Federal Practitioner

Idiopathic granulomatous lobular mastitis (IGLM) is a rare, chronic inflammatory breast disease first described in 1972.¹ IGLM usually affects women during reproductive years and has similar clinical features to breast cancer.² Ultrasonography and mammography yield nonspecific results and cannot adequately differentiate between malignancy and inflammation.³ Magnetic resonance imaging (MRI) is known to be more sensitive in detecting lesions in dense breasts; however, it does not differentiate between granulomatous lesions and other disorders.^4,5 Histopathology is the gold standard for diagnosis.^1-12

Infectious and autoimmune causes of granulomatous mastitis must be excluded before establishing an IGLM diagnosis. The clinical quandary that remains is how to adequately manage the disease. Although there are no defined treatment guidelines, current literature has proposed a multimodal strategy.^6,9 In this report, we describe a case of IGLM successfully treated with surgical excision after failed medical therapy.

Case Presentation

A 43-year-old gravida 5, para 4 White woman presented with a 2-week history of right breast tenderness, heaviness, warmth, and redness that was refractory to cephalexin and dicloxacillin. She had no personal or family history of breast cancer; never had breast surgery and breastfed all 4 children.

An examination of the right breast demonstrated erythema and an 8-cm tender mass in the right lower outer quadrant but no skin retraction or dimpling (Figure 1). The mammography, concerning for inflammatory breast cancer, was category BI-RADS 4 and demonstrated a suspicious right axillary lymph node (Figure 2).

A core needle breast biopsy revealed granulomatous mastitis (Figure 3A), without evidence of malignancy. Rheumatology and endocrinology excluded secondary causes of granulomatous mastitis (ie, sarcoidosis, tuberculosis, granulomatosis with polyangiitis, and other autoimmune conditions). A pituitary MRI to assess an elevated serum prolactin level showed no evidence of microadenoma.

After a prolonged course of 8 months of unsuccessful therapy with prednisone and methotrexate, the patient was referred for surgical excision. Culture and special stains (Gram stain, periodic acid-Schiff stain, acid-fast Bacillus culture, Fite stain, and Brown and Benn stain) of the breast tissue were negative for organisms (Figure 3B). Seven months after excision the patient was doing well and had no evidence of recurrence.

Discussion

IGLM is a rare, chronic benign inflammatory breast disease of unknown etiology and more commonly reported in individuals of Mediterranean descent.¹³ It is believed that hyperprolactinemia causing extravasation of fat and protein during milk letdown leads to lymphocyte and macrophage migration, resulting in a localized autoimmune response in the breast ducts.^10,14

There are 2 types of granulomatous mastitis: idiopathic and specific. Infectious, autoimmune, and malignant causes of granulomatous mastitis (ie, tuberculosis, sarcoidosis, Corynebacterium spp, granulomatosis with polyangiitis, systemic lupus erythematosus, Behçet disease, ductal ectasia, or granulomatous reaction in a carcinoma) must be excluded prior to establishing an IGLM diagnosis, as these can be fatal if left untreated.¹⁵ The most frequent findings on ultrasound and mammography are hypoechoic masses and focal asymmetric densities, respectively.^3,5 MRI has been proposed more for surveillance in patients with chronic IGLM.^4,5 Histopathology—featuring lobular noncaseating granulomas with epithelioid histiocytes; and multinucleated giant cells in a background of neutrophils, lymphocytes, plasma cells, and eosinophils—is the gold standard for diagnosing IGLM.^1-12

There are currently no universal treatment guidelines and management usually consists of observation, systemic and topical steroids, or surgery.^3,13 Topical and injectable steroids have been effective in treating both initial and recurrent IGLM in patients who are unable to be treated with systemic steroids.^16-18 Due to reported high recurrence rates with steroid tapers, adjunctive therapy with methotrexate, azathioprine, colchicine, and hydroxychloroquine have been proposed.^1,3-6,10-12

Additionally, antibiotics are recommended only in the management of IGLM when microbial co-infection is concerning, such as with Corynebacterium spp.^9,11,19-22 Histologically, this bacterium is distinct from IGLM and demonstrates granulomatous, neutrophilic inflammation within cystic spaces.^19-21 Wide surgical excision with negative margins is the only definitive treatment to reduce recurrence and expedite recovery time.^2,3,7-10 Notably, surgical excision has been associated with poor wound healing and occasional recurrence compared with medication alone.^5,11

Although IGLM is normally a benign process, chronic disease has been related (without causality) to infiltrating breast carcinoma.⁴ A proposed theory for the development of malignancy suggests that chronic inflammation leading to free radical formation can result in cellular dysplasia and cancer.²³

Conclusions

Fifty years after its first description, IGLM is still a poorly understood disease. There remains no consensus behind its etiology or management. In our case, we demonstrated a stepwise treatment progression, beginning with medical therapy before proceeding to surgical cure. Given concerns for poor wound healing and postsurgical infections, monitoring the response and recurrence to an initial trial of conservative medical treatment is not unreasonable. Because of possible risk for malignancy with chronic IGLM, patients should not delay surgical excision if their condition remains refractory to medical therapy alone.

Idiopathic granulomatous lobular mastitis (IGLM) is a rare, chronic inflammatory breast disease first described in 1972.¹ IGLM usually affects women during reproductive years and has similar clinical features to breast cancer.² Ultrasonography and mammography yield nonspecific results and cannot adequately differentiate between malignancy and inflammation.³ Magnetic resonance imaging (MRI) is known to be more sensitive in detecting lesions in dense breasts; however, it does not differentiate between granulomatous lesions and other disorders.^4,5 Histopathology is the gold standard for diagnosis.^1-12

Infectious and autoimmune causes of granulomatous mastitis must be excluded before establishing an IGLM diagnosis. The clinical quandary that remains is how to adequately manage the disease. Although there are no defined treatment guidelines, current literature has proposed a multimodal strategy.^6,9 In this report, we describe a case of IGLM successfully treated with surgical excision after failed medical therapy.

Case Presentation

A 43-year-old gravida 5, para 4 White woman presented with a 2-week history of right breast tenderness, heaviness, warmth, and redness that was refractory to cephalexin and dicloxacillin. She had no personal or family history of breast cancer; never had breast surgery and breastfed all 4 children.

An examination of the right breast demonstrated erythema and an 8-cm tender mass in the right lower outer quadrant but no skin retraction or dimpling (Figure 1). The mammography, concerning for inflammatory breast cancer, was category BI-RADS 4 and demonstrated a suspicious right axillary lymph node (Figure 2).

A core needle breast biopsy revealed granulomatous mastitis (Figure 3A), without evidence of malignancy. Rheumatology and endocrinology excluded secondary causes of granulomatous mastitis (ie, sarcoidosis, tuberculosis, granulomatosis with polyangiitis, and other autoimmune conditions). A pituitary MRI to assess an elevated serum prolactin level showed no evidence of microadenoma.

After a prolonged course of 8 months of unsuccessful therapy with prednisone and methotrexate, the patient was referred for surgical excision. Culture and special stains (Gram stain, periodic acid-Schiff stain, acid-fast Bacillus culture, Fite stain, and Brown and Benn stain) of the breast tissue were negative for organisms (Figure 3B). Seven months after excision the patient was doing well and had no evidence of recurrence.

Discussion

IGLM is a rare, chronic benign inflammatory breast disease of unknown etiology and more commonly reported in individuals of Mediterranean descent.¹³ It is believed that hyperprolactinemia causing extravasation of fat and protein during milk letdown leads to lymphocyte and macrophage migration, resulting in a localized autoimmune response in the breast ducts.^10,14

There are 2 types of granulomatous mastitis: idiopathic and specific. Infectious, autoimmune, and malignant causes of granulomatous mastitis (ie, tuberculosis, sarcoidosis, Corynebacterium spp, granulomatosis with polyangiitis, systemic lupus erythematosus, Behçet disease, ductal ectasia, or granulomatous reaction in a carcinoma) must be excluded prior to establishing an IGLM diagnosis, as these can be fatal if left untreated.¹⁵ The most frequent findings on ultrasound and mammography are hypoechoic masses and focal asymmetric densities, respectively.^3,5 MRI has been proposed more for surveillance in patients with chronic IGLM.^4,5 Histopathology—featuring lobular noncaseating granulomas with epithelioid histiocytes; and multinucleated giant cells in a background of neutrophils, lymphocytes, plasma cells, and eosinophils—is the gold standard for diagnosing IGLM.^1-12

There are currently no universal treatment guidelines and management usually consists of observation, systemic and topical steroids, or surgery.^3,13 Topical and injectable steroids have been effective in treating both initial and recurrent IGLM in patients who are unable to be treated with systemic steroids.^16-18 Due to reported high recurrence rates with steroid tapers, adjunctive therapy with methotrexate, azathioprine, colchicine, and hydroxychloroquine have been proposed.^1,3-6,10-12

Additionally, antibiotics are recommended only in the management of IGLM when microbial co-infection is concerning, such as with Corynebacterium spp.^9,11,19-22 Histologically, this bacterium is distinct from IGLM and demonstrates granulomatous, neutrophilic inflammation within cystic spaces.^19-21 Wide surgical excision with negative margins is the only definitive treatment to reduce recurrence and expedite recovery time.^2,3,7-10 Notably, surgical excision has been associated with poor wound healing and occasional recurrence compared with medication alone.^5,11

Although IGLM is normally a benign process, chronic disease has been related (without causality) to infiltrating breast carcinoma.⁴ A proposed theory for the development of malignancy suggests that chronic inflammation leading to free radical formation can result in cellular dysplasia and cancer.²³

Conclusions

Fifty years after its first description, IGLM is still a poorly understood disease. There remains no consensus behind its etiology or management. In our case, we demonstrated a stepwise treatment progression, beginning with medical therapy before proceeding to surgical cure. Given concerns for poor wound healing and postsurgical infections, monitoring the response and recurrence to an initial trial of conservative medical treatment is not unreasonable. Because of possible risk for malignancy with chronic IGLM, patients should not delay surgical excision if their condition remains refractory to medical therapy alone.

Idiopathic granulomatous lobular mastitis (IGLM) is a rare, chronic inflammatory breast disease first described in 1972.¹ IGLM usually affects women during reproductive years and has similar clinical features to breast cancer.² Ultrasonography and mammography yield nonspecific results and cannot adequately differentiate between malignancy and inflammation.³ Magnetic resonance imaging (MRI) is known to be more sensitive in detecting lesions in dense breasts; however, it does not differentiate between granulomatous lesions and other disorders.^4,5 Histopathology is the gold standard for diagnosis.^1-12

Infectious and autoimmune causes of granulomatous mastitis must be excluded before establishing an IGLM diagnosis. The clinical quandary that remains is how to adequately manage the disease. Although there are no defined treatment guidelines, current literature has proposed a multimodal strategy.^6,9 In this report, we describe a case of IGLM successfully treated with surgical excision after failed medical therapy.

Case Presentation

A 43-year-old gravida 5, para 4 White woman presented with a 2-week history of right breast tenderness, heaviness, warmth, and redness that was refractory to cephalexin and dicloxacillin. She had no personal or family history of breast cancer; never had breast surgery and breastfed all 4 children.

An examination of the right breast demonstrated erythema and an 8-cm tender mass in the right lower outer quadrant but no skin retraction or dimpling (Figure 1). The mammography, concerning for inflammatory breast cancer, was category BI-RADS 4 and demonstrated a suspicious right axillary lymph node (Figure 2).

A core needle breast biopsy revealed granulomatous mastitis (Figure 3A), without evidence of malignancy. Rheumatology and endocrinology excluded secondary causes of granulomatous mastitis (ie, sarcoidosis, tuberculosis, granulomatosis with polyangiitis, and other autoimmune conditions). A pituitary MRI to assess an elevated serum prolactin level showed no evidence of microadenoma.

After a prolonged course of 8 months of unsuccessful therapy with prednisone and methotrexate, the patient was referred for surgical excision. Culture and special stains (Gram stain, periodic acid-Schiff stain, acid-fast Bacillus culture, Fite stain, and Brown and Benn stain) of the breast tissue were negative for organisms (Figure 3B). Seven months after excision the patient was doing well and had no evidence of recurrence.

Discussion

IGLM is a rare, chronic benign inflammatory breast disease of unknown etiology and more commonly reported in individuals of Mediterranean descent.¹³ It is believed that hyperprolactinemia causing extravasation of fat and protein during milk letdown leads to lymphocyte and macrophage migration, resulting in a localized autoimmune response in the breast ducts.^10,14

There are 2 types of granulomatous mastitis: idiopathic and specific. Infectious, autoimmune, and malignant causes of granulomatous mastitis (ie, tuberculosis, sarcoidosis, Corynebacterium spp, granulomatosis with polyangiitis, systemic lupus erythematosus, Behçet disease, ductal ectasia, or granulomatous reaction in a carcinoma) must be excluded prior to establishing an IGLM diagnosis, as these can be fatal if left untreated.¹⁵ The most frequent findings on ultrasound and mammography are hypoechoic masses and focal asymmetric densities, respectively.^3,5 MRI has been proposed more for surveillance in patients with chronic IGLM.^4,5 Histopathology—featuring lobular noncaseating granulomas with epithelioid histiocytes; and multinucleated giant cells in a background of neutrophils, lymphocytes, plasma cells, and eosinophils—is the gold standard for diagnosing IGLM.^1-12

There are currently no universal treatment guidelines and management usually consists of observation, systemic and topical steroids, or surgery.^3,13 Topical and injectable steroids have been effective in treating both initial and recurrent IGLM in patients who are unable to be treated with systemic steroids.^16-18 Due to reported high recurrence rates with steroid tapers, adjunctive therapy with methotrexate, azathioprine, colchicine, and hydroxychloroquine have been proposed.^1,3-6,10-12

Additionally, antibiotics are recommended only in the management of IGLM when microbial co-infection is concerning, such as with Corynebacterium spp.^9,11,19-22 Histologically, this bacterium is distinct from IGLM and demonstrates granulomatous, neutrophilic inflammation within cystic spaces.^19-21 Wide surgical excision with negative margins is the only definitive treatment to reduce recurrence and expedite recovery time.^2,3,7-10 Notably, surgical excision has been associated with poor wound healing and occasional recurrence compared with medication alone.^5,11

Although IGLM is normally a benign process, chronic disease has been related (without causality) to infiltrating breast carcinoma.⁴ A proposed theory for the development of malignancy suggests that chronic inflammation leading to free radical formation can result in cellular dysplasia and cancer.²³

Conclusions

Fifty years after its first description, IGLM is still a poorly understood disease. There remains no consensus behind its etiology or management. In our case, we demonstrated a stepwise treatment progression, beginning with medical therapy before proceeding to surgical cure. Given concerns for poor wound healing and postsurgical infections, monitoring the response and recurrence to an initial trial of conservative medical treatment is not unreasonable. Because of possible risk for malignancy with chronic IGLM, patients should not delay surgical excision if their condition remains refractory to medical therapy alone.

Discussion

An interdisciplinary discussion regarding the diagnosis considered the clinical features of the patient along with the imaging characteristics. The histological examination demonstrating sarcomatoid features with the supporting immunohistochemistry that confirmed both mesenchymal and epithelioid presence and was used to make the diagnosis of pulmonary sarcomatoid carcinoma (PSC).

Conclusions

PSC is diagnostically challenging because it is rare and has an aggressive progression. Identification of this tumor requires knowledge of histological criteria to identify their subtypes. Immunohistochemistry has an important role in the classification and to rule out differential diagnoses, including metastatic spread. Nevertheless, a reliable diagnosis requires precise histopathological examination, reached with surgical resection. Therefore, a detailed history, physical examination, systematic investigation, and correlation with chest imaging are needed to avoid misdiagnosis.

Our case highlights the importance of keeping this rare, aggressive tumor as part of the differential diagnosis. In view of its natural history, heterogeneity, and low incidence, published cases of PSC are limited. Thus, further investigation could optimize rapid identification and treatment options.

Discussion

An interdisciplinary discussion regarding the diagnosis considered the clinical features of the patient along with the imaging characteristics. The histological examination demonstrating sarcomatoid features with the supporting immunohistochemistry that confirmed both mesenchymal and epithelioid presence and was used to make the diagnosis of pulmonary sarcomatoid carcinoma (PSC).

Conclusions

PSC is diagnostically challenging because it is rare and has an aggressive progression. Identification of this tumor requires knowledge of histological criteria to identify their subtypes. Immunohistochemistry has an important role in the classification and to rule out differential diagnoses, including metastatic spread. Nevertheless, a reliable diagnosis requires precise histopathological examination, reached with surgical resection. Therefore, a detailed history, physical examination, systematic investigation, and correlation with chest imaging are needed to avoid misdiagnosis.

Our case highlights the importance of keeping this rare, aggressive tumor as part of the differential diagnosis. In view of its natural history, heterogeneity, and low incidence, published cases of PSC are limited. Thus, further investigation could optimize rapid identification and treatment options.

Discussion

An interdisciplinary discussion regarding the diagnosis considered the clinical features of the patient along with the imaging characteristics. The histological examination demonstrating sarcomatoid features with the supporting immunohistochemistry that confirmed both mesenchymal and epithelioid presence and was used to make the diagnosis of pulmonary sarcomatoid carcinoma (PSC).

Conclusions

PSC is diagnostically challenging because it is rare and has an aggressive progression. Identification of this tumor requires knowledge of histological criteria to identify their subtypes. Immunohistochemistry has an important role in the classification and to rule out differential diagnoses, including metastatic spread. Nevertheless, a reliable diagnosis requires precise histopathological examination, reached with surgical resection. Therefore, a detailed history, physical examination, systematic investigation, and correlation with chest imaging are needed to avoid misdiagnosis.

Our case highlights the importance of keeping this rare, aggressive tumor as part of the differential diagnosis. In view of its natural history, heterogeneity, and low incidence, published cases of PSC are limited. Thus, further investigation could optimize rapid identification and treatment options.

Many patients take herbals as alternative supplements to boost energy and mood. There are increasing reports of unintended adverse effects related to these supplements, particularly to the liver.^1-3 A study by the Drug-Induced Liver Injury Network found that liver injury caused by herbals and dietary supplements has increased from 7% in 2004 to 20% in 2013.⁴

The supplement ashwagandha has become increasingly popular. Ashwagandha is extracted from the root of Withania somnifera (W somnifera). It is purported to have health benefits, such as improving men’s health and increasing strength, memory, and learning abilities while decreasing anxiety and counteracting chronic fatigue.^5,6 W somnifera generally has been considered safe, though recently, a few case reports suggest that it may lead to a cholestatic pattern of injury.^5-7

To date, the factors defining the population at risk for ashwagandha toxicity are unclear, and an understanding of how to diagnose drug-induced liver injury is still immature in clinical practice. The regulation and study of the herbal and dietary supplement industry remain challenging. While many so-called natural substances are well tolerated, others can have unanticipated and harmful adverse effects and drug interactions. Future research should not only identify potentially harmful substances, but also which patients may be at greatest risk.

Case Presentation

A 48-year-old man with a history of severe alcohol use disorder (AUD) complicated by fatty liver and withdrawal seizures and delirium tremens, hypertension, depression, and anxiety presented to the emergency department (ED) after 4 days of having jaundice, epigastric abdominal pain, dark urine, and pale stools. In the preceding months, he had increased his alcohol use to as many as 12 drinks daily due to depression. After experiencing a blackout, he stopped drinking 7 days before presenting to the ED. He felt withdrawal symptoms, including tremors, diaphoresis, abdominal pain, nausea, and vomiting. On the third day of withdrawals, he reported that he had started taking an over-the-counter testosterone-boosting supplement to increase his energy, which he referred to as TestBoost—a mix of 8 ingredients, including ashwagandha, eleuthero root, Hawthorn berry, longjack, ginseng root, mushroom extract, bindii, and horny goat weed. After taking the supplement for 2 days, he noticed that his urine darkened, his stools became paler, his abdominal pain worsened, and he became jaundiced. After 2 additional days without improvement, and still taking the supplement, he presented to the ED. He reported having no fever, chills, recent illness, chest pain, shortness of breath, melena, lower extremity swelling, recent travel, or any changes in medications.

The patient had a 100.1 °F temperature, 102 beats per minute pulse; 129/94 mm Hg blood pressure, 18 beats per minute respiratory rate, and 97% oxygen saturation on room air on admission. He was in no acute distress, though his examination was notable for generalized jaundice and scleral icterus. He was mildly tender to palpation in the epigastric and right upper quadrant region. He was alert and oriented without confusion. He did not have any asterixis or spider angiomas, though he had scattered bruises on his left flank and left calf. His laboratory results were notable for mildly elevated aspartate aminotransferase (AST), 58 U/L (reference range, 13-35); alanine transaminase (ALT), 49 U/L (reference range, 7-45); and alkaline phosphatase (ALP), 98 U/L (reference range 33-94); total bilirubin, 13.6 mg/dL (reference range, 0.2-1.0); direct bilirubin, 8.4 mg/dL (reference range, 0.2-1); and international normalized ratio (INR), 1.11 (reference range, 2-3). His white blood cell and platelet counts were not remarkable at 9790/μL (reference range, 4500-11,000) and 337,000/μL (reference range, 150,000-440,000), respectively. Abdominal ultrasound and computed tomography (CT) revealed fatty liver with contracted gallbladder and no biliary dilatation. Urine ethanol levels were negative. The gastrointestinal (GI) service was consulted and agreed that his cholestatic injury was nonobstructive and likely related to the ashwagandha component of his supplement. The recommendation was cessation with close outpatient follow-up.

The patient was not prescribed any additional medications, such as steroids or ursodiol. He ceased supplement use following hospitalization; but relapsed into alcohol use 1 month after his discharge. Within 3 weeks, his total bilirubin had improved to 2.87 mg/dL, though AST, ALT, and ALP worsened to 127 U/L, 152 U/L, and 140 U/L, respectively. According to the notes of his psychiatrist who saw him at the time the laboratory tests were drawn, he had remained sober since discharge. His acute hepatitis panel drawn on admission was negative, and he demonstrated immunity to hepatitis A and B. Urine toxicology was negative. Antinuclear antibody (ANA) test was negative 1 year prior to discharge. Epstein-Barr virus (EBV), cytomegalovirus (CMV), ANA, antismooth muscle antibody, and immunoglobulins were not checked as suspicion for these etiologies was low. The Roussel Uclaf Causality Assessment Method (RUCAM) score was calculated as 6 (+1 for timing, +2 for drop in total bilirubin, +1 for ethanol risk factor, 0 for no other drugs, 0 for rule out of other diseases, +2 for known hepatotoxicity, 0 no repeat administration) for this patient indicating probable adverse drug reaction liver injury (Tables 1 and 2). However, we acknowledge that CMV, EBV, and herpes simplex virus status were not tested.

The 8 ingredients contained in TestBoost aside from ashwagandha did not have any major known liver adverse effects per a major database of medications. The other ingredients include eleuthero root, Hawthorn berry (crataegus laevigata), longjack (eurycoma longifolla) root, American ginseng root (American panax ginseng—panax quinquefolius), and Cordyceps mycelium (mushroom) extract, bindii (Tribulus terrestris), and epimedium grandiflorum (horny goat weed).⁶ No assays were performed to confirm purity of the ingredients in the patient’s supplement container.

Alcoholic hepatitis is an important consideration in this patient with AUD, though the timing of symptoms with supplement use and the cholestatic injury pattern with normal INR seems more consistent with drug-induced injury. Viral, infectious, and obstructive etiologies also were investigated. Acute viral hepatitis was ruled out based on bloodwork. The normal hepatobiliary tree on both ultrasound and CT effectively ruled out acute cholecystitis, cholangitis, and choledocholithiasis and there was no further indication for magnetic resonance cholangiopancreatography. There was no hepatic vein clot suggestive of Budd-Chiari syndrome. Autoimmune hepatitis was thought to be unlikely given that the etiology of injury seemed cholestatic in nature. Given the timing of the liver injury relative to supplement use it is likely that ashwagandha was a causative factor of this patient’s liver injury overlaid on an already strained liver from increased alcohol abuse.

The patient did not follow up with the GI service as an outpatient. There are no reports that the patient continued using the testosterone booster. His bilirubin improved dramatically within 1.5 months while his liver enzymes peaked 3 weeks later, with ALT ≥ AST. During his next admission 3 months later, he had relapsed, and his liver enzymes had the classic 2:1 AST to ALT ratio.

Discussion

Generally, ashwagandha has been thought to be well tolerated and possibly hepatoprotective.^7-10 However, recent studies suggest potential for hepatotoxicity, though without clear guidance about which patients are most at risk.^5,11,12 A study by Inagaki and colleagues suggests the potential for dose-dependent mechanism of liver injury, and this is supported by in vitro CYP450 inhibition with high doses of W Somnifera extract.^11,13 We hypothesize that there may be a multihit process that makes some patients more susceptible to supplement harm, particularly those with repeated exposures and with ongoing exposure to hepatic toxins, such as AUD.¹⁴ Supplements should be used with more caution in these individuals.

Additionally, although there are no validated guidelines to confirm the diagnosis of drug-induced liver injury (DILI) from a manufactured medication or herbal remedy, the Council for International Organizations of Medical Sciences (CIOMS) developed RUCAM, a set of diagnostic criteria for DILI, which can be used to determine the probability of DILI based on pattern of injury.¹⁵ Although not widely used in clinical practice, RUCAM can help identify the possibility of DILI outside of expert consensus.¹⁶ It seems to have better discriminative ability than the Maria and Victorino scale, also used to identify DILI.^16,17 While there is no replacement for clinical judgment, these scales may aid in identifying potential causes of DILI. The National Institutes of Health also has a LiverTox online tool that can assist health care professionals in identifying potentially hepatotoxic substances.⁶

Conclusions

We present a patient with AUD who developed cholestatic liver injury after ashwagandha use. Crucial to the diagnostic process is quantifying the amount ingested before presentation and the presence of contaminants, which is currently difficult to quantify given the lack of mechanisms to test supplements expediently in this manner in the clinical setting, which also requires the patient to bring in the supplements directly. There is also a lack of regulation and uniformity in these products. A clinician may be inclined to measure ashwagandha serum levels; however, such a test is not available to our knowledge. Nonetheless, using clinical tools such as RUCAM and utilizing databases, such as LiverTox, may help clinicians identify and remove potentially unsafe supplements. While there are many possible synergies between current medical practice and herbal remedies, practitioners must take care to first do no harm, as outlined in our Hippocratic Oath.

Many patients take herbals as alternative supplements to boost energy and mood. There are increasing reports of unintended adverse effects related to these supplements, particularly to the liver.^1-3 A study by the Drug-Induced Liver Injury Network found that liver injury caused by herbals and dietary supplements has increased from 7% in 2004 to 20% in 2013.⁴

The supplement ashwagandha has become increasingly popular. Ashwagandha is extracted from the root of Withania somnifera (W somnifera). It is purported to have health benefits, such as improving men’s health and increasing strength, memory, and learning abilities while decreasing anxiety and counteracting chronic fatigue.^5,6 W somnifera generally has been considered safe, though recently, a few case reports suggest that it may lead to a cholestatic pattern of injury.^5-7

To date, the factors defining the population at risk for ashwagandha toxicity are unclear, and an understanding of how to diagnose drug-induced liver injury is still immature in clinical practice. The regulation and study of the herbal and dietary supplement industry remain challenging. While many so-called natural substances are well tolerated, others can have unanticipated and harmful adverse effects and drug interactions. Future research should not only identify potentially harmful substances, but also which patients may be at greatest risk.

Case Presentation

A 48-year-old man with a history of severe alcohol use disorder (AUD) complicated by fatty liver and withdrawal seizures and delirium tremens, hypertension, depression, and anxiety presented to the emergency department (ED) after 4 days of having jaundice, epigastric abdominal pain, dark urine, and pale stools. In the preceding months, he had increased his alcohol use to as many as 12 drinks daily due to depression. After experiencing a blackout, he stopped drinking 7 days before presenting to the ED. He felt withdrawal symptoms, including tremors, diaphoresis, abdominal pain, nausea, and vomiting. On the third day of withdrawals, he reported that he had started taking an over-the-counter testosterone-boosting supplement to increase his energy, which he referred to as TestBoost—a mix of 8 ingredients, including ashwagandha, eleuthero root, Hawthorn berry, longjack, ginseng root, mushroom extract, bindii, and horny goat weed. After taking the supplement for 2 days, he noticed that his urine darkened, his stools became paler, his abdominal pain worsened, and he became jaundiced. After 2 additional days without improvement, and still taking the supplement, he presented to the ED. He reported having no fever, chills, recent illness, chest pain, shortness of breath, melena, lower extremity swelling, recent travel, or any changes in medications.

The patient had a 100.1 °F temperature, 102 beats per minute pulse; 129/94 mm Hg blood pressure, 18 beats per minute respiratory rate, and 97% oxygen saturation on room air on admission. He was in no acute distress, though his examination was notable for generalized jaundice and scleral icterus. He was mildly tender to palpation in the epigastric and right upper quadrant region. He was alert and oriented without confusion. He did not have any asterixis or spider angiomas, though he had scattered bruises on his left flank and left calf. His laboratory results were notable for mildly elevated aspartate aminotransferase (AST), 58 U/L (reference range, 13-35); alanine transaminase (ALT), 49 U/L (reference range, 7-45); and alkaline phosphatase (ALP), 98 U/L (reference range 33-94); total bilirubin, 13.6 mg/dL (reference range, 0.2-1.0); direct bilirubin, 8.4 mg/dL (reference range, 0.2-1); and international normalized ratio (INR), 1.11 (reference range, 2-3). His white blood cell and platelet counts were not remarkable at 9790/μL (reference range, 4500-11,000) and 337,000/μL (reference range, 150,000-440,000), respectively. Abdominal ultrasound and computed tomography (CT) revealed fatty liver with contracted gallbladder and no biliary dilatation. Urine ethanol levels were negative. The gastrointestinal (GI) service was consulted and agreed that his cholestatic injury was nonobstructive and likely related to the ashwagandha component of his supplement. The recommendation was cessation with close outpatient follow-up.

The patient was not prescribed any additional medications, such as steroids or ursodiol. He ceased supplement use following hospitalization; but relapsed into alcohol use 1 month after his discharge. Within 3 weeks, his total bilirubin had improved to 2.87 mg/dL, though AST, ALT, and ALP worsened to 127 U/L, 152 U/L, and 140 U/L, respectively. According to the notes of his psychiatrist who saw him at the time the laboratory tests were drawn, he had remained sober since discharge. His acute hepatitis panel drawn on admission was negative, and he demonstrated immunity to hepatitis A and B. Urine toxicology was negative. Antinuclear antibody (ANA) test was negative 1 year prior to discharge. Epstein-Barr virus (EBV), cytomegalovirus (CMV), ANA, antismooth muscle antibody, and immunoglobulins were not checked as suspicion for these etiologies was low. The Roussel Uclaf Causality Assessment Method (RUCAM) score was calculated as 6 (+1 for timing, +2 for drop in total bilirubin, +1 for ethanol risk factor, 0 for no other drugs, 0 for rule out of other diseases, +2 for known hepatotoxicity, 0 no repeat administration) for this patient indicating probable adverse drug reaction liver injury (Tables 1 and 2). However, we acknowledge that CMV, EBV, and herpes simplex virus status were not tested.

The 8 ingredients contained in TestBoost aside from ashwagandha did not have any major known liver adverse effects per a major database of medications. The other ingredients include eleuthero root, Hawthorn berry (crataegus laevigata), longjack (eurycoma longifolla) root, American ginseng root (American panax ginseng—panax quinquefolius), and Cordyceps mycelium (mushroom) extract, bindii (Tribulus terrestris), and epimedium grandiflorum (horny goat weed).⁶ No assays were performed to confirm purity of the ingredients in the patient’s supplement container.

Alcoholic hepatitis is an important consideration in this patient with AUD, though the timing of symptoms with supplement use and the cholestatic injury pattern with normal INR seems more consistent with drug-induced injury. Viral, infectious, and obstructive etiologies also were investigated. Acute viral hepatitis was ruled out based on bloodwork. The normal hepatobiliary tree on both ultrasound and CT effectively ruled out acute cholecystitis, cholangitis, and choledocholithiasis and there was no further indication for magnetic resonance cholangiopancreatography. There was no hepatic vein clot suggestive of Budd-Chiari syndrome. Autoimmune hepatitis was thought to be unlikely given that the etiology of injury seemed cholestatic in nature. Given the timing of the liver injury relative to supplement use it is likely that ashwagandha was a causative factor of this patient’s liver injury overlaid on an already strained liver from increased alcohol abuse.

The patient did not follow up with the GI service as an outpatient. There are no reports that the patient continued using the testosterone booster. His bilirubin improved dramatically within 1.5 months while his liver enzymes peaked 3 weeks later, with ALT ≥ AST. During his next admission 3 months later, he had relapsed, and his liver enzymes had the classic 2:1 AST to ALT ratio.

Discussion

Generally, ashwagandha has been thought to be well tolerated and possibly hepatoprotective.^7-10 However, recent studies suggest potential for hepatotoxicity, though without clear guidance about which patients are most at risk.^5,11,12 A study by Inagaki and colleagues suggests the potential for dose-dependent mechanism of liver injury, and this is supported by in vitro CYP450 inhibition with high doses of W Somnifera extract.^11,13 We hypothesize that there may be a multihit process that makes some patients more susceptible to supplement harm, particularly those with repeated exposures and with ongoing exposure to hepatic toxins, such as AUD.¹⁴ Supplements should be used with more caution in these individuals.

Additionally, although there are no validated guidelines to confirm the diagnosis of drug-induced liver injury (DILI) from a manufactured medication or herbal remedy, the Council for International Organizations of Medical Sciences (CIOMS) developed RUCAM, a set of diagnostic criteria for DILI, which can be used to determine the probability of DILI based on pattern of injury.¹⁵ Although not widely used in clinical practice, RUCAM can help identify the possibility of DILI outside of expert consensus.¹⁶ It seems to have better discriminative ability than the Maria and Victorino scale, also used to identify DILI.^16,17 While there is no replacement for clinical judgment, these scales may aid in identifying potential causes of DILI. The National Institutes of Health also has a LiverTox online tool that can assist health care professionals in identifying potentially hepatotoxic substances.⁶

Conclusions

We present a patient with AUD who developed cholestatic liver injury after ashwagandha use. Crucial to the diagnostic process is quantifying the amount ingested before presentation and the presence of contaminants, which is currently difficult to quantify given the lack of mechanisms to test supplements expediently in this manner in the clinical setting, which also requires the patient to bring in the supplements directly. There is also a lack of regulation and uniformity in these products. A clinician may be inclined to measure ashwagandha serum levels; however, such a test is not available to our knowledge. Nonetheless, using clinical tools such as RUCAM and utilizing databases, such as LiverTox, may help clinicians identify and remove potentially unsafe supplements. While there are many possible synergies between current medical practice and herbal remedies, practitioners must take care to first do no harm, as outlined in our Hippocratic Oath.

Many patients take herbals as alternative supplements to boost energy and mood. There are increasing reports of unintended adverse effects related to these supplements, particularly to the liver.^1-3 A study by the Drug-Induced Liver Injury Network found that liver injury caused by herbals and dietary supplements has increased from 7% in 2004 to 20% in 2013.⁴

The supplement ashwagandha has become increasingly popular. Ashwagandha is extracted from the root of Withania somnifera (W somnifera). It is purported to have health benefits, such as improving men’s health and increasing strength, memory, and learning abilities while decreasing anxiety and counteracting chronic fatigue.^5,6 W somnifera generally has been considered safe, though recently, a few case reports suggest that it may lead to a cholestatic pattern of injury.^5-7

To date, the factors defining the population at risk for ashwagandha toxicity are unclear, and an understanding of how to diagnose drug-induced liver injury is still immature in clinical practice. The regulation and study of the herbal and dietary supplement industry remain challenging. While many so-called natural substances are well tolerated, others can have unanticipated and harmful adverse effects and drug interactions. Future research should not only identify potentially harmful substances, but also which patients may be at greatest risk.

Case Presentation

A 48-year-old man with a history of severe alcohol use disorder (AUD) complicated by fatty liver and withdrawal seizures and delirium tremens, hypertension, depression, and anxiety presented to the emergency department (ED) after 4 days of having jaundice, epigastric abdominal pain, dark urine, and pale stools. In the preceding months, he had increased his alcohol use to as many as 12 drinks daily due to depression. After experiencing a blackout, he stopped drinking 7 days before presenting to the ED. He felt withdrawal symptoms, including tremors, diaphoresis, abdominal pain, nausea, and vomiting. On the third day of withdrawals, he reported that he had started taking an over-the-counter testosterone-boosting supplement to increase his energy, which he referred to as TestBoost—a mix of 8 ingredients, including ashwagandha, eleuthero root, Hawthorn berry, longjack, ginseng root, mushroom extract, bindii, and horny goat weed. After taking the supplement for 2 days, he noticed that his urine darkened, his stools became paler, his abdominal pain worsened, and he became jaundiced. After 2 additional days without improvement, and still taking the supplement, he presented to the ED. He reported having no fever, chills, recent illness, chest pain, shortness of breath, melena, lower extremity swelling, recent travel, or any changes in medications.

The patient had a 100.1 °F temperature, 102 beats per minute pulse; 129/94 mm Hg blood pressure, 18 beats per minute respiratory rate, and 97% oxygen saturation on room air on admission. He was in no acute distress, though his examination was notable for generalized jaundice and scleral icterus. He was mildly tender to palpation in the epigastric and right upper quadrant region. He was alert and oriented without confusion. He did not have any asterixis or spider angiomas, though he had scattered bruises on his left flank and left calf. His laboratory results were notable for mildly elevated aspartate aminotransferase (AST), 58 U/L (reference range, 13-35); alanine transaminase (ALT), 49 U/L (reference range, 7-45); and alkaline phosphatase (ALP), 98 U/L (reference range 33-94); total bilirubin, 13.6 mg/dL (reference range, 0.2-1.0); direct bilirubin, 8.4 mg/dL (reference range, 0.2-1); and international normalized ratio (INR), 1.11 (reference range, 2-3). His white blood cell and platelet counts were not remarkable at 9790/μL (reference range, 4500-11,000) and 337,000/μL (reference range, 150,000-440,000), respectively. Abdominal ultrasound and computed tomography (CT) revealed fatty liver with contracted gallbladder and no biliary dilatation. Urine ethanol levels were negative. The gastrointestinal (GI) service was consulted and agreed that his cholestatic injury was nonobstructive and likely related to the ashwagandha component of his supplement. The recommendation was cessation with close outpatient follow-up.

The patient was not prescribed any additional medications, such as steroids or ursodiol. He ceased supplement use following hospitalization; but relapsed into alcohol use 1 month after his discharge. Within 3 weeks, his total bilirubin had improved to 2.87 mg/dL, though AST, ALT, and ALP worsened to 127 U/L, 152 U/L, and 140 U/L, respectively. According to the notes of his psychiatrist who saw him at the time the laboratory tests were drawn, he had remained sober since discharge. His acute hepatitis panel drawn on admission was negative, and he demonstrated immunity to hepatitis A and B. Urine toxicology was negative. Antinuclear antibody (ANA) test was negative 1 year prior to discharge. Epstein-Barr virus (EBV), cytomegalovirus (CMV), ANA, antismooth muscle antibody, and immunoglobulins were not checked as suspicion for these etiologies was low. The Roussel Uclaf Causality Assessment Method (RUCAM) score was calculated as 6 (+1 for timing, +2 for drop in total bilirubin, +1 for ethanol risk factor, 0 for no other drugs, 0 for rule out of other diseases, +2 for known hepatotoxicity, 0 no repeat administration) for this patient indicating probable adverse drug reaction liver injury (Tables 1 and 2). However, we acknowledge that CMV, EBV, and herpes simplex virus status were not tested.

The 8 ingredients contained in TestBoost aside from ashwagandha did not have any major known liver adverse effects per a major database of medications. The other ingredients include eleuthero root, Hawthorn berry (crataegus laevigata), longjack (eurycoma longifolla) root, American ginseng root (American panax ginseng—panax quinquefolius), and Cordyceps mycelium (mushroom) extract, bindii (Tribulus terrestris), and epimedium grandiflorum (horny goat weed).⁶ No assays were performed to confirm purity of the ingredients in the patient’s supplement container.

Alcoholic hepatitis is an important consideration in this patient with AUD, though the timing of symptoms with supplement use and the cholestatic injury pattern with normal INR seems more consistent with drug-induced injury. Viral, infectious, and obstructive etiologies also were investigated. Acute viral hepatitis was ruled out based on bloodwork. The normal hepatobiliary tree on both ultrasound and CT effectively ruled out acute cholecystitis, cholangitis, and choledocholithiasis and there was no further indication for magnetic resonance cholangiopancreatography. There was no hepatic vein clot suggestive of Budd-Chiari syndrome. Autoimmune hepatitis was thought to be unlikely given that the etiology of injury seemed cholestatic in nature. Given the timing of the liver injury relative to supplement use it is likely that ashwagandha was a causative factor of this patient’s liver injury overlaid on an already strained liver from increased alcohol abuse.

The patient did not follow up with the GI service as an outpatient. There are no reports that the patient continued using the testosterone booster. His bilirubin improved dramatically within 1.5 months while his liver enzymes peaked 3 weeks later, with ALT ≥ AST. During his next admission 3 months later, he had relapsed, and his liver enzymes had the classic 2:1 AST to ALT ratio.

Discussion

Generally, ashwagandha has been thought to be well tolerated and possibly hepatoprotective.^7-10 However, recent studies suggest potential for hepatotoxicity, though without clear guidance about which patients are most at risk.^5,11,12 A study by Inagaki and colleagues suggests the potential for dose-dependent mechanism of liver injury, and this is supported by in vitro CYP450 inhibition with high doses of W Somnifera extract.^11,13 We hypothesize that there may be a multihit process that makes some patients more susceptible to supplement harm, particularly those with repeated exposures and with ongoing exposure to hepatic toxins, such as AUD.¹⁴ Supplements should be used with more caution in these individuals.

Additionally, although there are no validated guidelines to confirm the diagnosis of drug-induced liver injury (DILI) from a manufactured medication or herbal remedy, the Council for International Organizations of Medical Sciences (CIOMS) developed RUCAM, a set of diagnostic criteria for DILI, which can be used to determine the probability of DILI based on pattern of injury.¹⁵ Although not widely used in clinical practice, RUCAM can help identify the possibility of DILI outside of expert consensus.¹⁶ It seems to have better discriminative ability than the Maria and Victorino scale, also used to identify DILI.^16,17 While there is no replacement for clinical judgment, these scales may aid in identifying potential causes of DILI. The National Institutes of Health also has a LiverTox online tool that can assist health care professionals in identifying potentially hepatotoxic substances.⁶

Conclusions

We present a patient with AUD who developed cholestatic liver injury after ashwagandha use. Crucial to the diagnostic process is quantifying the amount ingested before presentation and the presence of contaminants, which is currently difficult to quantify given the lack of mechanisms to test supplements expediently in this manner in the clinical setting, which also requires the patient to bring in the supplements directly. There is also a lack of regulation and uniformity in these products. A clinician may be inclined to measure ashwagandha serum levels; however, such a test is not available to our knowledge. Nonetheless, using clinical tools such as RUCAM and utilizing databases, such as LiverTox, may help clinicians identify and remove potentially unsafe supplements. While there are many possible synergies between current medical practice and herbal remedies, practitioners must take care to first do no harm, as outlined in our Hippocratic Oath.

Pharmacy call centers have been successfully implemented in outpatient and specialty pharmacy settings.¹ A centralized pharmacy call center gives patients immediate access to a pharmacist who can view their health records to answer specific questions or fulfill medication renewal requests.^2-4 Little literature exists to describe its use in an inpatient setting.

Inpatient pharmacies receive numerous calls from health care professionals and patients. Challenges related to phone calls in the inpatient pharmacy setting may include interruptions, distractions, low accountability, poor efficiency, lack of optimal resources, and staffing.⁵ An unequal distribution and lack of accountability may exist when answering phone calls for the inpatient pharmacy team, which may contribute to long hold times and call abandonment rates. Phone calls also may be directed inefficiently between clinical pharmacists (CPs) and pharmacy technicians. Team member time related to answering phone calls may not be captured or measured.

The Edward Hines, Jr. Veterans Affairs Hospital (EHJVAH) in Illinois offers primary, extended, and specialty care and is a tertiary care referral center. The facility operates 483 beds and serves 6 community-based outpatient clinics.

Implementation

A new inpatient pharmacy service phone line extension was implemented. Data used to report quality metrics were obtained from the Global Navigator (GNAV), an information system that records calls, tracks the performance of agents, and coordinates personnel scheduling. The effectiveness of the ACD system was evaluated by quality metric goals of mean speed to answer ≤ 30 seconds and mean abandonment rate ≤ 5%. This project was determined to be quality improvement and was not reviewed by the EHJVAH Institutional Review Board.

The ACD system was set up in December 2020. After a 1-month implementation period, metrics were reported to the inpatient pharmacy team and leadership. By January 2021, EHJVAH fully implemented an ACD phone system operated by inpatient pharmacy technicians and CPs. EHJVAH inpatient pharmacy includes CPs who practice without a scope of practice and board-certified pharmacy technicians in 3 shifts. The CPs and pharmacy technicians work in the central pharmacy (the main pharmacy and inpatient pharmacy vault) or are decentralized with responsibility for answering phone calls and making deliveries (pharmacy technicians).

The pharmacy leadership team decided to implement 1 phone line with 2 ACD splits. The first split was directed to pharmacy technicians and the second to CPs. The intention was to streamline calls to be directed to proper team members within the inpatient pharmacy. The CP line also was designed to back up the pharmacy technician line. These calls were equally distributed among staff based on a standard algorithm. The pharmacy greeting stated, “Thank you for contacting the inpatient pharmacy at Hines VA Hospital. For missing doses, unit stock requests, or to speak with a pharmacy technician, please press 1. For clinical questions, order verification, or to speak with a pharmacist, please press 2.” Each inpatient pharmacy team member had a unique system login.

Fourteen ACD phone stations were established in the main pharmacy and in decentralized locations for order verification. The stations were distributed across the pharmacy service to optimize workload, space, and resources.

Training and Communication

Before implementing the inpatient pharmacy ACD phone system, the CPs and pharmacy technicians received mandatory ACD training. After the training, pharmacy team members were required to sign off on the training document to indicate that they had completed the course. The pharmacy team was trained on the importance of staffing the phones continuously. As a 24-hour pharmacy service in the acute care setting, any call may be critical for patient care.

A hospital-wide memorandum was distributed via email to all unit managers and hospital staff to educate them on the new ACD phone system, which included a new phone line extension for the inpatient pharmacy. Additionally, the inpatient pharmacy team was trained on the proper way of communicating the ACD phone system process with the hospital staff. The inpatient pharmacy team was notified that there would be an educational period to explain the queue process to hospital staff. Occasionally, hospital staff believed they were speaking to an automated system and hung up before their call was answered. The inpatient pharmacy team was instructed to notify the hospital staff to stay on the line since their call would be answered in the order it was received. Once the inpatient pharmacy team received proper training and felt comfortable with the phone system, it was set up and integrated into the workflow.

Postimplementation Evaluation

Inpatient pharmacy ACD phone system data were collected for 2021. To evaluate the effectiveness of an ACD system, the pharmacy leadership team set up the following metrics and goals for inpatient CPs and inpatient pharmacy technicians for monthly call volume/abandonment rate, mean speed to answer, mean call volume by shift, and the mean abandonment rate by shift.

Discussion

Since implementing the inpatient pharmacy ACD phone system in January 2021, there have been successes and challenges. The implementation increased accountability and efficiency when answering pharmacy phone calls. An ACD uses an algorithm that ensures equitable distribution of phone calls between CPs and pharmacy technicians. Through this algorithm, the pharmacy team is held more accountable when answering incoming calls. Distributing phone calls equally allows for optimization and balances the workload. The ACD phone system also improved efficiency when answering incoming calls. By incorporating splits when a patient or health care professional calls, ACD routes the question to the appropriate staff member. As a result, CPs spend less time answering questions meant for pharmacy technicians and instead can answer clinical or order verification questions more efficiently.

ACD data also allow pharmacy leadership to assess staffing needs, depending on the call volume. Based on ACD data, the busiest time of day was 8:00 AM to 4:00 PM. Based on this information, pharmacy leadership plans to staff more appropriately to have more pharmacy technicians working during the first shift to attend to phone calls.

The mean call abandonment rate was 4.7% for both CPs and pharmacy technicians, which met the ≤ 5% goal. The highest call abandonment rate was from midnight to 8 AM, though this shift also experienced the lowest call volume. This trend may be attributed to fewer pharmacy team members available to meet the demands of the overnight shift.

Pharmacy technicians handled a higher total call volume, which may be attributed to more phone calls related to missing doses or unit stock requests compared with clinical questions or order verifications. This information may be beneficial to identify opportunities to improve pharmacy operations.

The main challenges encountered in the ACD implementation process were hardware installation and communication with hospital staff about the changes in the inpatient pharmacy phone system. To implement the new inpatient pharmacy ACD phone system, previous telephones and hardware were removed and replaced. Initially, hardware and installation delays made it difficult for the ACD phone system to operate efficiently in the early months of its implementation. The inpatient pharmacy team depends on the telecommunications system and computers for their daily activities. Delays and issues with the hardware and ACD phone system made it more difficult to provide patient care.

Communication is a continuous challenge to ensure that hospital staff are notified of the new inpatient pharmacy ACD phone number. Over time, the understanding and use of the new ACD phone system have increased dramatically, but there are still opportunities to capture any misdirected calls. Informal feedback was obtained at pharmacy huddles and 1-on-1 discussions with pharmacy staff, and the opinions were mixed. Members of the pharmacy staff expressed that the ACD phone system set up an effective way to triage phone calls. Another positive comment was that the system created a means of accountability for pharmacy phone calls. Critical feedback included challenges with triaging phone calls to appropriate pharmacists, because calls are assigned based on an algorithm, whereas clinical coverage is determined by designated unit daily assignments.

Limitations

There are potential limitations to this quality improvement project. This phone system may not apply to all inpatient hospital pharmacy settings. Potential limitations for implementation at other institutions may include but are not limited to, differing pharmacy practice models (centralized vs decentralized), implementation costs, and internal resources.

Future Goals

To improve the quality of service provided to patients and other hospital staff, the pharmacy leadership team can use the data to ensure that inpatient pharmacy technician resources are being used effectively during times of day with the greatest number of incoming ACD calls. The ACD phone system helps determine whether current resources are being used most efficiently and if they are not, can help identify areas of improvement.

The pharmacy leadership team plans on using reports for pharmacy team members to monitor performance. Reports on individual agent activity capture workload; this may be used as a performance-related metric for future performance plans.

Conclusions

The inpatient pharmacy ACD phone system at EHJVAH is a promising application of available technology. The implementation of the ACD system improved accountability, efficiency, work distribution, and the allocation of resources in the inpatient pharmacy service. The ACD phone system has yielded positive performance metrics including mean speed to answer ≤ 30 seconds and abandonment rate ≤ 5% over 12 months after implementation. With time, users of the inpatient pharmacy ACD phone system will become more comfortable with the technology, thus further improving the patient health care quality.

Pharmacy call centers have been successfully implemented in outpatient and specialty pharmacy settings.¹ A centralized pharmacy call center gives patients immediate access to a pharmacist who can view their health records to answer specific questions or fulfill medication renewal requests.^2-4 Little literature exists to describe its use in an inpatient setting.

Inpatient pharmacies receive numerous calls from health care professionals and patients. Challenges related to phone calls in the inpatient pharmacy setting may include interruptions, distractions, low accountability, poor efficiency, lack of optimal resources, and staffing.⁵ An unequal distribution and lack of accountability may exist when answering phone calls for the inpatient pharmacy team, which may contribute to long hold times and call abandonment rates. Phone calls also may be directed inefficiently between clinical pharmacists (CPs) and pharmacy technicians. Team member time related to answering phone calls may not be captured or measured.

The Edward Hines, Jr. Veterans Affairs Hospital (EHJVAH) in Illinois offers primary, extended, and specialty care and is a tertiary care referral center. The facility operates 483 beds and serves 6 community-based outpatient clinics.

Implementation

A new inpatient pharmacy service phone line extension was implemented. Data used to report quality metrics were obtained from the Global Navigator (GNAV), an information system that records calls, tracks the performance of agents, and coordinates personnel scheduling. The effectiveness of the ACD system was evaluated by quality metric goals of mean speed to answer ≤ 30 seconds and mean abandonment rate ≤ 5%. This project was determined to be quality improvement and was not reviewed by the EHJVAH Institutional Review Board.

The ACD system was set up in December 2020. After a 1-month implementation period, metrics were reported to the inpatient pharmacy team and leadership. By January 2021, EHJVAH fully implemented an ACD phone system operated by inpatient pharmacy technicians and CPs. EHJVAH inpatient pharmacy includes CPs who practice without a scope of practice and board-certified pharmacy technicians in 3 shifts. The CPs and pharmacy technicians work in the central pharmacy (the main pharmacy and inpatient pharmacy vault) or are decentralized with responsibility for answering phone calls and making deliveries (pharmacy technicians).

The pharmacy leadership team decided to implement 1 phone line with 2 ACD splits. The first split was directed to pharmacy technicians and the second to CPs. The intention was to streamline calls to be directed to proper team members within the inpatient pharmacy. The CP line also was designed to back up the pharmacy technician line. These calls were equally distributed among staff based on a standard algorithm. The pharmacy greeting stated, “Thank you for contacting the inpatient pharmacy at Hines VA Hospital. For missing doses, unit stock requests, or to speak with a pharmacy technician, please press 1. For clinical questions, order verification, or to speak with a pharmacist, please press 2.” Each inpatient pharmacy team member had a unique system login.

Fourteen ACD phone stations were established in the main pharmacy and in decentralized locations for order verification. The stations were distributed across the pharmacy service to optimize workload, space, and resources.

Training and Communication

Before implementing the inpatient pharmacy ACD phone system, the CPs and pharmacy technicians received mandatory ACD training. After the training, pharmacy team members were required to sign off on the training document to indicate that they had completed the course. The pharmacy team was trained on the importance of staffing the phones continuously. As a 24-hour pharmacy service in the acute care setting, any call may be critical for patient care.

A hospital-wide memorandum was distributed via email to all unit managers and hospital staff to educate them on the new ACD phone system, which included a new phone line extension for the inpatient pharmacy. Additionally, the inpatient pharmacy team was trained on the proper way of communicating the ACD phone system process with the hospital staff. The inpatient pharmacy team was notified that there would be an educational period to explain the queue process to hospital staff. Occasionally, hospital staff believed they were speaking to an automated system and hung up before their call was answered. The inpatient pharmacy team was instructed to notify the hospital staff to stay on the line since their call would be answered in the order it was received. Once the inpatient pharmacy team received proper training and felt comfortable with the phone system, it was set up and integrated into the workflow.

Postimplementation Evaluation

Inpatient pharmacy ACD phone system data were collected for 2021. To evaluate the effectiveness of an ACD system, the pharmacy leadership team set up the following metrics and goals for inpatient CPs and inpatient pharmacy technicians for monthly call volume/abandonment rate, mean speed to answer, mean call volume by shift, and the mean abandonment rate by shift.

Discussion

Since implementing the inpatient pharmacy ACD phone system in January 2021, there have been successes and challenges. The implementation increased accountability and efficiency when answering pharmacy phone calls. An ACD uses an algorithm that ensures equitable distribution of phone calls between CPs and pharmacy technicians. Through this algorithm, the pharmacy team is held more accountable when answering incoming calls. Distributing phone calls equally allows for optimization and balances the workload. The ACD phone system also improved efficiency when answering incoming calls. By incorporating splits when a patient or health care professional calls, ACD routes the question to the appropriate staff member. As a result, CPs spend less time answering questions meant for pharmacy technicians and instead can answer clinical or order verification questions more efficiently.

ACD data also allow pharmacy leadership to assess staffing needs, depending on the call volume. Based on ACD data, the busiest time of day was 8:00 AM to 4:00 PM. Based on this information, pharmacy leadership plans to staff more appropriately to have more pharmacy technicians working during the first shift to attend to phone calls.

The mean call abandonment rate was 4.7% for both CPs and pharmacy technicians, which met the ≤ 5% goal. The highest call abandonment rate was from midnight to 8 AM, though this shift also experienced the lowest call volume. This trend may be attributed to fewer pharmacy team members available to meet the demands of the overnight shift.

Pharmacy technicians handled a higher total call volume, which may be attributed to more phone calls related to missing doses or unit stock requests compared with clinical questions or order verifications. This information may be beneficial to identify opportunities to improve pharmacy operations.

The main challenges encountered in the ACD implementation process were hardware installation and communication with hospital staff about the changes in the inpatient pharmacy phone system. To implement the new inpatient pharmacy ACD phone system, previous telephones and hardware were removed and replaced. Initially, hardware and installation delays made it difficult for the ACD phone system to operate efficiently in the early months of its implementation. The inpatient pharmacy team depends on the telecommunications system and computers for their daily activities. Delays and issues with the hardware and ACD phone system made it more difficult to provide patient care.

Communication is a continuous challenge to ensure that hospital staff are notified of the new inpatient pharmacy ACD phone number. Over time, the understanding and use of the new ACD phone system have increased dramatically, but there are still opportunities to capture any misdirected calls. Informal feedback was obtained at pharmacy huddles and 1-on-1 discussions with pharmacy staff, and the opinions were mixed. Members of the pharmacy staff expressed that the ACD phone system set up an effective way to triage phone calls. Another positive comment was that the system created a means of accountability for pharmacy phone calls. Critical feedback included challenges with triaging phone calls to appropriate pharmacists, because calls are assigned based on an algorithm, whereas clinical coverage is determined by designated unit daily assignments.

Limitations

There are potential limitations to this quality improvement project. This phone system may not apply to all inpatient hospital pharmacy settings. Potential limitations for implementation at other institutions may include but are not limited to, differing pharmacy practice models (centralized vs decentralized), implementation costs, and internal resources.

Future Goals

To improve the quality of service provided to patients and other hospital staff, the pharmacy leadership team can use the data to ensure that inpatient pharmacy technician resources are being used effectively during times of day with the greatest number of incoming ACD calls. The ACD phone system helps determine whether current resources are being used most efficiently and if they are not, can help identify areas of improvement.

The pharmacy leadership team plans on using reports for pharmacy team members to monitor performance. Reports on individual agent activity capture workload; this may be used as a performance-related metric for future performance plans.

Conclusions

The inpatient pharmacy ACD phone system at EHJVAH is a promising application of available technology. The implementation of the ACD system improved accountability, efficiency, work distribution, and the allocation of resources in the inpatient pharmacy service. The ACD phone system has yielded positive performance metrics including mean speed to answer ≤ 30 seconds and abandonment rate ≤ 5% over 12 months after implementation. With time, users of the inpatient pharmacy ACD phone system will become more comfortable with the technology, thus further improving the patient health care quality.

Pharmacy call centers have been successfully implemented in outpatient and specialty pharmacy settings.¹ A centralized pharmacy call center gives patients immediate access to a pharmacist who can view their health records to answer specific questions or fulfill medication renewal requests.^2-4 Little literature exists to describe its use in an inpatient setting.

Inpatient pharmacies receive numerous calls from health care professionals and patients. Challenges related to phone calls in the inpatient pharmacy setting may include interruptions, distractions, low accountability, poor efficiency, lack of optimal resources, and staffing.⁵ An unequal distribution and lack of accountability may exist when answering phone calls for the inpatient pharmacy team, which may contribute to long hold times and call abandonment rates. Phone calls also may be directed inefficiently between clinical pharmacists (CPs) and pharmacy technicians. Team member time related to answering phone calls may not be captured or measured.

The Edward Hines, Jr. Veterans Affairs Hospital (EHJVAH) in Illinois offers primary, extended, and specialty care and is a tertiary care referral center. The facility operates 483 beds and serves 6 community-based outpatient clinics.

Implementation

A new inpatient pharmacy service phone line extension was implemented. Data used to report quality metrics were obtained from the Global Navigator (GNAV), an information system that records calls, tracks the performance of agents, and coordinates personnel scheduling. The effectiveness of the ACD system was evaluated by quality metric goals of mean speed to answer ≤ 30 seconds and mean abandonment rate ≤ 5%. This project was determined to be quality improvement and was not reviewed by the EHJVAH Institutional Review Board.

The ACD system was set up in December 2020. After a 1-month implementation period, metrics were reported to the inpatient pharmacy team and leadership. By January 2021, EHJVAH fully implemented an ACD phone system operated by inpatient pharmacy technicians and CPs. EHJVAH inpatient pharmacy includes CPs who practice without a scope of practice and board-certified pharmacy technicians in 3 shifts. The CPs and pharmacy technicians work in the central pharmacy (the main pharmacy and inpatient pharmacy vault) or are decentralized with responsibility for answering phone calls and making deliveries (pharmacy technicians).

The pharmacy leadership team decided to implement 1 phone line with 2 ACD splits. The first split was directed to pharmacy technicians and the second to CPs. The intention was to streamline calls to be directed to proper team members within the inpatient pharmacy. The CP line also was designed to back up the pharmacy technician line. These calls were equally distributed among staff based on a standard algorithm. The pharmacy greeting stated, “Thank you for contacting the inpatient pharmacy at Hines VA Hospital. For missing doses, unit stock requests, or to speak with a pharmacy technician, please press 1. For clinical questions, order verification, or to speak with a pharmacist, please press 2.” Each inpatient pharmacy team member had a unique system login.

Fourteen ACD phone stations were established in the main pharmacy and in decentralized locations for order verification. The stations were distributed across the pharmacy service to optimize workload, space, and resources.

Training and Communication

Before implementing the inpatient pharmacy ACD phone system, the CPs and pharmacy technicians received mandatory ACD training. After the training, pharmacy team members were required to sign off on the training document to indicate that they had completed the course. The pharmacy team was trained on the importance of staffing the phones continuously. As a 24-hour pharmacy service in the acute care setting, any call may be critical for patient care.

A hospital-wide memorandum was distributed via email to all unit managers and hospital staff to educate them on the new ACD phone system, which included a new phone line extension for the inpatient pharmacy. Additionally, the inpatient pharmacy team was trained on the proper way of communicating the ACD phone system process with the hospital staff. The inpatient pharmacy team was notified that there would be an educational period to explain the queue process to hospital staff. Occasionally, hospital staff believed they were speaking to an automated system and hung up before their call was answered. The inpatient pharmacy team was instructed to notify the hospital staff to stay on the line since their call would be answered in the order it was received. Once the inpatient pharmacy team received proper training and felt comfortable with the phone system, it was set up and integrated into the workflow.

Postimplementation Evaluation

Inpatient pharmacy ACD phone system data were collected for 2021. To evaluate the effectiveness of an ACD system, the pharmacy leadership team set up the following metrics and goals for inpatient CPs and inpatient pharmacy technicians for monthly call volume/abandonment rate, mean speed to answer, mean call volume by shift, and the mean abandonment rate by shift.

Discussion

Since implementing the inpatient pharmacy ACD phone system in January 2021, there have been successes and challenges. The implementation increased accountability and efficiency when answering pharmacy phone calls. An ACD uses an algorithm that ensures equitable distribution of phone calls between CPs and pharmacy technicians. Through this algorithm, the pharmacy team is held more accountable when answering incoming calls. Distributing phone calls equally allows for optimization and balances the workload. The ACD phone system also improved efficiency when answering incoming calls. By incorporating splits when a patient or health care professional calls, ACD routes the question to the appropriate staff member. As a result, CPs spend less time answering questions meant for pharmacy technicians and instead can answer clinical or order verification questions more efficiently.

ACD data also allow pharmacy leadership to assess staffing needs, depending on the call volume. Based on ACD data, the busiest time of day was 8:00 AM to 4:00 PM. Based on this information, pharmacy leadership plans to staff more appropriately to have more pharmacy technicians working during the first shift to attend to phone calls.

The mean call abandonment rate was 4.7% for both CPs and pharmacy technicians, which met the ≤ 5% goal. The highest call abandonment rate was from midnight to 8 AM, though this shift also experienced the lowest call volume. This trend may be attributed to fewer pharmacy team members available to meet the demands of the overnight shift.

Pharmacy technicians handled a higher total call volume, which may be attributed to more phone calls related to missing doses or unit stock requests compared with clinical questions or order verifications. This information may be beneficial to identify opportunities to improve pharmacy operations.

The main challenges encountered in the ACD implementation process were hardware installation and communication with hospital staff about the changes in the inpatient pharmacy phone system. To implement the new inpatient pharmacy ACD phone system, previous telephones and hardware were removed and replaced. Initially, hardware and installation delays made it difficult for the ACD phone system to operate efficiently in the early months of its implementation. The inpatient pharmacy team depends on the telecommunications system and computers for their daily activities. Delays and issues with the hardware and ACD phone system made it more difficult to provide patient care.

Communication is a continuous challenge to ensure that hospital staff are notified of the new inpatient pharmacy ACD phone number. Over time, the understanding and use of the new ACD phone system have increased dramatically, but there are still opportunities to capture any misdirected calls. Informal feedback was obtained at pharmacy huddles and 1-on-1 discussions with pharmacy staff, and the opinions were mixed. Members of the pharmacy staff expressed that the ACD phone system set up an effective way to triage phone calls. Another positive comment was that the system created a means of accountability for pharmacy phone calls. Critical feedback included challenges with triaging phone calls to appropriate pharmacists, because calls are assigned based on an algorithm, whereas clinical coverage is determined by designated unit daily assignments.

Limitations

There are potential limitations to this quality improvement project. This phone system may not apply to all inpatient hospital pharmacy settings. Potential limitations for implementation at other institutions may include but are not limited to, differing pharmacy practice models (centralized vs decentralized), implementation costs, and internal resources.

Future Goals

To improve the quality of service provided to patients and other hospital staff, the pharmacy leadership team can use the data to ensure that inpatient pharmacy technician resources are being used effectively during times of day with the greatest number of incoming ACD calls. The ACD phone system helps determine whether current resources are being used most efficiently and if they are not, can help identify areas of improvement.

The pharmacy leadership team plans on using reports for pharmacy team members to monitor performance. Reports on individual agent activity capture workload; this may be used as a performance-related metric for future performance plans.

Conclusions

The inpatient pharmacy ACD phone system at EHJVAH is a promising application of available technology. The implementation of the ACD system improved accountability, efficiency, work distribution, and the allocation of resources in the inpatient pharmacy service. The ACD phone system has yielded positive performance metrics including mean speed to answer ≤ 30 seconds and abandonment rate ≤ 5% over 12 months after implementation. With time, users of the inpatient pharmacy ACD phone system will become more comfortable with the technology, thus further improving the patient health care quality.

Early in the COVID-19 pandemic, monoclonal antibody (Mab) therapy was the only outpatient therapy for patients with COVID-19 experiencing mild-to-moderate symptoms. The Blocking Viral Attachment and Cell Entry with SARS-CoV-2 Neutralizing Antibodies (BLAZE-1) and the REGN-COV2 (Regeneron) clinical trials found participants treated with Mab had a shorter duration of symptoms and fewer hospitalizations compared with those receiving placebo.^1,2 Mab therapy was most efficacious early in the disease course, and the initial US Food and Drug Administration (FDA) Emergency Use Authorization (EUA) of Mab therapies required use within 10 days of symptom onset.³

The impact of the COVID-19 pandemic has been felt disproportionately among marginalized racial and ethnic groups in the US. The COVID-19 Associated Hospitalization Surveillance Network found that non-Hispanic Black persons have significantly higher rates of hospitalization and death by COVID-19 compared with White persons.^4-7 However, marginalized groups are underrepresented in the receipt of therapeutic agents for COVID-19. From March 2020 through August 2021, the mean monthly Mab use among Black patients (2.8%) was lower compared with White patients (4.0%), and Black patients received Mab 22.4% less often than White patients.⁷

The Mab clinical trials BLAZE-1 and REGN-COV2 study populations consisted of > 80% White participants.^1,2 Receipt of COVID-19 outpatient treatments may not align with the disease burden in marginalized racial and ethnic groups, leading to health disparities. Although not exhaustive, reasons for these disparities include patient, health care practitioner, and systems-level issues: patient awareness, trust, and engagement with the health care system; health care practitioner awareness and advocacy to pursue COVID-19 treatment for the patient; and health care capacity to provide the medication and service.⁷

Here, we describe a novel, quality improvement initiative at the Atlanta Veterans Affairs Health Care System (AVAHCS) in Georgia that paired a proactive laboratory-based surveillance strategy to identify and engage veterans for Mab. By centralizing the surveillance and outreach process, we sought to reduce barriers to the Mab referral process and optimize access to life-saving medication.

Implementation

AVAHCS serves a diverse population of more than 129,000 (50.8% non-Hispanic Black veterans, 37.5% White veterans, and 11.7% of other races) at a main medical campus and 18 surrounding community-based outpatient clinics. From December 28, 2020, to August 31, 2021, veterans with a positive COVID-19 nasopharyngeal polymerase chain reaction (PCR) test at AVAHCS were screened daily. A central Mab team consisting of infectious disease (ID) clinical pharmacists and physicians reviewed daily lists of positive laboratory results and identified high-risk individuals for Mab eligibility, using the FDA EUA inclusion criteria. Eligible patients were called by a Mab team member to discuss Mab treatment, provide anticipatory guidance, obtain verbal consent, and schedule the infusion. Conventional referrals from non-Mab team members (eg, primary care physicians) were also accepted into the screening process and underwent the same procedures and risk prioritization strategy as those identified by the Mab team.

Clinic resources allowed for 1 to 2 patients per day to be given Mab, increasing to a maximum of 5 patients per day during the COVID-19 Delta variant surge. We followed our best clinical judgment in prioritizing patient selection, and we aligned our practice with the standards of our affiliated partner, Emory University. In circumstances where patients who were Mab-eligible outnumbered infusion availability, patients were prioritized using the Veterans Health Administration (VHA) COVID-19 (VACO) Index for 30-day COVID-19 mortality.⁸ As COVID-19 variants developed resistance to the recommended Mab infusions, bamlanivimab, bamlanivimab-etesevimab, or casirivimab-imdevimab, local protocols adapted to EUA revisions. The Mab team also adopted FDA eligibility criteria revisions as they were available.^9,10

We describe the outcomes of our centralized screening process for Mab therapy, as measured by screening, uptake, and time to receipt of Mab from screening. We also describe the demographic and clinical characteristics of Mab recipients. Clinical outcomes include postinfusion adverse events (AEs) at day 1 and day 7, emergency department (ED) visits, inpatient hospitalization, and death.

Results

The Mab team screened 2028 veterans who were COVID-19 positive between December 28, 2020, and August 31, 2021, and identified 289 veterans (14%) who met the EUA criteria. One hundred thirty-two veterans (46%) completed Mab infusion, and of the remaining 145 veterans, 124 (86%) declined treatment, and 21 (14%) veterans did not complete Mab infusion largely due to not keeping the appointment. The Mab team active surveillance strategy identified 101 of 132 infusion candidates (77%); 82% had outpatient Mab infusion.

The mean age of veterans who received Mab was 55 years (range, 29-90), and 75% of veterans were aged ≥ 65 years; most were male (84%) and 86 (65%) identified as non-Hispanic Black individuals (Table 1).

Discussion

This novel initiative to optimize access to outpatient COVID-19 treatment demonstrated how the Mab team proactively screened and reached out to eligible veterans with COVID-19 promptly. This approach removed layers in the traditional referral process that could be barriers to accessing care. More than three-quarters of patients who received Mab were identified through this strategy, and the uptake was high at 46%. Conventional passive referrals were suboptimal for identifying candidates, which was also the case at a neighboring institution.

In an Emory University study, referrals to the Mab clinic were made through a traditional, decentralized referral system and resulted in a lower uptake of Mab treatment (4.6%).¹¹ One of the key advantages of the AVAHCS program was that we were able to provide individual education about COVID-19 and counsel on the benefits and risks of therapy. Having a structured, telehealth follow-up plan provided additional reassurance and support to the patient. These personalized patient connections likely helped increase acceptance of the Mab therapy.

Our surveillance and outreach strategy had high uptake among Black patients (65%), which exceeded the proportion of AVAHCS Black veterans (54%).¹² In the Emory study, just 30% of the participants were Black patients.¹¹ In a study of bamlanivimab use in Chicago, Black individuals represented just 11% of the study population. White patients were more likely to receive bamlanivimab compared with others races, and the likelihood of receiving bamlanivimab was significantly worse for Black patients (odds ratio, 0.28) compared with White patients.¹³ These studies highlight the disparity in COVID-19 outpatient treatment that does not reflect the racial and minority group representation of the community at large.

Limitations

The VHA medication allocation system at times created a significant mismatch in supply and demand, which significantly limited the AVAHCS Mab program. VHA facilities nationwide with Mab programs received discrete allocations through the US Department of Health and Human Services via VHA pharmacy benefits management services. Despite our large catchment, AVAHCS was allocated 6 or fewer doses of Mab per week during the evaluated period.

Without formal national guidance in the early period of Mab, the AVAHCS Mab team conferred with Emory University Mab clinicians as well as at other VHA facilities in the country to develop an optimal approach to resource allocation. The Mab team considered all EUA criteria to be as inclusive as possible. However, during times of high demand, our utilitarian approach tried to identify the highest-risk patients who would benefit the most from Mab. The VACO index was validated in early 2021, which facilitated decision making when demand was greater than supply. One limitation of the VACO index is its exclusion of several original Mab EUA criteria, including weight, hypertension, and nonmalignancy-related immunosuppression, into its algorithm.^3,8

Conclusions

Through proactive screening and direct outreach to patients, the AVAHCS was able to achieve timely administration of Mab infusion that was well within the initial EUA time frame of 10 days and comparable with the time frame in the REGN-COV2 and BLAZE-1 trials. Improving access to resources by changing the referral structure helped engage veterans who may have otherwise missed the time frame for Mab therapy. The experience of the Mab infusion program at the AVAHCS provided valuable insight into how a health care system could effectively screen a large population and distribute the limited resource of Mab therapy in a timely and proportionate fashion among its represented demographic groups.

Acknowledgments

The authors acknowledge the Veterans Health Administration VISN 7 Clinical Resource Hub and Tele Primary Care group for their support.

Early in the COVID-19 pandemic, monoclonal antibody (Mab) therapy was the only outpatient therapy for patients with COVID-19 experiencing mild-to-moderate symptoms. The Blocking Viral Attachment and Cell Entry with SARS-CoV-2 Neutralizing Antibodies (BLAZE-1) and the REGN-COV2 (Regeneron) clinical trials found participants treated with Mab had a shorter duration of symptoms and fewer hospitalizations compared with those receiving placebo.^1,2 Mab therapy was most efficacious early in the disease course, and the initial US Food and Drug Administration (FDA) Emergency Use Authorization (EUA) of Mab therapies required use within 10 days of symptom onset.³

The impact of the COVID-19 pandemic has been felt disproportionately among marginalized racial and ethnic groups in the US. The COVID-19 Associated Hospitalization Surveillance Network found that non-Hispanic Black persons have significantly higher rates of hospitalization and death by COVID-19 compared with White persons.^4-7 However, marginalized groups are underrepresented in the receipt of therapeutic agents for COVID-19. From March 2020 through August 2021, the mean monthly Mab use among Black patients (2.8%) was lower compared with White patients (4.0%), and Black patients received Mab 22.4% less often than White patients.⁷

The Mab clinical trials BLAZE-1 and REGN-COV2 study populations consisted of > 80% White participants.^1,2 Receipt of COVID-19 outpatient treatments may not align with the disease burden in marginalized racial and ethnic groups, leading to health disparities. Although not exhaustive, reasons for these disparities include patient, health care practitioner, and systems-level issues: patient awareness, trust, and engagement with the health care system; health care practitioner awareness and advocacy to pursue COVID-19 treatment for the patient; and health care capacity to provide the medication and service.⁷

Here, we describe a novel, quality improvement initiative at the Atlanta Veterans Affairs Health Care System (AVAHCS) in Georgia that paired a proactive laboratory-based surveillance strategy to identify and engage veterans for Mab. By centralizing the surveillance and outreach process, we sought to reduce barriers to the Mab referral process and optimize access to life-saving medication.

Implementation

AVAHCS serves a diverse population of more than 129,000 (50.8% non-Hispanic Black veterans, 37.5% White veterans, and 11.7% of other races) at a main medical campus and 18 surrounding community-based outpatient clinics. From December 28, 2020, to August 31, 2021, veterans with a positive COVID-19 nasopharyngeal polymerase chain reaction (PCR) test at AVAHCS were screened daily. A central Mab team consisting of infectious disease (ID) clinical pharmacists and physicians reviewed daily lists of positive laboratory results and identified high-risk individuals for Mab eligibility, using the FDA EUA inclusion criteria. Eligible patients were called by a Mab team member to discuss Mab treatment, provide anticipatory guidance, obtain verbal consent, and schedule the infusion. Conventional referrals from non-Mab team members (eg, primary care physicians) were also accepted into the screening process and underwent the same procedures and risk prioritization strategy as those identified by the Mab team.

Clinic resources allowed for 1 to 2 patients per day to be given Mab, increasing to a maximum of 5 patients per day during the COVID-19 Delta variant surge. We followed our best clinical judgment in prioritizing patient selection, and we aligned our practice with the standards of our affiliated partner, Emory University. In circumstances where patients who were Mab-eligible outnumbered infusion availability, patients were prioritized using the Veterans Health Administration (VHA) COVID-19 (VACO) Index for 30-day COVID-19 mortality.⁸ As COVID-19 variants developed resistance to the recommended Mab infusions, bamlanivimab, bamlanivimab-etesevimab, or casirivimab-imdevimab, local protocols adapted to EUA revisions. The Mab team also adopted FDA eligibility criteria revisions as they were available.^9,10

We describe the outcomes of our centralized screening process for Mab therapy, as measured by screening, uptake, and time to receipt of Mab from screening. We also describe the demographic and clinical characteristics of Mab recipients. Clinical outcomes include postinfusion adverse events (AEs) at day 1 and day 7, emergency department (ED) visits, inpatient hospitalization, and death.

Results

The Mab team screened 2028 veterans who were COVID-19 positive between December 28, 2020, and August 31, 2021, and identified 289 veterans (14%) who met the EUA criteria. One hundred thirty-two veterans (46%) completed Mab infusion, and of the remaining 145 veterans, 124 (86%) declined treatment, and 21 (14%) veterans did not complete Mab infusion largely due to not keeping the appointment. The Mab team active surveillance strategy identified 101 of 132 infusion candidates (77%); 82% had outpatient Mab infusion.

The mean age of veterans who received Mab was 55 years (range, 29-90), and 75% of veterans were aged ≥ 65 years; most were male (84%) and 86 (65%) identified as non-Hispanic Black individuals (Table 1).

Discussion

This novel initiative to optimize access to outpatient COVID-19 treatment demonstrated how the Mab team proactively screened and reached out to eligible veterans with COVID-19 promptly. This approach removed layers in the traditional referral process that could be barriers to accessing care. More than three-quarters of patients who received Mab were identified through this strategy, and the uptake was high at 46%. Conventional passive referrals were suboptimal for identifying candidates, which was also the case at a neighboring institution.

In an Emory University study, referrals to the Mab clinic were made through a traditional, decentralized referral system and resulted in a lower uptake of Mab treatment (4.6%).¹¹ One of the key advantages of the AVAHCS program was that we were able to provide individual education about COVID-19 and counsel on the benefits and risks of therapy. Having a structured, telehealth follow-up plan provided additional reassurance and support to the patient. These personalized patient connections likely helped increase acceptance of the Mab therapy.

Our surveillance and outreach strategy had high uptake among Black patients (65%), which exceeded the proportion of AVAHCS Black veterans (54%).¹² In the Emory study, just 30% of the participants were Black patients.¹¹ In a study of bamlanivimab use in Chicago, Black individuals represented just 11% of the study population. White patients were more likely to receive bamlanivimab compared with others races, and the likelihood of receiving bamlanivimab was significantly worse for Black patients (odds ratio, 0.28) compared with White patients.¹³ These studies highlight the disparity in COVID-19 outpatient treatment that does not reflect the racial and minority group representation of the community at large.

Limitations

The VHA medication allocation system at times created a significant mismatch in supply and demand, which significantly limited the AVAHCS Mab program. VHA facilities nationwide with Mab programs received discrete allocations through the US Department of Health and Human Services via VHA pharmacy benefits management services. Despite our large catchment, AVAHCS was allocated 6 or fewer doses of Mab per week during the evaluated period.

Without formal national guidance in the early period of Mab, the AVAHCS Mab team conferred with Emory University Mab clinicians as well as at other VHA facilities in the country to develop an optimal approach to resource allocation. The Mab team considered all EUA criteria to be as inclusive as possible. However, during times of high demand, our utilitarian approach tried to identify the highest-risk patients who would benefit the most from Mab. The VACO index was validated in early 2021, which facilitated decision making when demand was greater than supply. One limitation of the VACO index is its exclusion of several original Mab EUA criteria, including weight, hypertension, and nonmalignancy-related immunosuppression, into its algorithm.^3,8

Conclusions

Through proactive screening and direct outreach to patients, the AVAHCS was able to achieve timely administration of Mab infusion that was well within the initial EUA time frame of 10 days and comparable with the time frame in the REGN-COV2 and BLAZE-1 trials. Improving access to resources by changing the referral structure helped engage veterans who may have otherwise missed the time frame for Mab therapy. The experience of the Mab infusion program at the AVAHCS provided valuable insight into how a health care system could effectively screen a large population and distribute the limited resource of Mab therapy in a timely and proportionate fashion among its represented demographic groups.

Acknowledgments

The authors acknowledge the Veterans Health Administration VISN 7 Clinical Resource Hub and Tele Primary Care group for their support.

Early in the COVID-19 pandemic, monoclonal antibody (Mab) therapy was the only outpatient therapy for patients with COVID-19 experiencing mild-to-moderate symptoms. The Blocking Viral Attachment and Cell Entry with SARS-CoV-2 Neutralizing Antibodies (BLAZE-1) and the REGN-COV2 (Regeneron) clinical trials found participants treated with Mab had a shorter duration of symptoms and fewer hospitalizations compared with those receiving placebo.^1,2 Mab therapy was most efficacious early in the disease course, and the initial US Food and Drug Administration (FDA) Emergency Use Authorization (EUA) of Mab therapies required use within 10 days of symptom onset.³

The impact of the COVID-19 pandemic has been felt disproportionately among marginalized racial and ethnic groups in the US. The COVID-19 Associated Hospitalization Surveillance Network found that non-Hispanic Black persons have significantly higher rates of hospitalization and death by COVID-19 compared with White persons.^4-7 However, marginalized groups are underrepresented in the receipt of therapeutic agents for COVID-19. From March 2020 through August 2021, the mean monthly Mab use among Black patients (2.8%) was lower compared with White patients (4.0%), and Black patients received Mab 22.4% less often than White patients.⁷

The Mab clinical trials BLAZE-1 and REGN-COV2 study populations consisted of > 80% White participants.^1,2 Receipt of COVID-19 outpatient treatments may not align with the disease burden in marginalized racial and ethnic groups, leading to health disparities. Although not exhaustive, reasons for these disparities include patient, health care practitioner, and systems-level issues: patient awareness, trust, and engagement with the health care system; health care practitioner awareness and advocacy to pursue COVID-19 treatment for the patient; and health care capacity to provide the medication and service.⁷

Here, we describe a novel, quality improvement initiative at the Atlanta Veterans Affairs Health Care System (AVAHCS) in Georgia that paired a proactive laboratory-based surveillance strategy to identify and engage veterans for Mab. By centralizing the surveillance and outreach process, we sought to reduce barriers to the Mab referral process and optimize access to life-saving medication.

Implementation

AVAHCS serves a diverse population of more than 129,000 (50.8% non-Hispanic Black veterans, 37.5% White veterans, and 11.7% of other races) at a main medical campus and 18 surrounding community-based outpatient clinics. From December 28, 2020, to August 31, 2021, veterans with a positive COVID-19 nasopharyngeal polymerase chain reaction (PCR) test at AVAHCS were screened daily. A central Mab team consisting of infectious disease (ID) clinical pharmacists and physicians reviewed daily lists of positive laboratory results and identified high-risk individuals for Mab eligibility, using the FDA EUA inclusion criteria. Eligible patients were called by a Mab team member to discuss Mab treatment, provide anticipatory guidance, obtain verbal consent, and schedule the infusion. Conventional referrals from non-Mab team members (eg, primary care physicians) were also accepted into the screening process and underwent the same procedures and risk prioritization strategy as those identified by the Mab team.

Clinic resources allowed for 1 to 2 patients per day to be given Mab, increasing to a maximum of 5 patients per day during the COVID-19 Delta variant surge. We followed our best clinical judgment in prioritizing patient selection, and we aligned our practice with the standards of our affiliated partner, Emory University. In circumstances where patients who were Mab-eligible outnumbered infusion availability, patients were prioritized using the Veterans Health Administration (VHA) COVID-19 (VACO) Index for 30-day COVID-19 mortality.⁸ As COVID-19 variants developed resistance to the recommended Mab infusions, bamlanivimab, bamlanivimab-etesevimab, or casirivimab-imdevimab, local protocols adapted to EUA revisions. The Mab team also adopted FDA eligibility criteria revisions as they were available.^9,10

We describe the outcomes of our centralized screening process for Mab therapy, as measured by screening, uptake, and time to receipt of Mab from screening. We also describe the demographic and clinical characteristics of Mab recipients. Clinical outcomes include postinfusion adverse events (AEs) at day 1 and day 7, emergency department (ED) visits, inpatient hospitalization, and death.

Results

The Mab team screened 2028 veterans who were COVID-19 positive between December 28, 2020, and August 31, 2021, and identified 289 veterans (14%) who met the EUA criteria. One hundred thirty-two veterans (46%) completed Mab infusion, and of the remaining 145 veterans, 124 (86%) declined treatment, and 21 (14%) veterans did not complete Mab infusion largely due to not keeping the appointment. The Mab team active surveillance strategy identified 101 of 132 infusion candidates (77%); 82% had outpatient Mab infusion.

The mean age of veterans who received Mab was 55 years (range, 29-90), and 75% of veterans were aged ≥ 65 years; most were male (84%) and 86 (65%) identified as non-Hispanic Black individuals (Table 1).

Discussion

This novel initiative to optimize access to outpatient COVID-19 treatment demonstrated how the Mab team proactively screened and reached out to eligible veterans with COVID-19 promptly. This approach removed layers in the traditional referral process that could be barriers to accessing care. More than three-quarters of patients who received Mab were identified through this strategy, and the uptake was high at 46%. Conventional passive referrals were suboptimal for identifying candidates, which was also the case at a neighboring institution.

In an Emory University study, referrals to the Mab clinic were made through a traditional, decentralized referral system and resulted in a lower uptake of Mab treatment (4.6%).¹¹ One of the key advantages of the AVAHCS program was that we were able to provide individual education about COVID-19 and counsel on the benefits and risks of therapy. Having a structured, telehealth follow-up plan provided additional reassurance and support to the patient. These personalized patient connections likely helped increase acceptance of the Mab therapy.

Our surveillance and outreach strategy had high uptake among Black patients (65%), which exceeded the proportion of AVAHCS Black veterans (54%).¹² In the Emory study, just 30% of the participants were Black patients.¹¹ In a study of bamlanivimab use in Chicago, Black individuals represented just 11% of the study population. White patients were more likely to receive bamlanivimab compared with others races, and the likelihood of receiving bamlanivimab was significantly worse for Black patients (odds ratio, 0.28) compared with White patients.¹³ These studies highlight the disparity in COVID-19 outpatient treatment that does not reflect the racial and minority group representation of the community at large.

Limitations

The VHA medication allocation system at times created a significant mismatch in supply and demand, which significantly limited the AVAHCS Mab program. VHA facilities nationwide with Mab programs received discrete allocations through the US Department of Health and Human Services via VHA pharmacy benefits management services. Despite our large catchment, AVAHCS was allocated 6 or fewer doses of Mab per week during the evaluated period.

Without formal national guidance in the early period of Mab, the AVAHCS Mab team conferred with Emory University Mab clinicians as well as at other VHA facilities in the country to develop an optimal approach to resource allocation. The Mab team considered all EUA criteria to be as inclusive as possible. However, during times of high demand, our utilitarian approach tried to identify the highest-risk patients who would benefit the most from Mab. The VACO index was validated in early 2021, which facilitated decision making when demand was greater than supply. One limitation of the VACO index is its exclusion of several original Mab EUA criteria, including weight, hypertension, and nonmalignancy-related immunosuppression, into its algorithm.^3,8

Conclusions

Through proactive screening and direct outreach to patients, the AVAHCS was able to achieve timely administration of Mab infusion that was well within the initial EUA time frame of 10 days and comparable with the time frame in the REGN-COV2 and BLAZE-1 trials. Improving access to resources by changing the referral structure helped engage veterans who may have otherwise missed the time frame for Mab therapy. The experience of the Mab infusion program at the AVAHCS provided valuable insight into how a health care system could effectively screen a large population and distribute the limited resource of Mab therapy in a timely and proportionate fashion among its represented demographic groups.

Acknowledgments

The authors acknowledge the Veterans Health Administration VISN 7 Clinical Resource Hub and Tele Primary Care group for their support.