The Journal of Family Practice

With an estimated lifetime prevalence of 17% to 30%,¹ dizziness is a relatively common clinical symptom, but the underlying cause can be difficult to diagnose. That’s because patients’ descriptions of dizziness are often imprecise, and this symptom is associated with a wide range of conditions. A careful history and physical examination are key to diagnosis, as is an understanding of the mechanisms of dizziness.

This article covers the range of diagnoses that should be considered when a patient presents with dizziness, and provides insight regarding features of the patient’s history that can better elucidate the specific etiology.

What do patients mean when they say, “I feel dizzy”?

“Dizziness” is a vague term, and patients who report dizziness should be asked to further describe the sensation. Patients may use the word dizziness in an attempt to describe many sensations, including faintness, giddiness, light-headedness, or unsteadiness.² In 1972, Drachman and Hart proposed a classification system for dizziness that describes 4 categories—presyncope, vertigo, disequilibrium, and atypical (TABLE 1).³ These classifications are still commonly used today, and the discussion that follows describes potential causes of dizziness in each of these 4 categories. A stepwise approach for evaluating a patient who reports dizziness can be found in the ALGORITHM.^3-6

Syncopal-related dizziness can have a cardiovascular cause

Presyncope is a feeling of impending loss of consciousness that’s sometimes accompanied by generalized muscle weakness and/or partial vision loss. Taking a careful history regarding the events surrounding the episode should distinguish this type of dizziness, and doing so is essential because most of the underlying pathogenesis involves the cardiovascular system and requires specific interventions.

Dysrhythmias can cause syncope and may or may not be accompanied by a feeling of palpitations. Diagnosis is made by electrocardiogram (EKG) followed by the use of a Holter monitor.

Vasovagal syncope is caused by a sudden slowing of the pulse that’s the result of stimulation of the vagal nerve. It can occur from direct stimulation of the nerve from palpation (or strangulation), or from an intense autonomic discharge, as when people are frightened or confronted with something upsetting (eg, the sight of blood.)

Orthostatic hypotension results from a change in body position in which either autonomic mechanisms cannot maintain venous tone, causing a sudden drop in blood pressure, or in which the heart cannot compensate by speeding up, as when a patient is taking a beta-adrenergic antagonist or has first-degree heart block. It can also result from hypovolemia.

Measuring the patient’s blood pressure in the recumbent, seated, and standing positions can verify the diagnosis if an episode occurred soon before the examination. This kind of dizziness can be treated by instructing the patient to rise slowly, or by making appropriate medication adjustments. If conservative measures fail, medications such as midodrine or droxidopa can be tried.⁷

Hypoglycemia, hypoxia, or hyperventilation can also precipitate syncopal symptoms. Taking a careful history to assess for the presence of seizure-related features such as tonic/clonic movements or loss of bowel and bladder control can be helpful in distinguishing this form of dizziness.

Vertigo can have a central or peripheral cause

Vertigo is dizziness that is characterized by the sensation of spinning. The presence of vertigo implies disease of the inner ear or central nervous system. The “wiring diagram” of the vestibulo-ocular reflex is fairly straightforward, but sorting out the symptoms that arise from lesions within the system can be a diagnostic challenge. Vertigo has classically been divided into causes that are central (originating in the central nervous system) or peripheral (originating in the peripheral nervous system).

The HINTS (Head Impulse, Nystagmus, and Test of Skew) protocol is a group of 3 tests that can be used to differentiate central from peripheral vertigo (TABLE 2).^8,9 To perform the head impulse test, the examiner asks the patient to focus his gaze on a target and then rapidly turns the patient’s head to the side, watching the eyes for any corrective movements.¹⁰ When the eyes make a corrective saccade, the test is considered to be positive for a peripheral lesion.

Horizontal nystagmus is assessed by having the patient look in the direction of the fast phase of the nystagmus. If the nystagmus increases in intensity, then the test is considered positive for a peripheral lesion.

A careful description of the circumstances surrounding the dizziness episode can help identify underlying conditions such as orthostasis, hypoglycemia, or hyperventilation.

Vertigo can have many possible causes

Finally, the “test of skew” is performed by again having the patient fixate on the examiner’s nose. Each eye is tested by being covered, and then uncovered. If the uncovered eye has to move to refocus on the examiner’s nose, then the test is positive for a central lesion. A positive head impulse, positive horizontal nystagmus, and negative test of skew is 100% sensitive and 96% specific for a peripheral lesion.¹¹

Benign paroxysmal positional vertigo (BPPV) is vertigo that is triggered by movement of the head. It occurs when otoconia that are normally embedded in gel in the utricle become dislodged and migrate into the 3 fluid-filled semicircular canals, where they interfere with the normal fluid movement these canals use to sense head motion, causing the inner ear to send false signals to the brain.¹²

Diagnosis is confirmed by performing the Dix-Hallpike maneuver to elicit nystagmus. The patient is moved from a seated to a supine position with her head turned 45 degrees to the right and held for 30 seconds. For a demonstration of the Dix-Hallpike maneuver, see https://youtu.be/8RYB2QlO1N4. The Dix-Hallpike maneuver is also the first step of a treatment for BBPV known as the Epley maneuver. (See “The Epley maneuver: A procedure for treating BPPV”.^13,14)

Labyrinthitis—inflammation of the inner ear that can cause vertigo—is suggested by an acute, non-recurrent episode of dizziness that is often preceded by an upper respiratory infection. If the external canal is extremely painful and/or develops a vesicular rash, the patient might have herpes zoster of the geniculate ganglion (Ramsay Hunt syndrome type 2).

Dizziness related to presyncope often involves a cardiovascular pathology, such as a dysrhythmia or orthostatic hypotension.

Vertigo can have many possible causes

Vestibular migraine and Meniere’s disease. When a patient who has a history of migraines experiences symptoms of vertigo, vestibular migraine should be suspected, and treatment should focus on migraine therapy rather than vestibular therapy.¹⁵

Symptoms of Meniere’s disease and vestibular migraine can overlap.¹⁶ The current definition of Meniere’s disease requires ≥2 definitive episodes of vertigo with hearing loss plus tinnitus and/or aural symptoms.¹⁷ Thirty percent of vertigo episodes in patients with Meniere's disease can be attributed to BPPV.¹⁸

Acoustic neuroma. In addition to vertigo, acoustic neuroma is often associated with gradual hearing loss, tinnitus, and facial numbness (from compression of cranial nerve V preoperatively) or facial weakness (from compression of cranial nerve VII postoperatively). Unilateral hearing loss should prompt evaluation with magnetic resonance imaging.

“Acoustic neuroma” is a misnomer. The lesion arises from the vestibular (not the acoustic) portion of the 8th cranial nerve, and isn’t a neuroma; it is a schwannoma.¹⁹ Although it actually arises peripherally within the vestibular canal, it typically expands centrally and compresses other nerves centrally, which can make the clinical diagnosis more challenging if one were using the classical schema of differentiating between peripheral and central causes of vertigo.

Age-related vestibular loss occurs when the aging process causes deterioration of most of the components of the vestibulo-ocular reflex, resulting in dizziness and vertigo. Usually, the cerebral override mechanisms can compensate for the degeneration.

Other causes of vertigo include cerebellar infarction (3% of patients with vertigo),²⁰ sound-induced vertigo (Tullio phenomenon),²¹ obstructive sleep apnea,²² and systemic sclerosis.²³ Diabetes can cause a reduction in vestibular sensitivity that is evidenced by an increased reliance on visual stimuli to resolve vestibulo-visual conflict.²⁴

Disequilibrium

Disequilibrium is predominantly a loss of balance. Patients with disequilibrium have the feeling that they are about to fall, specifically without the sensation of spinning. They may appear to sway, and will reach out for something to support them. Disequilibrium can be a component of vertigo, or it may suggest a more specific diagnosis, such as ataxia, which is a lack of coordination when walking.

Atypical causes of dizziness

A positive head impulse test is highly suggestive of a peripheral lesion.

“Light-headedness” may have an element of euphoria or may be indistinguishable from the early part of a syncopal episode. Because other causes of light-headedness can be difficult to distinguish from presyncope, it is important to consider syncope in the differential diagnosis.

The differential of light-headedness can also include panic attack, early hyperventilation, and toxin exposure (such as diphenylarsinic acid,²⁵ pregabalin,²⁶ or paint thinner²⁷).

CORRESPONDENCE
Shannon Paul Starr, MD, Louisiana State University Health Sciences Center, 200 W. Esplanade #412, Kenner, LA 70065; sstarr@lsuhsc.edu.

With an estimated lifetime prevalence of 17% to 30%,¹ dizziness is a relatively common clinical symptom, but the underlying cause can be difficult to diagnose. That’s because patients’ descriptions of dizziness are often imprecise, and this symptom is associated with a wide range of conditions. A careful history and physical examination are key to diagnosis, as is an understanding of the mechanisms of dizziness.

This article covers the range of diagnoses that should be considered when a patient presents with dizziness, and provides insight regarding features of the patient’s history that can better elucidate the specific etiology.

What do patients mean when they say, “I feel dizzy”?

“Dizziness” is a vague term, and patients who report dizziness should be asked to further describe the sensation. Patients may use the word dizziness in an attempt to describe many sensations, including faintness, giddiness, light-headedness, or unsteadiness.² In 1972, Drachman and Hart proposed a classification system for dizziness that describes 4 categories—presyncope, vertigo, disequilibrium, and atypical (TABLE 1).³ These classifications are still commonly used today, and the discussion that follows describes potential causes of dizziness in each of these 4 categories. A stepwise approach for evaluating a patient who reports dizziness can be found in the ALGORITHM.^3-6

Syncopal-related dizziness can have a cardiovascular cause

Presyncope is a feeling of impending loss of consciousness that’s sometimes accompanied by generalized muscle weakness and/or partial vision loss. Taking a careful history regarding the events surrounding the episode should distinguish this type of dizziness, and doing so is essential because most of the underlying pathogenesis involves the cardiovascular system and requires specific interventions.

Dysrhythmias can cause syncope and may or may not be accompanied by a feeling of palpitations. Diagnosis is made by electrocardiogram (EKG) followed by the use of a Holter monitor.

Vasovagal syncope is caused by a sudden slowing of the pulse that’s the result of stimulation of the vagal nerve. It can occur from direct stimulation of the nerve from palpation (or strangulation), or from an intense autonomic discharge, as when people are frightened or confronted with something upsetting (eg, the sight of blood.)

Orthostatic hypotension results from a change in body position in which either autonomic mechanisms cannot maintain venous tone, causing a sudden drop in blood pressure, or in which the heart cannot compensate by speeding up, as when a patient is taking a beta-adrenergic antagonist or has first-degree heart block. It can also result from hypovolemia.

Measuring the patient’s blood pressure in the recumbent, seated, and standing positions can verify the diagnosis if an episode occurred soon before the examination. This kind of dizziness can be treated by instructing the patient to rise slowly, or by making appropriate medication adjustments. If conservative measures fail, medications such as midodrine or droxidopa can be tried.⁷

Hypoglycemia, hypoxia, or hyperventilation can also precipitate syncopal symptoms. Taking a careful history to assess for the presence of seizure-related features such as tonic/clonic movements or loss of bowel and bladder control can be helpful in distinguishing this form of dizziness.

Vertigo can have a central or peripheral cause

Vertigo is dizziness that is characterized by the sensation of spinning. The presence of vertigo implies disease of the inner ear or central nervous system. The “wiring diagram” of the vestibulo-ocular reflex is fairly straightforward, but sorting out the symptoms that arise from lesions within the system can be a diagnostic challenge. Vertigo has classically been divided into causes that are central (originating in the central nervous system) or peripheral (originating in the peripheral nervous system).

The HINTS (Head Impulse, Nystagmus, and Test of Skew) protocol is a group of 3 tests that can be used to differentiate central from peripheral vertigo (TABLE 2).^8,9 To perform the head impulse test, the examiner asks the patient to focus his gaze on a target and then rapidly turns the patient’s head to the side, watching the eyes for any corrective movements.¹⁰ When the eyes make a corrective saccade, the test is considered to be positive for a peripheral lesion.

Horizontal nystagmus is assessed by having the patient look in the direction of the fast phase of the nystagmus. If the nystagmus increases in intensity, then the test is considered positive for a peripheral lesion.

A careful description of the circumstances surrounding the dizziness episode can help identify underlying conditions such as orthostasis, hypoglycemia, or hyperventilation.

Vertigo can have many possible causes

Finally, the “test of skew” is performed by again having the patient fixate on the examiner’s nose. Each eye is tested by being covered, and then uncovered. If the uncovered eye has to move to refocus on the examiner’s nose, then the test is positive for a central lesion. A positive head impulse, positive horizontal nystagmus, and negative test of skew is 100% sensitive and 96% specific for a peripheral lesion.¹¹

Benign paroxysmal positional vertigo (BPPV) is vertigo that is triggered by movement of the head. It occurs when otoconia that are normally embedded in gel in the utricle become dislodged and migrate into the 3 fluid-filled semicircular canals, where they interfere with the normal fluid movement these canals use to sense head motion, causing the inner ear to send false signals to the brain.¹²

Diagnosis is confirmed by performing the Dix-Hallpike maneuver to elicit nystagmus. The patient is moved from a seated to a supine position with her head turned 45 degrees to the right and held for 30 seconds. For a demonstration of the Dix-Hallpike maneuver, see https://youtu.be/8RYB2QlO1N4. The Dix-Hallpike maneuver is also the first step of a treatment for BBPV known as the Epley maneuver. (See “The Epley maneuver: A procedure for treating BPPV”.^13,14)

Labyrinthitis—inflammation of the inner ear that can cause vertigo—is suggested by an acute, non-recurrent episode of dizziness that is often preceded by an upper respiratory infection. If the external canal is extremely painful and/or develops a vesicular rash, the patient might have herpes zoster of the geniculate ganglion (Ramsay Hunt syndrome type 2).

Dizziness related to presyncope often involves a cardiovascular pathology, such as a dysrhythmia or orthostatic hypotension.

Vertigo can have many possible causes

Vestibular migraine and Meniere’s disease. When a patient who has a history of migraines experiences symptoms of vertigo, vestibular migraine should be suspected, and treatment should focus on migraine therapy rather than vestibular therapy.¹⁵

Symptoms of Meniere’s disease and vestibular migraine can overlap.¹⁶ The current definition of Meniere’s disease requires ≥2 definitive episodes of vertigo with hearing loss plus tinnitus and/or aural symptoms.¹⁷ Thirty percent of vertigo episodes in patients with Meniere's disease can be attributed to BPPV.¹⁸

Acoustic neuroma. In addition to vertigo, acoustic neuroma is often associated with gradual hearing loss, tinnitus, and facial numbness (from compression of cranial nerve V preoperatively) or facial weakness (from compression of cranial nerve VII postoperatively). Unilateral hearing loss should prompt evaluation with magnetic resonance imaging.

“Acoustic neuroma” is a misnomer. The lesion arises from the vestibular (not the acoustic) portion of the 8th cranial nerve, and isn’t a neuroma; it is a schwannoma.¹⁹ Although it actually arises peripherally within the vestibular canal, it typically expands centrally and compresses other nerves centrally, which can make the clinical diagnosis more challenging if one were using the classical schema of differentiating between peripheral and central causes of vertigo.

Age-related vestibular loss occurs when the aging process causes deterioration of most of the components of the vestibulo-ocular reflex, resulting in dizziness and vertigo. Usually, the cerebral override mechanisms can compensate for the degeneration.

Other causes of vertigo include cerebellar infarction (3% of patients with vertigo),²⁰ sound-induced vertigo (Tullio phenomenon),²¹ obstructive sleep apnea,²² and systemic sclerosis.²³ Diabetes can cause a reduction in vestibular sensitivity that is evidenced by an increased reliance on visual stimuli to resolve vestibulo-visual conflict.²⁴

Disequilibrium

Disequilibrium is predominantly a loss of balance. Patients with disequilibrium have the feeling that they are about to fall, specifically without the sensation of spinning. They may appear to sway, and will reach out for something to support them. Disequilibrium can be a component of vertigo, or it may suggest a more specific diagnosis, such as ataxia, which is a lack of coordination when walking.

Atypical causes of dizziness

A positive head impulse test is highly suggestive of a peripheral lesion.

“Light-headedness” may have an element of euphoria or may be indistinguishable from the early part of a syncopal episode. Because other causes of light-headedness can be difficult to distinguish from presyncope, it is important to consider syncope in the differential diagnosis.

The differential of light-headedness can also include panic attack, early hyperventilation, and toxin exposure (such as diphenylarsinic acid,²⁵ pregabalin,²⁶ or paint thinner²⁷).

CORRESPONDENCE
Shannon Paul Starr, MD, Louisiana State University Health Sciences Center, 200 W. Esplanade #412, Kenner, LA 70065; sstarr@lsuhsc.edu.

With an estimated lifetime prevalence of 17% to 30%,¹ dizziness is a relatively common clinical symptom, but the underlying cause can be difficult to diagnose. That’s because patients’ descriptions of dizziness are often imprecise, and this symptom is associated with a wide range of conditions. A careful history and physical examination are key to diagnosis, as is an understanding of the mechanisms of dizziness.

This article covers the range of diagnoses that should be considered when a patient presents with dizziness, and provides insight regarding features of the patient’s history that can better elucidate the specific etiology.

What do patients mean when they say, “I feel dizzy”?

“Dizziness” is a vague term, and patients who report dizziness should be asked to further describe the sensation. Patients may use the word dizziness in an attempt to describe many sensations, including faintness, giddiness, light-headedness, or unsteadiness.² In 1972, Drachman and Hart proposed a classification system for dizziness that describes 4 categories—presyncope, vertigo, disequilibrium, and atypical (TABLE 1).³ These classifications are still commonly used today, and the discussion that follows describes potential causes of dizziness in each of these 4 categories. A stepwise approach for evaluating a patient who reports dizziness can be found in the ALGORITHM.^3-6

Syncopal-related dizziness can have a cardiovascular cause

Presyncope is a feeling of impending loss of consciousness that’s sometimes accompanied by generalized muscle weakness and/or partial vision loss. Taking a careful history regarding the events surrounding the episode should distinguish this type of dizziness, and doing so is essential because most of the underlying pathogenesis involves the cardiovascular system and requires specific interventions.

Dysrhythmias can cause syncope and may or may not be accompanied by a feeling of palpitations. Diagnosis is made by electrocardiogram (EKG) followed by the use of a Holter monitor.

Vasovagal syncope is caused by a sudden slowing of the pulse that’s the result of stimulation of the vagal nerve. It can occur from direct stimulation of the nerve from palpation (or strangulation), or from an intense autonomic discharge, as when people are frightened or confronted with something upsetting (eg, the sight of blood.)

Orthostatic hypotension results from a change in body position in which either autonomic mechanisms cannot maintain venous tone, causing a sudden drop in blood pressure, or in which the heart cannot compensate by speeding up, as when a patient is taking a beta-adrenergic antagonist or has first-degree heart block. It can also result from hypovolemia.

Measuring the patient’s blood pressure in the recumbent, seated, and standing positions can verify the diagnosis if an episode occurred soon before the examination. This kind of dizziness can be treated by instructing the patient to rise slowly, or by making appropriate medication adjustments. If conservative measures fail, medications such as midodrine or droxidopa can be tried.⁷

Hypoglycemia, hypoxia, or hyperventilation can also precipitate syncopal symptoms. Taking a careful history to assess for the presence of seizure-related features such as tonic/clonic movements or loss of bowel and bladder control can be helpful in distinguishing this form of dizziness.

Vertigo can have a central or peripheral cause

Vertigo is dizziness that is characterized by the sensation of spinning. The presence of vertigo implies disease of the inner ear or central nervous system. The “wiring diagram” of the vestibulo-ocular reflex is fairly straightforward, but sorting out the symptoms that arise from lesions within the system can be a diagnostic challenge. Vertigo has classically been divided into causes that are central (originating in the central nervous system) or peripheral (originating in the peripheral nervous system).

The HINTS (Head Impulse, Nystagmus, and Test of Skew) protocol is a group of 3 tests that can be used to differentiate central from peripheral vertigo (TABLE 2).^8,9 To perform the head impulse test, the examiner asks the patient to focus his gaze on a target and then rapidly turns the patient’s head to the side, watching the eyes for any corrective movements.¹⁰ When the eyes make a corrective saccade, the test is considered to be positive for a peripheral lesion.

Horizontal nystagmus is assessed by having the patient look in the direction of the fast phase of the nystagmus. If the nystagmus increases in intensity, then the test is considered positive for a peripheral lesion.

A careful description of the circumstances surrounding the dizziness episode can help identify underlying conditions such as orthostasis, hypoglycemia, or hyperventilation.

Vertigo can have many possible causes

Finally, the “test of skew” is performed by again having the patient fixate on the examiner’s nose. Each eye is tested by being covered, and then uncovered. If the uncovered eye has to move to refocus on the examiner’s nose, then the test is positive for a central lesion. A positive head impulse, positive horizontal nystagmus, and negative test of skew is 100% sensitive and 96% specific for a peripheral lesion.¹¹

Benign paroxysmal positional vertigo (BPPV) is vertigo that is triggered by movement of the head. It occurs when otoconia that are normally embedded in gel in the utricle become dislodged and migrate into the 3 fluid-filled semicircular canals, where they interfere with the normal fluid movement these canals use to sense head motion, causing the inner ear to send false signals to the brain.¹²

Diagnosis is confirmed by performing the Dix-Hallpike maneuver to elicit nystagmus. The patient is moved from a seated to a supine position with her head turned 45 degrees to the right and held for 30 seconds. For a demonstration of the Dix-Hallpike maneuver, see https://youtu.be/8RYB2QlO1N4. The Dix-Hallpike maneuver is also the first step of a treatment for BBPV known as the Epley maneuver. (See “The Epley maneuver: A procedure for treating BPPV”.^13,14)

Labyrinthitis—inflammation of the inner ear that can cause vertigo—is suggested by an acute, non-recurrent episode of dizziness that is often preceded by an upper respiratory infection. If the external canal is extremely painful and/or develops a vesicular rash, the patient might have herpes zoster of the geniculate ganglion (Ramsay Hunt syndrome type 2).

Dizziness related to presyncope often involves a cardiovascular pathology, such as a dysrhythmia or orthostatic hypotension.

Vertigo can have many possible causes

Vestibular migraine and Meniere’s disease. When a patient who has a history of migraines experiences symptoms of vertigo, vestibular migraine should be suspected, and treatment should focus on migraine therapy rather than vestibular therapy.¹⁵

Symptoms of Meniere’s disease and vestibular migraine can overlap.¹⁶ The current definition of Meniere’s disease requires ≥2 definitive episodes of vertigo with hearing loss plus tinnitus and/or aural symptoms.¹⁷ Thirty percent of vertigo episodes in patients with Meniere's disease can be attributed to BPPV.¹⁸

Acoustic neuroma. In addition to vertigo, acoustic neuroma is often associated with gradual hearing loss, tinnitus, and facial numbness (from compression of cranial nerve V preoperatively) or facial weakness (from compression of cranial nerve VII postoperatively). Unilateral hearing loss should prompt evaluation with magnetic resonance imaging.

“Acoustic neuroma” is a misnomer. The lesion arises from the vestibular (not the acoustic) portion of the 8th cranial nerve, and isn’t a neuroma; it is a schwannoma.¹⁹ Although it actually arises peripherally within the vestibular canal, it typically expands centrally and compresses other nerves centrally, which can make the clinical diagnosis more challenging if one were using the classical schema of differentiating between peripheral and central causes of vertigo.

Age-related vestibular loss occurs when the aging process causes deterioration of most of the components of the vestibulo-ocular reflex, resulting in dizziness and vertigo. Usually, the cerebral override mechanisms can compensate for the degeneration.

Other causes of vertigo include cerebellar infarction (3% of patients with vertigo),²⁰ sound-induced vertigo (Tullio phenomenon),²¹ obstructive sleep apnea,²² and systemic sclerosis.²³ Diabetes can cause a reduction in vestibular sensitivity that is evidenced by an increased reliance on visual stimuli to resolve vestibulo-visual conflict.²⁴

Disequilibrium

Disequilibrium is predominantly a loss of balance. Patients with disequilibrium have the feeling that they are about to fall, specifically without the sensation of spinning. They may appear to sway, and will reach out for something to support them. Disequilibrium can be a component of vertigo, or it may suggest a more specific diagnosis, such as ataxia, which is a lack of coordination when walking.

Atypical causes of dizziness

A positive head impulse test is highly suggestive of a peripheral lesion.

“Light-headedness” may have an element of euphoria or may be indistinguishable from the early part of a syncopal episode. Because other causes of light-headedness can be difficult to distinguish from presyncope, it is important to consider syncope in the differential diagnosis.

The differential of light-headedness can also include panic attack, early hyperventilation, and toxin exposure (such as diphenylarsinic acid,²⁵ pregabalin,²⁶ or paint thinner²⁷).

CORRESPONDENCE
Shannon Paul Starr, MD, Louisiana State University Health Sciences Center, 200 W. Esplanade #412, Kenner, LA 70065; sstarr@lsuhsc.edu.

Antispasmodics and tricyclics alleviate abdominal pain

A 2011 Cochrane review of 56 randomized controlled trials (RCTs) with 3725 patients compared bulking agents, antispasmodics, or antidepressants with placebo for treating IBS.¹ The pooled results from 13 RCTs with 1392 patients (65% female, mean age 45 years) showed that more patients had improved abdominal pain with antispasmodics than placebo over treatment periods varying from 6 days to 6 months (58% vs 46%; relative risk [RR]=1.3; 95% confidence interval [CI], 1.1-1.6; number needed to treat [NNT]=7).

The clinical relevance of the antispasmodic data is limited because the antispasmodics found effective for abdominal pain aren’t available in the United States. The pooled results from 8 RCTs with 517 patients (72% female, mean age 40) demonstrated greater improvement of abdominal pain with tricyclic and selective serotonin reuptake inhibitor antidepressants than placebo over 6 to 12 weeks (54% vs 37%; RR=1.5; 95% CI, 1.1–2.1; NNT=5). However, subgroup analysis found a statistically significant benefit for tricyclic antidepressants (4 trials; N=320; RR=1.3; 95% CI, 1.0-1.6) but no benefit for SSRIs (4 trials; N=197; RR=2.3; 95% CI, 0.79-6.7).

Effective antispasmodics aren’t available in the United States

A 2012 meta-analysis of 23 RCTs with 2585 patients examined the effect of antispasmodic agents, alone or in combination, to treat IBS.² Pooled results from 13 RCTs with 2394 patients (69% female, ages 16 years or older) favored treatment with antispasmodics over placebo for abdominal pain (odds ratio [OR]=1.5; 95% CI, 1.3-1.8). No difference in adverse events was found between antispasmodics and placebo (9 trials; N=2239; OR=0.74; 95% CI, 0.54-0.98). The antispasmodics found effective for abdominal pain in this meta-analysis aren’t available in the United States.

Peppermint oil helps, but can cause heartburn

A 2013 meta-analysis of 9 RCTs with 726 patients compared various doses of enteric-coated peppermint oil with placebo over a minimum of 2 weeks’ treatment.³ Five RCTs with 357 patients (62% female, 13.4% children) demonstrated improvement of abdominal pain in 57% of patients taking peppermint oil compared with 27% receiving placebo (RR=2.1; 95% CI, 1.6-2.8; NNT=4 at 2 to 8 weeks). No statistically significant heterogeneity was identified among the treatment groups.

Pooled analysis found that peppermint oil patients were more likely than placebo patients to experience an adverse event (7 trials; N=474; 22% vs 13%; RR=1.7; 95% CI, 1.3-2.4), but that the events were generally mild and transient. The most frequently reported adverse event was heartburn.

Antispasmodics and tricyclics alleviate abdominal pain

A 2011 Cochrane review of 56 randomized controlled trials (RCTs) with 3725 patients compared bulking agents, antispasmodics, or antidepressants with placebo for treating IBS.¹ The pooled results from 13 RCTs with 1392 patients (65% female, mean age 45 years) showed that more patients had improved abdominal pain with antispasmodics than placebo over treatment periods varying from 6 days to 6 months (58% vs 46%; relative risk [RR]=1.3; 95% confidence interval [CI], 1.1-1.6; number needed to treat [NNT]=7).

The clinical relevance of the antispasmodic data is limited because the antispasmodics found effective for abdominal pain aren’t available in the United States. The pooled results from 8 RCTs with 517 patients (72% female, mean age 40) demonstrated greater improvement of abdominal pain with tricyclic and selective serotonin reuptake inhibitor antidepressants than placebo over 6 to 12 weeks (54% vs 37%; RR=1.5; 95% CI, 1.1–2.1; NNT=5). However, subgroup analysis found a statistically significant benefit for tricyclic antidepressants (4 trials; N=320; RR=1.3; 95% CI, 1.0-1.6) but no benefit for SSRIs (4 trials; N=197; RR=2.3; 95% CI, 0.79-6.7).

Effective antispasmodics aren’t available in the United States

A 2012 meta-analysis of 23 RCTs with 2585 patients examined the effect of antispasmodic agents, alone or in combination, to treat IBS.² Pooled results from 13 RCTs with 2394 patients (69% female, ages 16 years or older) favored treatment with antispasmodics over placebo for abdominal pain (odds ratio [OR]=1.5; 95% CI, 1.3-1.8). No difference in adverse events was found between antispasmodics and placebo (9 trials; N=2239; OR=0.74; 95% CI, 0.54-0.98). The antispasmodics found effective for abdominal pain in this meta-analysis aren’t available in the United States.

Peppermint oil helps, but can cause heartburn

A 2013 meta-analysis of 9 RCTs with 726 patients compared various doses of enteric-coated peppermint oil with placebo over a minimum of 2 weeks’ treatment.³ Five RCTs with 357 patients (62% female, 13.4% children) demonstrated improvement of abdominal pain in 57% of patients taking peppermint oil compared with 27% receiving placebo (RR=2.1; 95% CI, 1.6-2.8; NNT=4 at 2 to 8 weeks). No statistically significant heterogeneity was identified among the treatment groups.

Pooled analysis found that peppermint oil patients were more likely than placebo patients to experience an adverse event (7 trials; N=474; 22% vs 13%; RR=1.7; 95% CI, 1.3-2.4), but that the events were generally mild and transient. The most frequently reported adverse event was heartburn.

Antispasmodics and tricyclics alleviate abdominal pain

A 2011 Cochrane review of 56 randomized controlled trials (RCTs) with 3725 patients compared bulking agents, antispasmodics, or antidepressants with placebo for treating IBS.¹ The pooled results from 13 RCTs with 1392 patients (65% female, mean age 45 years) showed that more patients had improved abdominal pain with antispasmodics than placebo over treatment periods varying from 6 days to 6 months (58% vs 46%; relative risk [RR]=1.3; 95% confidence interval [CI], 1.1-1.6; number needed to treat [NNT]=7).

The clinical relevance of the antispasmodic data is limited because the antispasmodics found effective for abdominal pain aren’t available in the United States. The pooled results from 8 RCTs with 517 patients (72% female, mean age 40) demonstrated greater improvement of abdominal pain with tricyclic and selective serotonin reuptake inhibitor antidepressants than placebo over 6 to 12 weeks (54% vs 37%; RR=1.5; 95% CI, 1.1–2.1; NNT=5). However, subgroup analysis found a statistically significant benefit for tricyclic antidepressants (4 trials; N=320; RR=1.3; 95% CI, 1.0-1.6) but no benefit for SSRIs (4 trials; N=197; RR=2.3; 95% CI, 0.79-6.7).

Effective antispasmodics aren’t available in the United States

A 2012 meta-analysis of 23 RCTs with 2585 patients examined the effect of antispasmodic agents, alone or in combination, to treat IBS.² Pooled results from 13 RCTs with 2394 patients (69% female, ages 16 years or older) favored treatment with antispasmodics over placebo for abdominal pain (odds ratio [OR]=1.5; 95% CI, 1.3-1.8). No difference in adverse events was found between antispasmodics and placebo (9 trials; N=2239; OR=0.74; 95% CI, 0.54-0.98). The antispasmodics found effective for abdominal pain in this meta-analysis aren’t available in the United States.

Peppermint oil helps, but can cause heartburn

A 2013 meta-analysis of 9 RCTs with 726 patients compared various doses of enteric-coated peppermint oil with placebo over a minimum of 2 weeks’ treatment.³ Five RCTs with 357 patients (62% female, 13.4% children) demonstrated improvement of abdominal pain in 57% of patients taking peppermint oil compared with 27% receiving placebo (RR=2.1; 95% CI, 1.6-2.8; NNT=4 at 2 to 8 weeks). No statistically significant heterogeneity was identified among the treatment groups.

Pooled analysis found that peppermint oil patients were more likely than placebo patients to experience an adverse event (7 trials; N=474; 22% vs 13%; RR=1.7; 95% CI, 1.3-2.4), but that the events were generally mild and transient. The most frequently reported adverse event was heartburn.

EVIDENCE SUMMARY

Six cohort studies found that patients who underwent ASD for subacromial impingement had improved pain and function scores at 4.5 to 12 years after surgery (TABLE^1-7). Weaknesses of the overall data set include use of heterogeneous outcome measures across studies, lack of sham surgical controls, and lack of blinding.

ASD improves pain and function slightly more than other treatments

One prospective and one retrospective cohort trial compared ASD with another intervention. In the prospective trial, ASD was associated with a 10% better combined pain and function score than open acromioplasty at 12 years.¹ In the retrospective trial, ASD was also associated with a 10% better combined pain and function score than prolonged physical therapy in patients older than 57 years (the median age of study participants) but not patients younger than 57 years.²

Two other studies found improvements in pain and function

Two other prospective cohort studies didn’t use a comparison group but followed changes in standardized shoulder pain and function scores for 5 to 6 years after ASD. In one study, pain decreased 6 points on a 10-point visual analog scale by 6 months postop (P<.001).³ In both studies, a 9% to 10% improvement in function was seen between 6 months and 5 to 6 years after surgery.^3,4

A third cohort study that asked patients about overall pain and satisfaction 8 to 11 years after ASD found that most were “very” or “quite” satisfied and half were pain-free.^5,6

Rotator cuff tears found less likely with ASD

An anatomic study obtained ultrasounds of patients 13 to 17 years after ASD and compared the findings to rotator cuff ultrasounds of the general population.⁷ Patients who had ASD were 22% less likely to demonstrate rotator cuff tears at the end of the study (no statistics were reported to measure significance).

RECOMMENDATIONS

Guidelines from the Washington State Department of Labor and Industry state that patients who should undergo isolated subacromial decompression (with or without acromioplasty) need to have documented subacromial impingement syndrome with magnetic resonance imaging evidence of rotator cuff tendonopathy or tear, have undergone 12 weeks of conservative therapy (including at least active assisted range of motion and home-based exercises), and have had a subacromial injection with a local anesthetic that has provided documented relief of pain.⁸

No current guidelines are available from national or international orthopedic or sports medicine organizations.

EVIDENCE SUMMARY

Six cohort studies found that patients who underwent ASD for subacromial impingement had improved pain and function scores at 4.5 to 12 years after surgery (TABLE^1-7). Weaknesses of the overall data set include use of heterogeneous outcome measures across studies, lack of sham surgical controls, and lack of blinding.

ASD improves pain and function slightly more than other treatments

One prospective and one retrospective cohort trial compared ASD with another intervention. In the prospective trial, ASD was associated with a 10% better combined pain and function score than open acromioplasty at 12 years.¹ In the retrospective trial, ASD was also associated with a 10% better combined pain and function score than prolonged physical therapy in patients older than 57 years (the median age of study participants) but not patients younger than 57 years.²

Two other studies found improvements in pain and function

Two other prospective cohort studies didn’t use a comparison group but followed changes in standardized shoulder pain and function scores for 5 to 6 years after ASD. In one study, pain decreased 6 points on a 10-point visual analog scale by 6 months postop (P<.001).³ In both studies, a 9% to 10% improvement in function was seen between 6 months and 5 to 6 years after surgery.^3,4

A third cohort study that asked patients about overall pain and satisfaction 8 to 11 years after ASD found that most were “very” or “quite” satisfied and half were pain-free.^5,6

Rotator cuff tears found less likely with ASD

An anatomic study obtained ultrasounds of patients 13 to 17 years after ASD and compared the findings to rotator cuff ultrasounds of the general population.⁷ Patients who had ASD were 22% less likely to demonstrate rotator cuff tears at the end of the study (no statistics were reported to measure significance).

RECOMMENDATIONS

Guidelines from the Washington State Department of Labor and Industry state that patients who should undergo isolated subacromial decompression (with or without acromioplasty) need to have documented subacromial impingement syndrome with magnetic resonance imaging evidence of rotator cuff tendonopathy or tear, have undergone 12 weeks of conservative therapy (including at least active assisted range of motion and home-based exercises), and have had a subacromial injection with a local anesthetic that has provided documented relief of pain.⁸

No current guidelines are available from national or international orthopedic or sports medicine organizations.

EVIDENCE SUMMARY

Six cohort studies found that patients who underwent ASD for subacromial impingement had improved pain and function scores at 4.5 to 12 years after surgery (TABLE^1-7). Weaknesses of the overall data set include use of heterogeneous outcome measures across studies, lack of sham surgical controls, and lack of blinding.

ASD improves pain and function slightly more than other treatments

One prospective and one retrospective cohort trial compared ASD with another intervention. In the prospective trial, ASD was associated with a 10% better combined pain and function score than open acromioplasty at 12 years.¹ In the retrospective trial, ASD was also associated with a 10% better combined pain and function score than prolonged physical therapy in patients older than 57 years (the median age of study participants) but not patients younger than 57 years.²

Two other studies found improvements in pain and function

Two other prospective cohort studies didn’t use a comparison group but followed changes in standardized shoulder pain and function scores for 5 to 6 years after ASD. In one study, pain decreased 6 points on a 10-point visual analog scale by 6 months postop (P<.001).³ In both studies, a 9% to 10% improvement in function was seen between 6 months and 5 to 6 years after surgery.^3,4

A third cohort study that asked patients about overall pain and satisfaction 8 to 11 years after ASD found that most were “very” or “quite” satisfied and half were pain-free.^5,6

Rotator cuff tears found less likely with ASD

An anatomic study obtained ultrasounds of patients 13 to 17 years after ASD and compared the findings to rotator cuff ultrasounds of the general population.⁷ Patients who had ASD were 22% less likely to demonstrate rotator cuff tears at the end of the study (no statistics were reported to measure significance).

RECOMMENDATIONS

Guidelines from the Washington State Department of Labor and Industry state that patients who should undergo isolated subacromial decompression (with or without acromioplasty) need to have documented subacromial impingement syndrome with magnetic resonance imaging evidence of rotator cuff tendonopathy or tear, have undergone 12 weeks of conservative therapy (including at least active assisted range of motion and home-based exercises), and have had a subacromial injection with a local anesthetic that has provided documented relief of pain.⁸

No current guidelines are available from national or international orthopedic or sports medicine organizations.

ABSTRACT

Purpose Few studies have quantitatively examined the degree to which the use of the computer affects patients’ satisfaction with the clinician and the quality of the visit. We conducted a study to examine this association.

Methods Twenty-three clinicians (21 internal medicine physicians, 2 nurse practitioners) were recruited from 4 Veteran Affairs Medical Center (VAMC) clinics located in San Diego, Calif. Five to 6 patients for most clinicians (one patient each for 2 of the clinicians) were recruited to participate in a study of patient-physician communication. The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen, and through the use of the Morae usability testing software system, which recorded clinician clicks and scrolls on the computer. After the visit, patients were asked to complete a satisfaction survey.

Results The final sample consisted of 126 consultations. Total patient satisfaction (beta=0.014; P=.027) and patient satisfaction with patient-centered communication (beta=0.02; P=.02) were significantly associated with higher clinician “gaze time” at the patient. A higher percentage of gaze time during a visit (controlling for the length of the visit) was significantly associated with greater satisfaction with patient-centered communication (beta=0.628; P=.033).

Conclusions Higher clinician gaze time at the patient predicted greater patient satisfaction. This suggests that clinicians would be well served to refine their multitasking skills so that they communicate in a patient-centered manner while performing necessary computer-related tasks. These findings also have important implications for clinical training with respect to using an electronic health record (EHR) system in ways that do not impede the one-on-one conversation between clinician and patient.

Primary care physicians’ use of electronic health record (EHR) systems has markedly increased in recent years. For example, a 2008 study of more than 1000 randomly selected practicing physicians in Massachusetts found that 33% utilized an EHR.¹ Many physicians believe that EHR systems are beneficial to patient care,² and several studies have supported this perception, showing clear benefits of EHR use. A study of one component of EHR systems—computerized physician order entry (CPOE)—found that CPOEs resulted in a >50% decrease in serious medication errors.³ Other errors have declined with the use of EHR systems, as well; Virapongse et al¹ found a trend towards fewer paid malpractice claims against physicians who used an EHR compared to those physicians using paper charting.

EHR systems may also improve efficiency. In a study of a health maintenance organization (HMO) model, initiating an EHR system improved efficiency by decreasing office visits.⁴ Widespread adoption of EHR systems could save an estimated $81 billion annually through reductions in errors and adverse events, and improved preventive care and chronic disease management.⁵ In a survey of approximately 300 patients who had been evaluated at a family medicine clinic for hypertension, high blood pressure without hypertension, or hyperlipidemia, 75% indicated that they felt EHRs had a positive impact on their care.⁶

Higher clinician gaze time at the patient predicted greater patient satisfaction.

However, some clinicians are concerned about the possible negative impact of EHR systems on health care. One major concern is that EHR systems might increase physician workload⁷ and the amount of time spent using a computer during patient visits. A study that examined physician EHR use found that while time spent on certain tasks, such as prescription writing and lab ordering, was reduced, there was an overall increase in time spent on computer tasks related to charting, preventive care, and chronic disease management.⁸ Baron et al⁹ also found an increase in time spent using the EHR during each clinic session in one private practice setting.

Physicians are also concerned that EHR systems might interfere with the patient-physician interaction (eg, maintaining eye contact, paying attention to patients’ concerns) by directing the physician’s attention away from the patient and toward the computer.¹⁰ In one study, this concern increased after physicians started utilizing a new EHR system.¹¹ Although a survey of inpatients indicated that residents engaged in greater patient-physician communication after an EHR was implemented,¹² a separate study conducted in an outpatient setting found physicians spent less time looking at patients after converting from a paper-based system to an EHR system.¹³

Very few studies have quantitatively examined the association of patient satisfaction with clinician EHR usage. The goal of this study was to examine the correlation of patient satisfaction with actual EHR usage in an ambulatory setting. The data reported in this paper are part of a larger study aimed at understanding EHR use in a VAMC.

METHODS

Study design and sample

The study participants were clinicians in 4 VAMC community clinics located in San Diego, Calif. Twenty-three clinicians (21 general internal medicine physicians and 2 nurse practitioners) were enrolled in the study. Most clinicians identified 5 to 6 patients from their practices to participate in the study (2 participants identified only one patient each). All patients were visiting their clinician for either an acute visit or a follow-up visit.

Although there were slight variations in clinic room size and shape, all rooms were equipped with a compact desk against a wall, a rolling desk chair, a desktop computer with keyboard and mouse, and a second, fixed chair placed diagonal to the physician’s chair. Two rooms had dual monitors. There was a standard examination table in all examination rooms.

The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen. A usability testing software system (Morae) was used to record clinicians’ computer activities, including mouse clicks and scrolls on the computer. The Computerized Patient Records System (CPRS) was the EHR used by all clinicians in this study.

At the end of the visit, patients were asked to complete a satisfaction survey with questions in 3 domains: the physician’s engagement in patient-centered communication, the physician’s clinical skills, and the physician’s interpersonal skills.

Data analysis

Descriptive statistics were used to document patient characteristics, the clinicians’ EHR usage (total number of mouse clicks and scrolls during the visit) and interaction with the patient (gaze time at EHR vs at patient and companion), and to summarize patient satisfaction with the visit. To account for clinician cluster effect, a linear mixed effects model was used to assess the associations between patient satisfaction with the clinician and 2 variables: the amount of clinician time spent viewing or using the computer and the clinician time spent interacting with the patient.

We also assessed the above associations by controlling for visit length. Visit lengths not significant at P<.10 were reported as unadjusted analyses.

All analyses were performed using R statistical software, with a P value of <.05 interpreted as statistically significant.

RESULTS

Satisfaction surveys and video and Morae data were collected for 126 individual patient office visits to the 23 participating physicians and nurses. A majority of the patients who participated in the study were older (mean: 60.5 years; standard deviation [SD]=13.4 years), male as expected in a VA setting (96.8%), Caucasian (65.1%), and had at least some college education (81.7%, TABLE 1).

Patients rated their satisfaction in 3 domains—patient-centered communication, physician clinical skills, and physician interpersonal skills—using a 1 to 5 scale (1=least satisfied, 5=most satisfied). Patients in this study were highly satisfied with their physician or nurse in all 3 domains and overall (TABLE 2), with an average satisfaction score of 4.52 ± 0.51 for patient-centered communication, 4.71 ± 0.56 for physician clinical skills, 4.86 ± 0.32 for physician interpersonal skills, and 4.64 ± 0.38 for total satisfaction.

The physicians and nurses used their EHR system extensively during the visits as delineated by the number of clicks and scrolls on the computer. The average number of clicks and scrolls was 192, with a maximum of 685 clicks and scrolls during one visit. The average visit lasted 30.7 minutes, and on average the clinician spent 12.7 minutes (SD: 8.22 minutes), or an average of about 39.4% of total visit time, viewing or working on the EHR; an average of 10.8 minutes (SD: 5.63 minutes), or an average of about 36.3% of total visit time, was spent interacting with the patient (TABLE 3).

Without adjusting for visit length, patient satisfaction with the clinicians’ patientcentered communication (beta=0.02; P=.02) and total satisfaction (beta=0.014; P=.027) were significantly associated with clinicians’ gaze time at the patient; more clinician gaze time at the patient resulted in greater patient satisfaction (TABLE 4). Adding visit length to the above models had no significant effect (P>.10); therefore, we did not include it in the models.

Patient satisfaction with clinicians’ interpersonal skills was positively associated with gaze time at the patient (beta=0.013, P =.017) without adjusting for visit length. Since the normal assumption of residuals was not plausible based on a normal probability plot, we also assessed the association by dichotomizing the score (5=very satisfied vs <5=not very satisfied) and this significance disappeared. This association was not significant while controlling for visit length.

The percentage of gaze time at the patient (the fraction of patient gaze time over the entire visit) was not significantly associated with patient-centered communication (beta=0.483, P=.12, TABLE 4) when not adjusted for visit length. After adjusting for visit length (P=.052), the association became significant (beta=0.628, P=.033); thus, the higher percentage of time the clinician spent interacting with the patient, the more satisfied the patient was.

DISCUSSION

In this study, patients were highly satisfied with their clinicians despite often high usage of the EHR. Gadd and Penrod¹¹ reported that patients perceived no impact on communication or eye contact with the clinician despite the initiation of an EHR system in 6 large academic medical practices. Another study demonstrated no significant differences in patient satisfaction with their physicians when comparing patients whose physicians used a paper charting system with those who used an EHR system.¹⁴

The average number of computer clicks and scrolls per visit was 192, with a maximum of 685 clicks and scrolls during one visit.

The fact that patients demonstrated high levels of satisfaction with patient-clinician communication even for clinicians with high EHR usage is somewhat surprising. However, Hsu et al¹⁵ found patients’ satisfaction with their clinicians’ communication about medical issues and familiarity with them increased 7 months after implementing an EHR system. In a different study that analyzed videotaped interactions between patients and 5 physicians, the patients found it disturbing not knowing what their doctor was doing when he or she worked on the computer, and preferred being able to see the computer screen.¹⁶ This study suggests that it’s advisable for clinicians to describe what they are doing when they use the computer, so that patients better understand how this time spent inputting data actually benefits them.

EHRs can be time-consuming. Physicians and nurses in our study interacted with the EHR a great deal during the office visit, as evidenced by the large average number of clicks and scrolls. This finding confirms clinicians’ perceptions of the amount of work the EHR system requires. For example, in a semi-structured interview of physicians regarding their use of a VA EHR system,¹⁰ one respondent noted that the reminders in the EHR required hundreds of clicks.

In our study, the average number of clicks and scrolls during the visit was 192, with some clinicians registering hundreds more. In fact, concerns about the time involved in the use of the EHR and about the adequacy of data collection may lead some clinicians who currently don’t have an EHR system to be reluctant to integrate one into their practices.¹⁷

In this study, patients were highly satisfied with their clinicians, despite often high usage of the EHR.

Makoul et al¹⁸ found that compared with physicians who used a paper chart, physicians who used an EHR system were more active in clarifying information from the patient and encouraging patient questions during visits, although the study found a trend toward less active roles in more patient-centered communication when using an EHR system. This latter finding is similar to the concerns raised in our study.

Clinical and communication skills are factors, too. One study found that compared to patients who were cared for by more experienced physicians, patients seen by residents using EHRs were more likely to feel that the physician spent less time talking with them and examining them; they were also more likely to report that the visit felt less personal.¹⁹ Another study found that clinicians with poor baseline communication skills had more difficulties interacting with patients when an EHR system was introduced than those who had better baseline communication skills.²⁰

Training needed to improve communication during EHR use. Research has shown that when used properly and thoughtfully, EHR use can result in greater patient engagement.²¹ But, as noted above, there are challenges, suggesting a need for training clinicians to more successfully use an EHR system while simultaneously communicating with their patients.

Study limitations. This study was conducted at a single site, using a single EHR system deployed in the VA clinics. We cannot generalize our findings to other sites or types of clinic systems. Other EHR systems may have different functionalities, which may affect the time required to provide the same type of medical care.

In addition, the study involved only 23 physicians and nurses in a single health system. Other clinicians may have patterns different from those we studied, although a wide range of patterns was seen among the participants, as demonstrated by the large variation in the number of clicks and scrolls. Another limitation is that study patients were not randomly selected, but rather referred by the provider, and the visits were not blinded to either the provider or patient. This may cause some selection bias.

EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication.

In this study of VA clinicians’ EHR use, patients expressed satisfaction with the clinicians’ clinical skills and patient-centered communication when the clinician spent more time and a greater percentage of the visit engaging the patient. EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication between the clinician and the patient, and to ease the workload of EHR documentation. In the meantime, clinicians should be trained in how to expedite their use of the EHR during the clinical visit as well as outside of the exam room in order to improve their patients’ satisfaction.

CORRESPONDENCE
Neil J. Farber, MD, University of California, San Diego, 8939 Villa La Jolla Drive, La Jolla, CA 92037; nfarber@ucsd.edu.

ABSTRACT

Purpose Few studies have quantitatively examined the degree to which the use of the computer affects patients’ satisfaction with the clinician and the quality of the visit. We conducted a study to examine this association.

Methods Twenty-three clinicians (21 internal medicine physicians, 2 nurse practitioners) were recruited from 4 Veteran Affairs Medical Center (VAMC) clinics located in San Diego, Calif. Five to 6 patients for most clinicians (one patient each for 2 of the clinicians) were recruited to participate in a study of patient-physician communication. The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen, and through the use of the Morae usability testing software system, which recorded clinician clicks and scrolls on the computer. After the visit, patients were asked to complete a satisfaction survey.

Results The final sample consisted of 126 consultations. Total patient satisfaction (beta=0.014; P=.027) and patient satisfaction with patient-centered communication (beta=0.02; P=.02) were significantly associated with higher clinician “gaze time” at the patient. A higher percentage of gaze time during a visit (controlling for the length of the visit) was significantly associated with greater satisfaction with patient-centered communication (beta=0.628; P=.033).

Conclusions Higher clinician gaze time at the patient predicted greater patient satisfaction. This suggests that clinicians would be well served to refine their multitasking skills so that they communicate in a patient-centered manner while performing necessary computer-related tasks. These findings also have important implications for clinical training with respect to using an electronic health record (EHR) system in ways that do not impede the one-on-one conversation between clinician and patient.

Primary care physicians’ use of electronic health record (EHR) systems has markedly increased in recent years. For example, a 2008 study of more than 1000 randomly selected practicing physicians in Massachusetts found that 33% utilized an EHR.¹ Many physicians believe that EHR systems are beneficial to patient care,² and several studies have supported this perception, showing clear benefits of EHR use. A study of one component of EHR systems—computerized physician order entry (CPOE)—found that CPOEs resulted in a >50% decrease in serious medication errors.³ Other errors have declined with the use of EHR systems, as well; Virapongse et al¹ found a trend towards fewer paid malpractice claims against physicians who used an EHR compared to those physicians using paper charting.

EHR systems may also improve efficiency. In a study of a health maintenance organization (HMO) model, initiating an EHR system improved efficiency by decreasing office visits.⁴ Widespread adoption of EHR systems could save an estimated $81 billion annually through reductions in errors and adverse events, and improved preventive care and chronic disease management.⁵ In a survey of approximately 300 patients who had been evaluated at a family medicine clinic for hypertension, high blood pressure without hypertension, or hyperlipidemia, 75% indicated that they felt EHRs had a positive impact on their care.⁶

Higher clinician gaze time at the patient predicted greater patient satisfaction.

However, some clinicians are concerned about the possible negative impact of EHR systems on health care. One major concern is that EHR systems might increase physician workload⁷ and the amount of time spent using a computer during patient visits. A study that examined physician EHR use found that while time spent on certain tasks, such as prescription writing and lab ordering, was reduced, there was an overall increase in time spent on computer tasks related to charting, preventive care, and chronic disease management.⁸ Baron et al⁹ also found an increase in time spent using the EHR during each clinic session in one private practice setting.

Physicians are also concerned that EHR systems might interfere with the patient-physician interaction (eg, maintaining eye contact, paying attention to patients’ concerns) by directing the physician’s attention away from the patient and toward the computer.¹⁰ In one study, this concern increased after physicians started utilizing a new EHR system.¹¹ Although a survey of inpatients indicated that residents engaged in greater patient-physician communication after an EHR was implemented,¹² a separate study conducted in an outpatient setting found physicians spent less time looking at patients after converting from a paper-based system to an EHR system.¹³

Very few studies have quantitatively examined the association of patient satisfaction with clinician EHR usage. The goal of this study was to examine the correlation of patient satisfaction with actual EHR usage in an ambulatory setting. The data reported in this paper are part of a larger study aimed at understanding EHR use in a VAMC.

METHODS

Study design and sample

The study participants were clinicians in 4 VAMC community clinics located in San Diego, Calif. Twenty-three clinicians (21 general internal medicine physicians and 2 nurse practitioners) were enrolled in the study. Most clinicians identified 5 to 6 patients from their practices to participate in the study (2 participants identified only one patient each). All patients were visiting their clinician for either an acute visit or a follow-up visit.

Although there were slight variations in clinic room size and shape, all rooms were equipped with a compact desk against a wall, a rolling desk chair, a desktop computer with keyboard and mouse, and a second, fixed chair placed diagonal to the physician’s chair. Two rooms had dual monitors. There was a standard examination table in all examination rooms.

The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen. A usability testing software system (Morae) was used to record clinicians’ computer activities, including mouse clicks and scrolls on the computer. The Computerized Patient Records System (CPRS) was the EHR used by all clinicians in this study.

At the end of the visit, patients were asked to complete a satisfaction survey with questions in 3 domains: the physician’s engagement in patient-centered communication, the physician’s clinical skills, and the physician’s interpersonal skills.

Data analysis

Descriptive statistics were used to document patient characteristics, the clinicians’ EHR usage (total number of mouse clicks and scrolls during the visit) and interaction with the patient (gaze time at EHR vs at patient and companion), and to summarize patient satisfaction with the visit. To account for clinician cluster effect, a linear mixed effects model was used to assess the associations between patient satisfaction with the clinician and 2 variables: the amount of clinician time spent viewing or using the computer and the clinician time spent interacting with the patient.

We also assessed the above associations by controlling for visit length. Visit lengths not significant at P<.10 were reported as unadjusted analyses.

All analyses were performed using R statistical software, with a P value of <.05 interpreted as statistically significant.

RESULTS

Satisfaction surveys and video and Morae data were collected for 126 individual patient office visits to the 23 participating physicians and nurses. A majority of the patients who participated in the study were older (mean: 60.5 years; standard deviation [SD]=13.4 years), male as expected in a VA setting (96.8%), Caucasian (65.1%), and had at least some college education (81.7%, TABLE 1).

Patients rated their satisfaction in 3 domains—patient-centered communication, physician clinical skills, and physician interpersonal skills—using a 1 to 5 scale (1=least satisfied, 5=most satisfied). Patients in this study were highly satisfied with their physician or nurse in all 3 domains and overall (TABLE 2), with an average satisfaction score of 4.52 ± 0.51 for patient-centered communication, 4.71 ± 0.56 for physician clinical skills, 4.86 ± 0.32 for physician interpersonal skills, and 4.64 ± 0.38 for total satisfaction.

The physicians and nurses used their EHR system extensively during the visits as delineated by the number of clicks and scrolls on the computer. The average number of clicks and scrolls was 192, with a maximum of 685 clicks and scrolls during one visit. The average visit lasted 30.7 minutes, and on average the clinician spent 12.7 minutes (SD: 8.22 minutes), or an average of about 39.4% of total visit time, viewing or working on the EHR; an average of 10.8 minutes (SD: 5.63 minutes), or an average of about 36.3% of total visit time, was spent interacting with the patient (TABLE 3).

Without adjusting for visit length, patient satisfaction with the clinicians’ patientcentered communication (beta=0.02; P=.02) and total satisfaction (beta=0.014; P=.027) were significantly associated with clinicians’ gaze time at the patient; more clinician gaze time at the patient resulted in greater patient satisfaction (TABLE 4). Adding visit length to the above models had no significant effect (P>.10); therefore, we did not include it in the models.

Patient satisfaction with clinicians’ interpersonal skills was positively associated with gaze time at the patient (beta=0.013, P =.017) without adjusting for visit length. Since the normal assumption of residuals was not plausible based on a normal probability plot, we also assessed the association by dichotomizing the score (5=very satisfied vs <5=not very satisfied) and this significance disappeared. This association was not significant while controlling for visit length.

The percentage of gaze time at the patient (the fraction of patient gaze time over the entire visit) was not significantly associated with patient-centered communication (beta=0.483, P=.12, TABLE 4) when not adjusted for visit length. After adjusting for visit length (P=.052), the association became significant (beta=0.628, P=.033); thus, the higher percentage of time the clinician spent interacting with the patient, the more satisfied the patient was.

DISCUSSION

In this study, patients were highly satisfied with their clinicians despite often high usage of the EHR. Gadd and Penrod¹¹ reported that patients perceived no impact on communication or eye contact with the clinician despite the initiation of an EHR system in 6 large academic medical practices. Another study demonstrated no significant differences in patient satisfaction with their physicians when comparing patients whose physicians used a paper charting system with those who used an EHR system.¹⁴

The average number of computer clicks and scrolls per visit was 192, with a maximum of 685 clicks and scrolls during one visit.

The fact that patients demonstrated high levels of satisfaction with patient-clinician communication even for clinicians with high EHR usage is somewhat surprising. However, Hsu et al¹⁵ found patients’ satisfaction with their clinicians’ communication about medical issues and familiarity with them increased 7 months after implementing an EHR system. In a different study that analyzed videotaped interactions between patients and 5 physicians, the patients found it disturbing not knowing what their doctor was doing when he or she worked on the computer, and preferred being able to see the computer screen.¹⁶ This study suggests that it’s advisable for clinicians to describe what they are doing when they use the computer, so that patients better understand how this time spent inputting data actually benefits them.

EHRs can be time-consuming. Physicians and nurses in our study interacted with the EHR a great deal during the office visit, as evidenced by the large average number of clicks and scrolls. This finding confirms clinicians’ perceptions of the amount of work the EHR system requires. For example, in a semi-structured interview of physicians regarding their use of a VA EHR system,¹⁰ one respondent noted that the reminders in the EHR required hundreds of clicks.

In our study, the average number of clicks and scrolls during the visit was 192, with some clinicians registering hundreds more. In fact, concerns about the time involved in the use of the EHR and about the adequacy of data collection may lead some clinicians who currently don’t have an EHR system to be reluctant to integrate one into their practices.¹⁷

In this study, patients were highly satisfied with their clinicians, despite often high usage of the EHR.

Makoul et al¹⁸ found that compared with physicians who used a paper chart, physicians who used an EHR system were more active in clarifying information from the patient and encouraging patient questions during visits, although the study found a trend toward less active roles in more patient-centered communication when using an EHR system. This latter finding is similar to the concerns raised in our study.

Clinical and communication skills are factors, too. One study found that compared to patients who were cared for by more experienced physicians, patients seen by residents using EHRs were more likely to feel that the physician spent less time talking with them and examining them; they were also more likely to report that the visit felt less personal.¹⁹ Another study found that clinicians with poor baseline communication skills had more difficulties interacting with patients when an EHR system was introduced than those who had better baseline communication skills.²⁰

Training needed to improve communication during EHR use. Research has shown that when used properly and thoughtfully, EHR use can result in greater patient engagement.²¹ But, as noted above, there are challenges, suggesting a need for training clinicians to more successfully use an EHR system while simultaneously communicating with their patients.

Study limitations. This study was conducted at a single site, using a single EHR system deployed in the VA clinics. We cannot generalize our findings to other sites or types of clinic systems. Other EHR systems may have different functionalities, which may affect the time required to provide the same type of medical care.

In addition, the study involved only 23 physicians and nurses in a single health system. Other clinicians may have patterns different from those we studied, although a wide range of patterns was seen among the participants, as demonstrated by the large variation in the number of clicks and scrolls. Another limitation is that study patients were not randomly selected, but rather referred by the provider, and the visits were not blinded to either the provider or patient. This may cause some selection bias.

EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication.

In this study of VA clinicians’ EHR use, patients expressed satisfaction with the clinicians’ clinical skills and patient-centered communication when the clinician spent more time and a greater percentage of the visit engaging the patient. EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication between the clinician and the patient, and to ease the workload of EHR documentation. In the meantime, clinicians should be trained in how to expedite their use of the EHR during the clinical visit as well as outside of the exam room in order to improve their patients’ satisfaction.

CORRESPONDENCE
Neil J. Farber, MD, University of California, San Diego, 8939 Villa La Jolla Drive, La Jolla, CA 92037; nfarber@ucsd.edu.

ABSTRACT

Purpose Few studies have quantitatively examined the degree to which the use of the computer affects patients’ satisfaction with the clinician and the quality of the visit. We conducted a study to examine this association.

Methods Twenty-three clinicians (21 internal medicine physicians, 2 nurse practitioners) were recruited from 4 Veteran Affairs Medical Center (VAMC) clinics located in San Diego, Calif. Five to 6 patients for most clinicians (one patient each for 2 of the clinicians) were recruited to participate in a study of patient-physician communication. The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen, and through the use of the Morae usability testing software system, which recorded clinician clicks and scrolls on the computer. After the visit, patients were asked to complete a satisfaction survey.

Results The final sample consisted of 126 consultations. Total patient satisfaction (beta=0.014; P=.027) and patient satisfaction with patient-centered communication (beta=0.02; P=.02) were significantly associated with higher clinician “gaze time” at the patient. A higher percentage of gaze time during a visit (controlling for the length of the visit) was significantly associated with greater satisfaction with patient-centered communication (beta=0.628; P=.033).

Conclusions Higher clinician gaze time at the patient predicted greater patient satisfaction. This suggests that clinicians would be well served to refine their multitasking skills so that they communicate in a patient-centered manner while performing necessary computer-related tasks. These findings also have important implications for clinical training with respect to using an electronic health record (EHR) system in ways that do not impede the one-on-one conversation between clinician and patient.

Primary care physicians’ use of electronic health record (EHR) systems has markedly increased in recent years. For example, a 2008 study of more than 1000 randomly selected practicing physicians in Massachusetts found that 33% utilized an EHR.¹ Many physicians believe that EHR systems are beneficial to patient care,² and several studies have supported this perception, showing clear benefits of EHR use. A study of one component of EHR systems—computerized physician order entry (CPOE)—found that CPOEs resulted in a >50% decrease in serious medication errors.³ Other errors have declined with the use of EHR systems, as well; Virapongse et al¹ found a trend towards fewer paid malpractice claims against physicians who used an EHR compared to those physicians using paper charting.

EHR systems may also improve efficiency. In a study of a health maintenance organization (HMO) model, initiating an EHR system improved efficiency by decreasing office visits.⁴ Widespread adoption of EHR systems could save an estimated $81 billion annually through reductions in errors and adverse events, and improved preventive care and chronic disease management.⁵ In a survey of approximately 300 patients who had been evaluated at a family medicine clinic for hypertension, high blood pressure without hypertension, or hyperlipidemia, 75% indicated that they felt EHRs had a positive impact on their care.⁶

Higher clinician gaze time at the patient predicted greater patient satisfaction.

However, some clinicians are concerned about the possible negative impact of EHR systems on health care. One major concern is that EHR systems might increase physician workload⁷ and the amount of time spent using a computer during patient visits. A study that examined physician EHR use found that while time spent on certain tasks, such as prescription writing and lab ordering, was reduced, there was an overall increase in time spent on computer tasks related to charting, preventive care, and chronic disease management.⁸ Baron et al⁹ also found an increase in time spent using the EHR during each clinic session in one private practice setting.

Physicians are also concerned that EHR systems might interfere with the patient-physician interaction (eg, maintaining eye contact, paying attention to patients’ concerns) by directing the physician’s attention away from the patient and toward the computer.¹⁰ In one study, this concern increased after physicians started utilizing a new EHR system.¹¹ Although a survey of inpatients indicated that residents engaged in greater patient-physician communication after an EHR was implemented,¹² a separate study conducted in an outpatient setting found physicians spent less time looking at patients after converting from a paper-based system to an EHR system.¹³

Very few studies have quantitatively examined the association of patient satisfaction with clinician EHR usage. The goal of this study was to examine the correlation of patient satisfaction with actual EHR usage in an ambulatory setting. The data reported in this paper are part of a larger study aimed at understanding EHR use in a VAMC.

METHODS

Study design and sample

The study participants were clinicians in 4 VAMC community clinics located in San Diego, Calif. Twenty-three clinicians (21 general internal medicine physicians and 2 nurse practitioners) were enrolled in the study. Most clinicians identified 5 to 6 patients from their practices to participate in the study (2 participants identified only one patient each). All patients were visiting their clinician for either an acute visit or a follow-up visit.

Although there were slight variations in clinic room size and shape, all rooms were equipped with a compact desk against a wall, a rolling desk chair, a desktop computer with keyboard and mouse, and a second, fixed chair placed diagonal to the physician’s chair. Two rooms had dual monitors. There was a standard examination table in all examination rooms.

The clinicians’ computer use and the patient-clinician interactions in the exam room were captured in real time via video recordings of the interactions and the computer screen. A usability testing software system (Morae) was used to record clinicians’ computer activities, including mouse clicks and scrolls on the computer. The Computerized Patient Records System (CPRS) was the EHR used by all clinicians in this study.

At the end of the visit, patients were asked to complete a satisfaction survey with questions in 3 domains: the physician’s engagement in patient-centered communication, the physician’s clinical skills, and the physician’s interpersonal skills.

Data analysis

Descriptive statistics were used to document patient characteristics, the clinicians’ EHR usage (total number of mouse clicks and scrolls during the visit) and interaction with the patient (gaze time at EHR vs at patient and companion), and to summarize patient satisfaction with the visit. To account for clinician cluster effect, a linear mixed effects model was used to assess the associations between patient satisfaction with the clinician and 2 variables: the amount of clinician time spent viewing or using the computer and the clinician time spent interacting with the patient.

We also assessed the above associations by controlling for visit length. Visit lengths not significant at P<.10 were reported as unadjusted analyses.

All analyses were performed using R statistical software, with a P value of <.05 interpreted as statistically significant.

RESULTS

Satisfaction surveys and video and Morae data were collected for 126 individual patient office visits to the 23 participating physicians and nurses. A majority of the patients who participated in the study were older (mean: 60.5 years; standard deviation [SD]=13.4 years), male as expected in a VA setting (96.8%), Caucasian (65.1%), and had at least some college education (81.7%, TABLE 1).

Patients rated their satisfaction in 3 domains—patient-centered communication, physician clinical skills, and physician interpersonal skills—using a 1 to 5 scale (1=least satisfied, 5=most satisfied). Patients in this study were highly satisfied with their physician or nurse in all 3 domains and overall (TABLE 2), with an average satisfaction score of 4.52 ± 0.51 for patient-centered communication, 4.71 ± 0.56 for physician clinical skills, 4.86 ± 0.32 for physician interpersonal skills, and 4.64 ± 0.38 for total satisfaction.

The physicians and nurses used their EHR system extensively during the visits as delineated by the number of clicks and scrolls on the computer. The average number of clicks and scrolls was 192, with a maximum of 685 clicks and scrolls during one visit. The average visit lasted 30.7 minutes, and on average the clinician spent 12.7 minutes (SD: 8.22 minutes), or an average of about 39.4% of total visit time, viewing or working on the EHR; an average of 10.8 minutes (SD: 5.63 minutes), or an average of about 36.3% of total visit time, was spent interacting with the patient (TABLE 3).

Without adjusting for visit length, patient satisfaction with the clinicians’ patientcentered communication (beta=0.02; P=.02) and total satisfaction (beta=0.014; P=.027) were significantly associated with clinicians’ gaze time at the patient; more clinician gaze time at the patient resulted in greater patient satisfaction (TABLE 4). Adding visit length to the above models had no significant effect (P>.10); therefore, we did not include it in the models.

Patient satisfaction with clinicians’ interpersonal skills was positively associated with gaze time at the patient (beta=0.013, P =.017) without adjusting for visit length. Since the normal assumption of residuals was not plausible based on a normal probability plot, we also assessed the association by dichotomizing the score (5=very satisfied vs <5=not very satisfied) and this significance disappeared. This association was not significant while controlling for visit length.

The percentage of gaze time at the patient (the fraction of patient gaze time over the entire visit) was not significantly associated with patient-centered communication (beta=0.483, P=.12, TABLE 4) when not adjusted for visit length. After adjusting for visit length (P=.052), the association became significant (beta=0.628, P=.033); thus, the higher percentage of time the clinician spent interacting with the patient, the more satisfied the patient was.

DISCUSSION

In this study, patients were highly satisfied with their clinicians despite often high usage of the EHR. Gadd and Penrod¹¹ reported that patients perceived no impact on communication or eye contact with the clinician despite the initiation of an EHR system in 6 large academic medical practices. Another study demonstrated no significant differences in patient satisfaction with their physicians when comparing patients whose physicians used a paper charting system with those who used an EHR system.¹⁴

The average number of computer clicks and scrolls per visit was 192, with a maximum of 685 clicks and scrolls during one visit.

The fact that patients demonstrated high levels of satisfaction with patient-clinician communication even for clinicians with high EHR usage is somewhat surprising. However, Hsu et al¹⁵ found patients’ satisfaction with their clinicians’ communication about medical issues and familiarity with them increased 7 months after implementing an EHR system. In a different study that analyzed videotaped interactions between patients and 5 physicians, the patients found it disturbing not knowing what their doctor was doing when he or she worked on the computer, and preferred being able to see the computer screen.¹⁶ This study suggests that it’s advisable for clinicians to describe what they are doing when they use the computer, so that patients better understand how this time spent inputting data actually benefits them.

EHRs can be time-consuming. Physicians and nurses in our study interacted with the EHR a great deal during the office visit, as evidenced by the large average number of clicks and scrolls. This finding confirms clinicians’ perceptions of the amount of work the EHR system requires. For example, in a semi-structured interview of physicians regarding their use of a VA EHR system,¹⁰ one respondent noted that the reminders in the EHR required hundreds of clicks.

In our study, the average number of clicks and scrolls during the visit was 192, with some clinicians registering hundreds more. In fact, concerns about the time involved in the use of the EHR and about the adequacy of data collection may lead some clinicians who currently don’t have an EHR system to be reluctant to integrate one into their practices.¹⁷

In this study, patients were highly satisfied with their clinicians, despite often high usage of the EHR.

Makoul et al¹⁸ found that compared with physicians who used a paper chart, physicians who used an EHR system were more active in clarifying information from the patient and encouraging patient questions during visits, although the study found a trend toward less active roles in more patient-centered communication when using an EHR system. This latter finding is similar to the concerns raised in our study.

Clinical and communication skills are factors, too. One study found that compared to patients who were cared for by more experienced physicians, patients seen by residents using EHRs were more likely to feel that the physician spent less time talking with them and examining them; they were also more likely to report that the visit felt less personal.¹⁹ Another study found that clinicians with poor baseline communication skills had more difficulties interacting with patients when an EHR system was introduced than those who had better baseline communication skills.²⁰

Training needed to improve communication during EHR use. Research has shown that when used properly and thoughtfully, EHR use can result in greater patient engagement.²¹ But, as noted above, there are challenges, suggesting a need for training clinicians to more successfully use an EHR system while simultaneously communicating with their patients.

Study limitations. This study was conducted at a single site, using a single EHR system deployed in the VA clinics. We cannot generalize our findings to other sites or types of clinic systems. Other EHR systems may have different functionalities, which may affect the time required to provide the same type of medical care.

In addition, the study involved only 23 physicians and nurses in a single health system. Other clinicians may have patterns different from those we studied, although a wide range of patterns was seen among the participants, as demonstrated by the large variation in the number of clicks and scrolls. Another limitation is that study patients were not randomly selected, but rather referred by the provider, and the visits were not blinded to either the provider or patient. This may cause some selection bias.

EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication.

In this study of VA clinicians’ EHR use, patients expressed satisfaction with the clinicians’ clinical skills and patient-centered communication when the clinician spent more time and a greater percentage of the visit engaging the patient. EHR systems need to be designed in a clinician-friendly manner that allows for increased time during the interaction for face-to-face communication between the clinician and the patient, and to ease the workload of EHR documentation. In the meantime, clinicians should be trained in how to expedite their use of the EHR during the clinical visit as well as outside of the exam room in order to improve their patients’ satisfaction.

CORRESPONDENCE
Neil J. Farber, MD, University of California, San Diego, 8939 Villa La Jolla Drive, La Jolla, CA 92037; nfarber@ucsd.edu.

When patients present with neurologic complaints in outpatient primary care practice, 2 key questions often arise: Should brain imaging be ordered, and if so, which study? Careful history-taking and physical exam findings can suggest differential diagnoses and help determine whether imaging studies could identify potential underlying causes. Further considerations in making a decision are the type of information each modality offers, the possible need for contrast media, benefits vs radiation exposure risks, potential contraindications, and cost and local availability. In this article, we present 3 common outpatient scenarios, and in each case we describe the evidence to support clinical decision-making about imaging.

The American College of Radiology (ACR) Appropriateness Criteria Web site (http://www.acr.org/Quality-Safety/Appropriateness-Criteria) provides radiation exposure information, numerical ratings of imaging studies for individual clinical scenarios, evidence tables, and reference tables for each of its recommendations.¹ ACR’s recommendations were developed by expert panels of diagnostic radiologists, interventional radiologists, and radiation oncologists, and designed to help physicians order the most appropriate imaging studies based on patients’ clinical conditions.² We used these criteria to develop an evaluation strategy for each of our clinical scenarios.

CASE 1 › Carrie D is a 45-year-old woman with a history of migraine without aura generally controlled with Excedrin (acetaminophen, aspirin, and caffeine). She arrives at your office with a 2-day history of severe headache over the top of her head and associated tingling sensation over the left side of her face, but with no vision changes, weakness, or slurred speech. She denies any prior history of numbness or tinging with past headaches. She is a business executive and reports that in the last few weeks, her company has been involved in a high-profile merger. On physical exam, her vital signs are within normal limits. She does not appear acutely ill, but on exam she reports diminished sensation to light touch over the left side of her face, left arm, and left leg compared with the right side.

›› What imaging options might you consider?