitness routine. The COVID-19 pandemic made life more difficult as she dealt with the risk of getting infected and the changing recommendations for treatment and prevention. After 20 years of practice, Dr. A wonders whether she should leave clinical medicine.

Dr. M is a single, 32-year-old family physician working at an academic medical center. Dr. M is unhappy in his job, is trying to grow his practice, and views himself as having little impact or autonomy. He finds himself lost while navigating the electronic health record (EHR) and struggles to be efficient in the clinic. Dr. M has multiple administrative responsibilities that require him to work evenings and weekends. Debt from medical school loans also motivates him to moonlight several weekends per month. Over the past few months, Dr. M has become frustrated and discouraged, making his depression more difficult to manage. He feels drained by the time he arrives home, where he lives alone. He has stopped exercising, socializing with friends, and dating. Dr. M often wonders if he is in the wrong profession.

Defining burnout, stress, and wellness

Dr. A and Dr. M are experiencing symptoms of burnout, common to physicians and other health care professionals. Recent studies showed an increase in burnout during the COVID-19 pandemic.^1,2 In a survey using the Maslach Burnout Inventory (MBI), approximately 44% of physicians reported at least one symptom of burnout.³ After adjusting for age, gender, relationship status, and hours worked per week, physicians were found to be at greater risk for burnout than nonphysician workers.³ The latest Medscape physician burnout survey found an increase in burnout among US physicians from 42% in 2021 to 47% in 2022 during the COVID-19 pandemic.¹ Rates of burnout were even higher among family physicians and other frontline (eg, emergency, infectious disease, and critical care) physicians.¹

Burnout has 3 key dimensions: (1) overwhelming exhaustion; (2) feelings of cynicism and detachment from the job; and (3) a sense of ineffectiveness and lack of accomplishment.⁴ The MBI is considered the standard tool for research in the field of burnout and has been repeatedly assessed for reliability and validity.⁴ The original MBI includes such items as: “I feel emotionally drained from my work,” “I feel like I’m working too hard on my job,” and “I worry that this job is hardening me emotionally.”⁵

According to the World Health Organization, burnout is an occupational phenomenon associated with chronic work-related stress that is not successfully managed.⁶ This definition emphasizes work stress as the cause of burnout, thus highlighting the importance of addressing the work environment.⁷ Physician burnout can affect physician health and wellness and the quality of patient care.^8-13 Because of the cost of burnout to individuals and the health care system, it is important to understand stressors that can lead to physician burnout.

Stress has been described as “physical, mental, or emotional strain or tension … when a person perceives that demands exceed the personal and social resources the individual is able to mobilize.”¹⁴ Work-related sources of stress affecting practicing physicians include long workdays, multiple bureaucratic tasks, lack of autonomy/control, and complex patients.^1,15

The COVID-19 pandemic is a stressor that increased physicians’ exposure to patient suffering and deaths and physicians’ vulnerability to disease at work.¹⁶ Physicians taking care of patients with COVID-19 risk infection and the possibility of infecting others.Online health records are another source of stress for many physicians.^17,18 Access to online health records on personal devices can blur the line between work and home. For each hour of direct patient contact, a physician spends an additional 2 hours interacting with an EHR.¹⁹ Among family physicians and other primary care physicians, increased EHR interaction outside clinic hours has been associated with decreased workplace satisfaction and increased rates of burnout.^11,19,20 Time spent on non-patient-facing clinical tasks, such as peer-to-peer reviews and billing queries, contributes more to burnout than clinic time alone.¹⁷

Continue to: These and other organizational factors...

A physician burnout survey found an increase in burnout among US physicians from 42% in 2021 to 47% in 2022 during the COVID-19 pandemic.

These and other organizational factors contribute to the stress experienced by physicians. Many describe themselves as feeling consumed by their work. At the beginning of the COVID-19 pandemic, physicians (and the rest of the health care team) had to quickly­ learn how to conduct virtual office visits. Clerical responsibilities increased as patients relied more on patient portals and telephone calls to receive care.

Who is predisposed to burnout? Although burnout is a work-related syndrome, studies have shown an increase in burnout associated with individual (ie, personal) factors. For example, female physicians have been shown to have higher rates of burnout compared with male physicians.^1,3 The stress of balancing the demands of the profession can begin during medical school and residency, with younger physicians having nearly­ twice the risk for stress-related symptoms when compared with older colleagues.^15,20-23 Having a child younger than 21 years old, and other personal factors related to balancing family and life demands, increases the likelihood of burnout.^11,21,22

Physicians with certain personality types and predispositions are at increased risk for burnout.^23-25 For example, neuroticism on the Big Five Personality Inventory (one of the most well-known of the psychology inventories) is associated with an increased risk for burnout. Neuroticism may manifest as sadness or related emotional dysregulation (eg, irritability, anxiety).²⁶ Other traits measured by the Big Five Personality Inventory include extraversion, agreeableness, conscientiousness, and openness to experience.²⁶

Physicians who were depressed were more likely to experience burnout symptoms (87.5%); however, only 26.2% of physicians experiencing burnout were diagnosed as having depression.

A history of depression is also associated with an increased risk for burnout.²⁷ Although depression and burnout are separate conditions, a 2016 study found significant overlap between the two.²⁷ Physicians in this study who were depressed were more likely to experience burnout symptoms (87.5%); however, only 26.2% of physicians experiencing burnout were diagnosed as having depression.²⁷ Rates of depression are higher among physicians when compared with nonphysicians, yet physicians are less likely to seek help due to fear of stigma and potential licensing concerns.^28,29 Because of this, when physicians experience depressive symptoms, they may respond by working harder rather than seeking professional counseling or emotional support. They might believe that “asking for help is a sign of weakness,” thus sacrificing their wellness.

Wellness encompasses a sense of thriving characterized by thoughts and feelings of contentment, joy, and fulfillment—and the absence of severe distress.³⁰ Wellness is a multifaceted condition that includes physical, psychological, and social aspects of an individual’s personal and professional life. Individuals experience a sense of wellness when they nurture their physical selves, minds, and relationships. People experience a sense of wellness when they balance their schedules, eat well, and maintain physical activity. Making time to enjoy family and friends also contributes to wellness.

Continue to: The culture of medicine often rewards...

The culture of medicine often rewards physician attitudes and behaviors that detract from wellness.³¹ Physicians internalize the culture of medicine that promotes perfectionism and downplays personal vulnerability.³² Physicians are reluctant to protect and preserve their wellness, believing self-sacrifice makes them good doctors. Physicians may spend countless hours counseling patients on the importance of wellness, but then work when ill or neglect their personal health needs and self-care—potentially decreasing their resilience and increasing the risk for burnout.³¹

Two paths to managing stress and preventing burnout

Patel and colleagues distinguish between 2 burnout intervention categories: (1) those that focus on individual physicians and (2) those that focus on the organizational environment.³³ We find these distinctions useful and offer strategies for enhancing individual physician wellness (TABLE 1^34-41). Similar to West and colleagues,¹¹ we offer strategies for addressing organizational sources of stress (TABLE 2^42-48). The following text describes these burnout intervention categories, emphasizing increasing self-care and changes that enable physicians to adapt effectively.

The recommendations outlined in this article are based on published stress and burnout literature, as well as the experiences of the authors. However, the number of randomized controlled studies of interventions aimed at reducing physician stress and burnout is limited. In addition, strategies proposed to reduce burnout in other professions may not address the unique stressors physicians encounter. Hence, our recommendations are limited. We have included interventions that seem optimal for individual physicians and the organizations that employ them.

Individual strategies target physical, psychological, and social wellness

Physician wellness strategies are divided into 3 categories: physical, psychological, and social wellness. Most strategies to improve physical wellness are widely known, evidence based, and recommended to patients by physicians.^34-36 For example, most physicians advise their patients to eat healthy balanced meals, avoid unhealthy foods and beverages, maintain a healthy body weight, get daily exercise and adequate sleep, avoid excessive alcohol use, and abstain from tobacco use. However, discrepancies between physicians’ advice to patients and their own behaviors are common. Simply stated, physicians are well advised to follow their own advice regarding physical self-care.

CBT and mindfulness are key to psychological wellness. Recommendations for enhancing psychological wellness are primarily derived from cognitive behavioral therapy (CBT) and mindfulness principles and practices.^37,38 CBT has been called the “gold standard” of psychotherapy, based on the breadth of research demonstrating that “no other form of psychotherapy has been shown to be systematically superior to CBT.”³⁹

Continue to: CBT is based on the premise...

CBT is based on the premise that individuals’ thoughts and beliefs largely determine how they feel (emotions) and act (behaviors). Certain thoughts lead to positive feelings and effective behaviors, while others lead to negative feelings and less effective behaviors. For example, when a physician has self-critical or helpless thoughts (eg, “I’m just no good at managing my life”), they are more likely to feel unhappy and abandon problem-solving. In contrast, when a physician has self-affirming or hopeful thoughts (eg, “This is difficult, but I have the personal resources to succeed”), they are more likely to feel confident and act to solve problems.

Physicians vacillate between these thoughts and beliefs, and their emotions and behaviors follow accordingly. When hyper-focused on “the hassles of medicine,” physicians feel defeated, depressed, and anxious about their work. In contrast, when physicians recognize and challenge problematic thoughts and focus on what they love about medicine, they feel good and interact with patients and coworkers in positive and self-reinforcing ways.

Mindfulness can help reduce psychological stress and increase personal fulfillment. Mindfulness is characterized as being in the present moment, fully accepting “what is,” and having a sense of gratitude and compassion for self and others.⁴⁰ In practice, mindfulness involves being intentional.

Dahl and colleagues⁴¹ describe a framework for human flourishing that includes 4 core dimensions of well-being (awareness, insight, connection, and purpose) that are all closely linked to mindful, intentional living. Based on their work, it is apparent that those who maintain a “heightened and flexible attentiveness” to their thoughts and feelings are likely to benefit by experiencing “improved mental health and psychological well-being.”⁴¹

However, the utility of CBT and mindfulness practices depends on receptivity to psychological interventions. Individuals who are not receptive may be hesitant to use these practices or likely will not benefit from them. Given these limitations of behavioral interventions, it would be helpful if more attention were paid to preventing and managing physician stress and burnout, especially through research focused on organizational changes.

Continue to: Supportive relationships are powerful

Supportive relationships are powerful. Finally, to enhance social wellness, it would be difficult to overstate the potential benefits of positive, supportive, close relationships.⁴² However, the demands of a career in medicine, starting in medical school, have the potential for inhibiting (rather than enhancing) close relationships.

Placing value on relationships with friends and family members is essential. As Dr. M began experiencing burnout, he felt increasingly lonely, yet he isolated himself from those who cared about him. Dr. A felt lonely at home, even though she was surrounded by family. Physicians are often reluctant to initiate vulnerable communication with others, believing “no one wants to hear about my problems.” However, by realizing the need for help and asking friends and family for emotional support, physicians can improve their wellness. Fostering supportive relationships can help provide the resilience needed to address organizational stressors.

Tackling organizational challenges

Long hours and pressure to see large numbers of patients (production demands) are a challenge across practice settings. Limiting work hours has been effective in improving the well-being of physician trainees but has had an inconsistent effect on burnout.^43,44

Organizations can offer flexible scheduling, and physicians considering limiting work hours may switch to part-time status or shift work. However, decreasing work hours may have the unintended consequence of increased stress as some physicians feel pressure to do more in less time.⁴⁵ Therefore, it’s important to set clear boundaries around work time and when and where work tasks are completed (eg, home vs office).

How we use technology matters. Given­ technology’s ever-increasing role in medicine, organizations must identify and use the most efficient, effective technology for managing clerical processes. When physicians participate in these decisions and share their experiences, technology is likely to be more user-friendly and impose less stress.⁴⁶

Continue to: If technology contributes to stress...