The Journal of Family Practice

Illustrative case

Your patient, a 42-year-old man with alcohol use disorder (AUD), detoxifies from alcohol during a recent hospitalization. He doesn’t want to resume drinking, but reports frequent cravings. Are there any medications you can prescribe to help prevent relapse?

Excessive alcohol consumption is responsible for 1 of every 10 deaths among US adults ages 20 to 64 years.² Twenty percent to 36% of patients seen in a primary care office have AUD.³ Up to 70% of people who quit with psychosocial support alone will relapse.³

Listen to Shannon Robinson, MD discuss: Sobriety meds—A look at compliance and cost

The US Preventive Services Task Force gives a grade B recommendation to screening all adults for AUD, indicating that physicians should provide this service.⁴ For patients with AUD who wish to abstain but struggle with cravings and relapse, the National Institute on Alcohol Abuse and Alcoholism (NIAAA) recommends considering medication as an adjunct to brief behavioral counseling.⁵

STUDY SUMMARY: Evidence shows naltrexone can prevent a return to drinking

In a meta-analysis, Jonas et al¹ reviewed 123 studies (N=22,803) of pharmacotherapy for AUD. After excluding 28 studies (7 were the only study of a given drug, one was a prospective cohort, and 20 had insufficient data), 95 randomized control trials were included in the analysis. Twenty-two were placebo-controlled for acamprosate (1000-3000 mg/d), 44 for naltrexone (50 mg/d oral, 100 mg/d oral, or injectable) and 4 compared the 2 drugs. Additional studies evaluated disulfiram as well as 23 other off-label medications such as valproic acid and topiramate.

Two investigators independently reviewed the studies, checking for completeness and accuracy. Studies were also analyzed for bias using predefined criteria; those with high or unclear risk of bias were excluded from the main analysis but included in the sensitivity analysis. Funnel plots showed no evidence of publication bias.

Participants were primarily recruited as inpatients and in most studies the mean age was in the 40s. Most patients were diagnosed with alcohol dependence based on criteria in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision (DSM-IV-TR); this diagnosis translates to likely moderate to severe AUD in DSM-5. Prior to starting medications, participants underwent detoxification or achieved at least 3 days of sobriety. Most studies included psychosocial intervention in addition to medication, but the types of interventions varied. The duration of the trials ranged from 12 to 52 weeks.

Researchers analyzed 5 drinking outcomes—return to any drinking, return to heavy drinking (defined as ≥4 drinks/d for women and ≥5 drinks/d for men), number of drinking days, number of heavy drinking days, and drinks per drinking day. They also evaluated health outcomes (accidents, injuries, quality of life, function, and mortality) and adverse effects.

Acamprosate and oral naltrexone (50 mg/d) significantly decreased return to any drinking, with a number needed to treat (NNT) of 12 (95% confidence interval [CI], 8-26) for acamprosate and 20 (95% CI, 11-500) for naltrexone. Oral naltrexone (50 mg/d) also decreased return to heavy drinking (NNT=12; 95% CI, 8-26), while acamprosate did not. Neither medication showed a decrease in heavy drinking days. In a post hoc subgroup analysis of acamprosate for return to any drinking, the drug appeared to be more effective in studies with a higher risk of bias and less effective in studies with a lower risk of bias; the 2 studies with the lowest risk of bias found no significant effect.

Disulfiram had no effect on any of the drinking outcomes analyzed.

Of the off-label medications, topiramate showed a decrease in drinking days (weighted mean difference [WMD]=-6.5%; 95% CI, -12.0% to -1.0%), heavy drinking days (WMD=-9.0%; 95% CI, -15.3% to -2.7%), and drinks per drinking day (WMD=-1.0; 95% CI, -1.6 to -0.48).

Oral naltrexone 50 mg/d significantly decreased the number of patients who resumed drinking after detoxification.

There were no significant differences in health outcomes for any of the medications. Adverse events were greater in treatment groups than placebo groups. Acamprosate was associated with increased risk of diarrhea (number needed to harm [NNH]=11; 95% CI, 6-34), vomiting (NNH=42; 95% CI, 24-143), and anxiety (NNH=7; 95% CI, 5-11). Naltrexone was associated with increased risk of nausea (NNH=9; 95% CI, 7-14), vomiting (NNH=24; 95% CI, 17-44), and dizziness (NNH 16; 95% CI, 12-28).

WHAT'S NEW: Consider prescribing naltrexone to prevent relapse

While previous studies suggested that pharmacotherapy could help patients with AUD remain abstinent, this methodologically rigorous meta-analysis compared the efficacy of several commonly used medications and found clear evidence favoring oral naltrexone. Prescribe oral naltrexone 50 mg/d to help patients with moderate to severe AUD avoid returning to any drinking or heavy drinking after alcohol detoxification. Acamprosate may also decrease return to drinking, although the evidence is not as strong (the studies with low bias showed no effect).

CAVEATS: Medication should be used with psychosocial treatments

Pharmacotherapy for AUD should be reserved for patients who want to quit drinking and used in conjunction with psychosocial intervention.³ Only one of the studies analyzed by Jonas et al¹ was conducted in primary care. That said, many of the psychosocial interventions—such as regular follow-up visits to encourage adherence and monitor for adverse effects, in conjunction with attendance at Alcoholics Anonymous meetings—could be done in primary care settings.

Comorbidities may limit therapy options. Naltrexone is contraindicated in acute hepatitis and liver failure, and in combination with opioids.⁵ Acamprosate is contraindicated in renal disease.⁵

CHALLENGES TO IMPLEMENTATION: Cost, adherence may be factors for some patients 

Perhaps the greatest hurdle in pharmacotherapy for AUD in primary care is a lack of familiarity with these medications. For physicians who are comfortable with prescribing these medications, implementation may be hindered by a lack of available psychosocial resources for successful abstinence.

Medications for alcohol use disorder should be reserved for patients who want to quit drinking, and should be combined with psychosocial interventions.

Additionally, the medications are expensive. The branded version of naltrexone 50 mg costs approximately $118 for a 30-day supply,⁶ and the branded version of acamprosate costs approximately $284 for a 30-day supply.⁷

As is the case with any chronic medical condition, medication adherence is a challenge. Naltrexone is taken once daily, while acamprosate is taken 3 times a day. The risk of relapse is high until 6 to 12 months of sobriety and then wanes over several years.⁵ The NIAAA recommends treatment for a minimum of 3 months.⁵

ACKNOWLEDGEMENT
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Illustrative case

Your patient, a 42-year-old man with alcohol use disorder (AUD), detoxifies from alcohol during a recent hospitalization. He doesn’t want to resume drinking, but reports frequent cravings. Are there any medications you can prescribe to help prevent relapse?

Excessive alcohol consumption is responsible for 1 of every 10 deaths among US adults ages 20 to 64 years.² Twenty percent to 36% of patients seen in a primary care office have AUD.³ Up to 70% of people who quit with psychosocial support alone will relapse.³

Listen to Shannon Robinson, MD discuss: Sobriety meds—A look at compliance and cost

The US Preventive Services Task Force gives a grade B recommendation to screening all adults for AUD, indicating that physicians should provide this service.⁴ For patients with AUD who wish to abstain but struggle with cravings and relapse, the National Institute on Alcohol Abuse and Alcoholism (NIAAA) recommends considering medication as an adjunct to brief behavioral counseling.⁵

STUDY SUMMARY: Evidence shows naltrexone can prevent a return to drinking

In a meta-analysis, Jonas et al¹ reviewed 123 studies (N=22,803) of pharmacotherapy for AUD. After excluding 28 studies (7 were the only study of a given drug, one was a prospective cohort, and 20 had insufficient data), 95 randomized control trials were included in the analysis. Twenty-two were placebo-controlled for acamprosate (1000-3000 mg/d), 44 for naltrexone (50 mg/d oral, 100 mg/d oral, or injectable) and 4 compared the 2 drugs. Additional studies evaluated disulfiram as well as 23 other off-label medications such as valproic acid and topiramate.

Two investigators independently reviewed the studies, checking for completeness and accuracy. Studies were also analyzed for bias using predefined criteria; those with high or unclear risk of bias were excluded from the main analysis but included in the sensitivity analysis. Funnel plots showed no evidence of publication bias.

Participants were primarily recruited as inpatients and in most studies the mean age was in the 40s. Most patients were diagnosed with alcohol dependence based on criteria in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision (DSM-IV-TR); this diagnosis translates to likely moderate to severe AUD in DSM-5. Prior to starting medications, participants underwent detoxification or achieved at least 3 days of sobriety. Most studies included psychosocial intervention in addition to medication, but the types of interventions varied. The duration of the trials ranged from 12 to 52 weeks.

Researchers analyzed 5 drinking outcomes—return to any drinking, return to heavy drinking (defined as ≥4 drinks/d for women and ≥5 drinks/d for men), number of drinking days, number of heavy drinking days, and drinks per drinking day. They also evaluated health outcomes (accidents, injuries, quality of life, function, and mortality) and adverse effects.

Acamprosate and oral naltrexone (50 mg/d) significantly decreased return to any drinking, with a number needed to treat (NNT) of 12 (95% confidence interval [CI], 8-26) for acamprosate and 20 (95% CI, 11-500) for naltrexone. Oral naltrexone (50 mg/d) also decreased return to heavy drinking (NNT=12; 95% CI, 8-26), while acamprosate did not. Neither medication showed a decrease in heavy drinking days. In a post hoc subgroup analysis of acamprosate for return to any drinking, the drug appeared to be more effective in studies with a higher risk of bias and less effective in studies with a lower risk of bias; the 2 studies with the lowest risk of bias found no significant effect.

Disulfiram had no effect on any of the drinking outcomes analyzed.

Of the off-label medications, topiramate showed a decrease in drinking days (weighted mean difference [WMD]=-6.5%; 95% CI, -12.0% to -1.0%), heavy drinking days (WMD=-9.0%; 95% CI, -15.3% to -2.7%), and drinks per drinking day (WMD=-1.0; 95% CI, -1.6 to -0.48).

Oral naltrexone 50 mg/d significantly decreased the number of patients who resumed drinking after detoxification.

There were no significant differences in health outcomes for any of the medications. Adverse events were greater in treatment groups than placebo groups. Acamprosate was associated with increased risk of diarrhea (number needed to harm [NNH]=11; 95% CI, 6-34), vomiting (NNH=42; 95% CI, 24-143), and anxiety (NNH=7; 95% CI, 5-11). Naltrexone was associated with increased risk of nausea (NNH=9; 95% CI, 7-14), vomiting (NNH=24; 95% CI, 17-44), and dizziness (NNH 16; 95% CI, 12-28).

WHAT'S NEW: Consider prescribing naltrexone to prevent relapse

While previous studies suggested that pharmacotherapy could help patients with AUD remain abstinent, this methodologically rigorous meta-analysis compared the efficacy of several commonly used medications and found clear evidence favoring oral naltrexone. Prescribe oral naltrexone 50 mg/d to help patients with moderate to severe AUD avoid returning to any drinking or heavy drinking after alcohol detoxification. Acamprosate may also decrease return to drinking, although the evidence is not as strong (the studies with low bias showed no effect).

CAVEATS: Medication should be used with psychosocial treatments

Pharmacotherapy for AUD should be reserved for patients who want to quit drinking and used in conjunction with psychosocial intervention.³ Only one of the studies analyzed by Jonas et al¹ was conducted in primary care. That said, many of the psychosocial interventions—such as regular follow-up visits to encourage adherence and monitor for adverse effects, in conjunction with attendance at Alcoholics Anonymous meetings—could be done in primary care settings.

Comorbidities may limit therapy options. Naltrexone is contraindicated in acute hepatitis and liver failure, and in combination with opioids.⁵ Acamprosate is contraindicated in renal disease.⁵

CHALLENGES TO IMPLEMENTATION: Cost, adherence may be factors for some patients 

Perhaps the greatest hurdle in pharmacotherapy for AUD in primary care is a lack of familiarity with these medications. For physicians who are comfortable with prescribing these medications, implementation may be hindered by a lack of available psychosocial resources for successful abstinence.

Medications for alcohol use disorder should be reserved for patients who want to quit drinking, and should be combined with psychosocial interventions.

Additionally, the medications are expensive. The branded version of naltrexone 50 mg costs approximately $118 for a 30-day supply,⁶ and the branded version of acamprosate costs approximately $284 for a 30-day supply.⁷

As is the case with any chronic medical condition, medication adherence is a challenge. Naltrexone is taken once daily, while acamprosate is taken 3 times a day. The risk of relapse is high until 6 to 12 months of sobriety and then wanes over several years.⁵ The NIAAA recommends treatment for a minimum of 3 months.⁵

ACKNOWLEDGEMENT
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Illustrative case

Your patient, a 42-year-old man with alcohol use disorder (AUD), detoxifies from alcohol during a recent hospitalization. He doesn’t want to resume drinking, but reports frequent cravings. Are there any medications you can prescribe to help prevent relapse?

Excessive alcohol consumption is responsible for 1 of every 10 deaths among US adults ages 20 to 64 years.² Twenty percent to 36% of patients seen in a primary care office have AUD.³ Up to 70% of people who quit with psychosocial support alone will relapse.³

Listen to Shannon Robinson, MD discuss: Sobriety meds—A look at compliance and cost

The US Preventive Services Task Force gives a grade B recommendation to screening all adults for AUD, indicating that physicians should provide this service.⁴ For patients with AUD who wish to abstain but struggle with cravings and relapse, the National Institute on Alcohol Abuse and Alcoholism (NIAAA) recommends considering medication as an adjunct to brief behavioral counseling.⁵

STUDY SUMMARY: Evidence shows naltrexone can prevent a return to drinking

In a meta-analysis, Jonas et al¹ reviewed 123 studies (N=22,803) of pharmacotherapy for AUD. After excluding 28 studies (7 were the only study of a given drug, one was a prospective cohort, and 20 had insufficient data), 95 randomized control trials were included in the analysis. Twenty-two were placebo-controlled for acamprosate (1000-3000 mg/d), 44 for naltrexone (50 mg/d oral, 100 mg/d oral, or injectable) and 4 compared the 2 drugs. Additional studies evaluated disulfiram as well as 23 other off-label medications such as valproic acid and topiramate.

Two investigators independently reviewed the studies, checking for completeness and accuracy. Studies were also analyzed for bias using predefined criteria; those with high or unclear risk of bias were excluded from the main analysis but included in the sensitivity analysis. Funnel plots showed no evidence of publication bias.

Participants were primarily recruited as inpatients and in most studies the mean age was in the 40s. Most patients were diagnosed with alcohol dependence based on criteria in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision (DSM-IV-TR); this diagnosis translates to likely moderate to severe AUD in DSM-5. Prior to starting medications, participants underwent detoxification or achieved at least 3 days of sobriety. Most studies included psychosocial intervention in addition to medication, but the types of interventions varied. The duration of the trials ranged from 12 to 52 weeks.

Researchers analyzed 5 drinking outcomes—return to any drinking, return to heavy drinking (defined as ≥4 drinks/d for women and ≥5 drinks/d for men), number of drinking days, number of heavy drinking days, and drinks per drinking day. They also evaluated health outcomes (accidents, injuries, quality of life, function, and mortality) and adverse effects.

Acamprosate and oral naltrexone (50 mg/d) significantly decreased return to any drinking, with a number needed to treat (NNT) of 12 (95% confidence interval [CI], 8-26) for acamprosate and 20 (95% CI, 11-500) for naltrexone. Oral naltrexone (50 mg/d) also decreased return to heavy drinking (NNT=12; 95% CI, 8-26), while acamprosate did not. Neither medication showed a decrease in heavy drinking days. In a post hoc subgroup analysis of acamprosate for return to any drinking, the drug appeared to be more effective in studies with a higher risk of bias and less effective in studies with a lower risk of bias; the 2 studies with the lowest risk of bias found no significant effect.

Disulfiram had no effect on any of the drinking outcomes analyzed.

Of the off-label medications, topiramate showed a decrease in drinking days (weighted mean difference [WMD]=-6.5%; 95% CI, -12.0% to -1.0%), heavy drinking days (WMD=-9.0%; 95% CI, -15.3% to -2.7%), and drinks per drinking day (WMD=-1.0; 95% CI, -1.6 to -0.48).

Oral naltrexone 50 mg/d significantly decreased the number of patients who resumed drinking after detoxification.

There were no significant differences in health outcomes for any of the medications. Adverse events were greater in treatment groups than placebo groups. Acamprosate was associated with increased risk of diarrhea (number needed to harm [NNH]=11; 95% CI, 6-34), vomiting (NNH=42; 95% CI, 24-143), and anxiety (NNH=7; 95% CI, 5-11). Naltrexone was associated with increased risk of nausea (NNH=9; 95% CI, 7-14), vomiting (NNH=24; 95% CI, 17-44), and dizziness (NNH 16; 95% CI, 12-28).

WHAT'S NEW: Consider prescribing naltrexone to prevent relapse

While previous studies suggested that pharmacotherapy could help patients with AUD remain abstinent, this methodologically rigorous meta-analysis compared the efficacy of several commonly used medications and found clear evidence favoring oral naltrexone. Prescribe oral naltrexone 50 mg/d to help patients with moderate to severe AUD avoid returning to any drinking or heavy drinking after alcohol detoxification. Acamprosate may also decrease return to drinking, although the evidence is not as strong (the studies with low bias showed no effect).

CAVEATS: Medication should be used with psychosocial treatments

Pharmacotherapy for AUD should be reserved for patients who want to quit drinking and used in conjunction with psychosocial intervention.³ Only one of the studies analyzed by Jonas et al¹ was conducted in primary care. That said, many of the psychosocial interventions—such as regular follow-up visits to encourage adherence and monitor for adverse effects, in conjunction with attendance at Alcoholics Anonymous meetings—could be done in primary care settings.

Comorbidities may limit therapy options. Naltrexone is contraindicated in acute hepatitis and liver failure, and in combination with opioids.⁵ Acamprosate is contraindicated in renal disease.⁵

CHALLENGES TO IMPLEMENTATION: Cost, adherence may be factors for some patients 

Perhaps the greatest hurdle in pharmacotherapy for AUD in primary care is a lack of familiarity with these medications. For physicians who are comfortable with prescribing these medications, implementation may be hindered by a lack of available psychosocial resources for successful abstinence.

Medications for alcohol use disorder should be reserved for patients who want to quit drinking, and should be combined with psychosocial interventions.

Additionally, the medications are expensive. The branded version of naltrexone 50 mg costs approximately $118 for a 30-day supply,⁶ and the branded version of acamprosate costs approximately $284 for a 30-day supply.⁷

As is the case with any chronic medical condition, medication adherence is a challenge. Naltrexone is taken once daily, while acamprosate is taken 3 times a day. The risk of relapse is high until 6 to 12 months of sobriety and then wanes over several years.⁵ The NIAAA recommends treatment for a minimum of 3 months.⁵

ACKNOWLEDGEMENT
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Failure to recognize impending MI has tragic consequences

A 55-YEAR-OLD WOMAN WENT TO HER MEDICAL CLINIC because she had heartburn and bilateral arm pain with numbness and tingling in her forearms. She said she’d had intermittent arm pain over the previous 7 to 10 days. A physician’s assistant diagnosed gastroesophageal reflux disease, gave the patient an antacid medication, and instructed her to return in 2 to 3 weeks. The patient came back to the clinic 10 days later with increased heartburn and continued arm pain with tingling. Because no clinicians were available to see her at that time, a prescription for ranitidine was called in and the patient was sent home. That evening, the patient died of a myocardial infarction (MI).

PLAINTIFF’S CLAIM There were specific, objective signs of an impending MI that were not recognized.

The patient should have been seen by a medical provider on the day of her death or referred to an emergency department.

THE DEFENSE No information about the defense is available.

VERDICT $275,000 California settlement.

COMMENT There was clearly an opportunity to make the correct diagnosis for this woman, especially when she returned a second time. The one lesson I have learned from reviewing malpractice cases for 15 years is that if a patient returns unimproved, you must up the ante with the evaluation. Start all over again and think through the entire history very carefully; you are likely to find a clue to the correct diagnosis.

Pulmonary embolism mistaken for respiratory infection

A 40-YEAR-OLD MAN SOUGHT TREATMENT FOR SYMPTOMS OF A COLD. He also complained of shortness of breath, dizziness, and pain in his left calf. His family physician (FP) treated him for a respiratory infection. Three days later, the patient returned to the office with continued shortness of breath. The FP scheduled a cardiac work-up. Two days before the work-up, the patient died from a pulmonary embolism (PE).

PLAINTIFF'S CLAIM No information about the plaintiff’s claim is available.

If a patient returns unimproved, start the evaluation over again and think through the entire history; you’ll likely find a clue to the correct diagnosis.

THE DEFENSE No information about the defense is available.

VERDICT $1.1 million Virginia settlement.

COMMENT PE has clearly unseated syphilis as “The Great Masquerader.” We cannot tell from this short synopsis how significant the patient’s calf pain was and whether or not there were any physical findings of deep vein thrombosis. However, when the patient returned 3 days later with increasing shortness of breath, PE should have been toward the top of the differential diagnosis.

Back spasms—or something far more serious?

A 47-YEAR-OLD WOMAN WENT TO THE EMERGENCY DEPARTMENT (ED) seeking treatment for severe back and abdominal pain. The patient had previously undergone gastric bypass surgery. The ED physician diagnosed back spasms, but admitted her to the hospital for observation. The next day, the patient died from a bowel obstruction.

PLAINTIFF'S CLAIM The ED physician failed to order testing and consult with a specialist to diagnose bowel obstruction, which is a known complication of gastric bypass surgery.

THE DEFENSE No information about the defense is available.

VERDICT $2.4 million Illinois verdict.

COMMENT Bowel obstruction with back pain only? And dead the next day from bowel obstruction? I can only presume the history was inadequate, which led to a failure to do an abdominal exam.

Failure to recognize impending MI has tragic consequences

A 55-YEAR-OLD WOMAN WENT TO HER MEDICAL CLINIC because she had heartburn and bilateral arm pain with numbness and tingling in her forearms. She said she’d had intermittent arm pain over the previous 7 to 10 days. A physician’s assistant diagnosed gastroesophageal reflux disease, gave the patient an antacid medication, and instructed her to return in 2 to 3 weeks. The patient came back to the clinic 10 days later with increased heartburn and continued arm pain with tingling. Because no clinicians were available to see her at that time, a prescription for ranitidine was called in and the patient was sent home. That evening, the patient died of a myocardial infarction (MI).

PLAINTIFF’S CLAIM There were specific, objective signs of an impending MI that were not recognized.

The patient should have been seen by a medical provider on the day of her death or referred to an emergency department.

THE DEFENSE No information about the defense is available.

VERDICT $275,000 California settlement.

COMMENT There was clearly an opportunity to make the correct diagnosis for this woman, especially when she returned a second time. The one lesson I have learned from reviewing malpractice cases for 15 years is that if a patient returns unimproved, you must up the ante with the evaluation. Start all over again and think through the entire history very carefully; you are likely to find a clue to the correct diagnosis.

Pulmonary embolism mistaken for respiratory infection

A 40-YEAR-OLD MAN SOUGHT TREATMENT FOR SYMPTOMS OF A COLD. He also complained of shortness of breath, dizziness, and pain in his left calf. His family physician (FP) treated him for a respiratory infection. Three days later, the patient returned to the office with continued shortness of breath. The FP scheduled a cardiac work-up. Two days before the work-up, the patient died from a pulmonary embolism (PE).

PLAINTIFF'S CLAIM No information about the plaintiff’s claim is available.

If a patient returns unimproved, start the evaluation over again and think through the entire history; you’ll likely find a clue to the correct diagnosis.

THE DEFENSE No information about the defense is available.

VERDICT $1.1 million Virginia settlement.

COMMENT PE has clearly unseated syphilis as “The Great Masquerader.” We cannot tell from this short synopsis how significant the patient’s calf pain was and whether or not there were any physical findings of deep vein thrombosis. However, when the patient returned 3 days later with increasing shortness of breath, PE should have been toward the top of the differential diagnosis.

Back spasms—or something far more serious?

A 47-YEAR-OLD WOMAN WENT TO THE EMERGENCY DEPARTMENT (ED) seeking treatment for severe back and abdominal pain. The patient had previously undergone gastric bypass surgery. The ED physician diagnosed back spasms, but admitted her to the hospital for observation. The next day, the patient died from a bowel obstruction.

PLAINTIFF'S CLAIM The ED physician failed to order testing and consult with a specialist to diagnose bowel obstruction, which is a known complication of gastric bypass surgery.

THE DEFENSE No information about the defense is available.

VERDICT $2.4 million Illinois verdict.

COMMENT Bowel obstruction with back pain only? And dead the next day from bowel obstruction? I can only presume the history was inadequate, which led to a failure to do an abdominal exam.

Failure to recognize impending MI has tragic consequences

A 55-YEAR-OLD WOMAN WENT TO HER MEDICAL CLINIC because she had heartburn and bilateral arm pain with numbness and tingling in her forearms. She said she’d had intermittent arm pain over the previous 7 to 10 days. A physician’s assistant diagnosed gastroesophageal reflux disease, gave the patient an antacid medication, and instructed her to return in 2 to 3 weeks. The patient came back to the clinic 10 days later with increased heartburn and continued arm pain with tingling. Because no clinicians were available to see her at that time, a prescription for ranitidine was called in and the patient was sent home. That evening, the patient died of a myocardial infarction (MI).

PLAINTIFF’S CLAIM There were specific, objective signs of an impending MI that were not recognized.

The patient should have been seen by a medical provider on the day of her death or referred to an emergency department.

THE DEFENSE No information about the defense is available.

VERDICT $275,000 California settlement.

COMMENT There was clearly an opportunity to make the correct diagnosis for this woman, especially when she returned a second time. The one lesson I have learned from reviewing malpractice cases for 15 years is that if a patient returns unimproved, you must up the ante with the evaluation. Start all over again and think through the entire history very carefully; you are likely to find a clue to the correct diagnosis.

Pulmonary embolism mistaken for respiratory infection

A 40-YEAR-OLD MAN SOUGHT TREATMENT FOR SYMPTOMS OF A COLD. He also complained of shortness of breath, dizziness, and pain in his left calf. His family physician (FP) treated him for a respiratory infection. Three days later, the patient returned to the office with continued shortness of breath. The FP scheduled a cardiac work-up. Two days before the work-up, the patient died from a pulmonary embolism (PE).

PLAINTIFF'S CLAIM No information about the plaintiff’s claim is available.

If a patient returns unimproved, start the evaluation over again and think through the entire history; you’ll likely find a clue to the correct diagnosis.

THE DEFENSE No information about the defense is available.

VERDICT $1.1 million Virginia settlement.

COMMENT PE has clearly unseated syphilis as “The Great Masquerader.” We cannot tell from this short synopsis how significant the patient’s calf pain was and whether or not there were any physical findings of deep vein thrombosis. However, when the patient returned 3 days later with increasing shortness of breath, PE should have been toward the top of the differential diagnosis.

Back spasms—or something far more serious?

A 47-YEAR-OLD WOMAN WENT TO THE EMERGENCY DEPARTMENT (ED) seeking treatment for severe back and abdominal pain. The patient had previously undergone gastric bypass surgery. The ED physician diagnosed back spasms, but admitted her to the hospital for observation. The next day, the patient died from a bowel obstruction.

PLAINTIFF'S CLAIM The ED physician failed to order testing and consult with a specialist to diagnose bowel obstruction, which is a known complication of gastric bypass surgery.

THE DEFENSE No information about the defense is available.

VERDICT $2.4 million Illinois verdict.

COMMENT Bowel obstruction with back pain only? And dead the next day from bowel obstruction? I can only presume the history was inadequate, which led to a failure to do an abdominal exam.

THE CASE

A 21-year-old college student was referred to us by the counseling center at our university for a psychiatric evaluation after 11 psychotherapy sessions over 3 months had failed to reduce her feelings of anxiety and panic.

During our evaluation, the patient described feeling “not quite right” for many months. She had been experiencing mental fogginess, fatigue, and worsening concentration/ memory. Her anxiety, which had been gradually increasing, was the result of being unsure about her gait. She first noticed this while walking down some bleachers; she felt dizzy, was afraid of falling, and couldn’t walk down without assistance. All episodes of “panic” occurred in situations where she experienced disequilibrium, unsteady gait, and fear of falling. She grew fearful of driving or going anywhere without assistance.

The patient had celiac disease that was well controlled with a gluten-free diet. She had no personal or family psychiatric history and no history of substance abuse.

THE DIAGNOSIS

Physical exam and lab studies, including a complete blood count, comprehensive metabolic panel, and thyrotropin and folate levels, were normal. Her homocysteine level was 11.8 μmol/L (reference range, 5.4-11.9 μmol/L) and vitamin B₁₂ level was 292 pg/mL (reference range, 200-1100 pg/mL). Her lab report included a note that read, “Although the reference range for vitamin B₁₂ is 200 to 1100 pg/mL, it has been reported that between 5% and 10% of patients with values between 200 and 400 pg/mL may experience neuropsychiatric and hematologic abnormalities due to occult B₁₂ deficiency; <1% of patients with values >400 pg/mL will have symptoms.”

Based on this vitamin B₁₂ level, the patient’s symptoms, and her borderline high homocysteine level, we diagnosed vitamin B₁₂ deficiency.

DISCUSSION

There are no recommendations by the US Preventive Services Task Force or any other major US medical society for routine vitamin B₁₂ screening.¹ In Canada, the Medical Services Commission of the British Columbia Ministry of Health recommends B₁₂ screening for patients who present with macrocytic anemia or unexplained neurologic symptoms (eg, paresthesia, numbness, poor motor coordination, memory lapses, or cognitive or personality changes).²

Vitamin B₁₂ deficiency can be caused by numerous conditions, including those that cause malabsorption (such as gastric bypass). It can also be caused by diseases such as human immunodeficiency virus infection or Crohn’s disease, long-term adherence to a vegetarian or vegan diet, or by any other lack of dietary intake.¹ The condition can cause hematologic-related signs and symptoms such as megaloblastic anemia, fatigue, and syncope. It also can have neurologic manifestations, including paresthesia, weakness, motor disturbances (including gait abnormalities), vision loss, and a wide range of cognitive and behavioral changes.¹ Anemia is uncommon because since 1998, the US Food and Drug Administration has required fortification of all enriched grain and cereal products with folic acid; thus, vitamin B₁₂ deficiency may proceed without anemia revealing its presence.¹

A controversial topic. Vitamin B₁₂ deficiency is a complicated and controversial subject. Specifically, there is uncertainty about the clinical importance of lower serum levels of vitamin B₁₂ (200-400 pg/mL), their impact on well-being, and the need for treatment. In addition to measuring a patient’s serum B₁₂ level, testing a second biomarker (such as homocysteine or methylmalonic acid) can be helpful in establishing a diagnosis of B₁₂ deficiency.¹ Levels of each of these are elevated in patients with B₁₂ deficiency.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety.

Although vitamin B₁₂ deficiency has been well studied in older patients,³ little has been published about the condition in young adults. National Health and Nutrition Examination Survey (NHANES) data from 2000 to 2004 shows that almost 40% of people ages 19 to 30 years have a B₁₂ level <400 pg/mL.¹ How many of these individuals are at risk of complications of B₁₂ deficiency is unknown.

B₁₂ supplementation might improve depression, anxiety

B₁₂ supplementation is inexpensive and has no significant adverse effects.¹ It can be administered orally, parenterally (intramuscularly or subcutaneously), or intranasally.¹ A common oral regimen is 1 mg/d; parental regimens vary widely, but might include a 1-mg injection once a week for 8 weeks, then once a month for life.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety. A Pakistani study randomized 73 patients with depression and “low normal” B₁₂ levels (190-300 pg/mL) to an antidepressant only (equivalent to imipramine 100-250 mg/d or fluoxetine 20-40 mg/d) or an antidepressant plus parenteral B₁₂ (1000 mcg once a week).⁴ At 3 months follow-up, 100% of the treatment group showed at least a 20% reduction in their Hamilton Depression Rating Scale (HAM-D) score, compared to 69% in the control arm (P<.001).⁴

A Swedish study analyzed the effects of several B vitamins, including 0.5 mg/d of B₁₂ vs placebo on mood in 65 celiac patients on a gluten-free diet who had borderline/low normal B₁₂ levels (>191 pg/mL).⁵ Patients who scored low on a measure of psychological well-being at the beginning of the study and who received B₁₂ experienced significant improvements in anxiety and depressed mood compared to those who received placebo.

Our patient

Because neurologic and psychiatric symptoms require assured compliance and urgent treatment, our patient received vitamin B₁₂ parenterally as cyanocobalamin 1 mg/mL. She was given this dosage intramuscularly once a day for 5 days, then once a week for 4 weeks. She will continue to receive it once a month indefinitely.

The patient was advised that if she wished to switch to oral therapy, she could do so after several months of parenteral treatment, as long as she had close follow-up with frequent B₁₂ measurements to assure that she was absorbing oral therapy. Her anxiety and mood symptoms resolved within one month, and her disequilibrium was almost entirely resolved within 3 months of treatment.

THE TAKEAWAY

Although more common in older patients, vitamin B₁₂ deficiency can also affect younger patients. “Low normal” B₁₂ levels (200-400 pg/mL) may affect psychological well-being.

Consider testing serum B₁₂ and a second biomarker—such as homocysteine or methylmalonic acid, if indicated—in patients who present with depressed mood, anxiety, cognitive symptoms, and/or fatigue. Vitamin B₁₂ supplementation can be administered orally and has no major adverse effects.

THE CASE

A 21-year-old college student was referred to us by the counseling center at our university for a psychiatric evaluation after 11 psychotherapy sessions over 3 months had failed to reduce her feelings of anxiety and panic.

During our evaluation, the patient described feeling “not quite right” for many months. She had been experiencing mental fogginess, fatigue, and worsening concentration/ memory. Her anxiety, which had been gradually increasing, was the result of being unsure about her gait. She first noticed this while walking down some bleachers; she felt dizzy, was afraid of falling, and couldn’t walk down without assistance. All episodes of “panic” occurred in situations where she experienced disequilibrium, unsteady gait, and fear of falling. She grew fearful of driving or going anywhere without assistance.

The patient had celiac disease that was well controlled with a gluten-free diet. She had no personal or family psychiatric history and no history of substance abuse.

THE DIAGNOSIS

Physical exam and lab studies, including a complete blood count, comprehensive metabolic panel, and thyrotropin and folate levels, were normal. Her homocysteine level was 11.8 μmol/L (reference range, 5.4-11.9 μmol/L) and vitamin B₁₂ level was 292 pg/mL (reference range, 200-1100 pg/mL). Her lab report included a note that read, “Although the reference range for vitamin B₁₂ is 200 to 1100 pg/mL, it has been reported that between 5% and 10% of patients with values between 200 and 400 pg/mL may experience neuropsychiatric and hematologic abnormalities due to occult B₁₂ deficiency; <1% of patients with values >400 pg/mL will have symptoms.”

Based on this vitamin B₁₂ level, the patient’s symptoms, and her borderline high homocysteine level, we diagnosed vitamin B₁₂ deficiency.

DISCUSSION

There are no recommendations by the US Preventive Services Task Force or any other major US medical society for routine vitamin B₁₂ screening.¹ In Canada, the Medical Services Commission of the British Columbia Ministry of Health recommends B₁₂ screening for patients who present with macrocytic anemia or unexplained neurologic symptoms (eg, paresthesia, numbness, poor motor coordination, memory lapses, or cognitive or personality changes).²

Vitamin B₁₂ deficiency can be caused by numerous conditions, including those that cause malabsorption (such as gastric bypass). It can also be caused by diseases such as human immunodeficiency virus infection or Crohn’s disease, long-term adherence to a vegetarian or vegan diet, or by any other lack of dietary intake.¹ The condition can cause hematologic-related signs and symptoms such as megaloblastic anemia, fatigue, and syncope. It also can have neurologic manifestations, including paresthesia, weakness, motor disturbances (including gait abnormalities), vision loss, and a wide range of cognitive and behavioral changes.¹ Anemia is uncommon because since 1998, the US Food and Drug Administration has required fortification of all enriched grain and cereal products with folic acid; thus, vitamin B₁₂ deficiency may proceed without anemia revealing its presence.¹

A controversial topic. Vitamin B₁₂ deficiency is a complicated and controversial subject. Specifically, there is uncertainty about the clinical importance of lower serum levels of vitamin B₁₂ (200-400 pg/mL), their impact on well-being, and the need for treatment. In addition to measuring a patient’s serum B₁₂ level, testing a second biomarker (such as homocysteine or methylmalonic acid) can be helpful in establishing a diagnosis of B₁₂ deficiency.¹ Levels of each of these are elevated in patients with B₁₂ deficiency.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety.

Although vitamin B₁₂ deficiency has been well studied in older patients,³ little has been published about the condition in young adults. National Health and Nutrition Examination Survey (NHANES) data from 2000 to 2004 shows that almost 40% of people ages 19 to 30 years have a B₁₂ level <400 pg/mL.¹ How many of these individuals are at risk of complications of B₁₂ deficiency is unknown.

B₁₂ supplementation might improve depression, anxiety

B₁₂ supplementation is inexpensive and has no significant adverse effects.¹ It can be administered orally, parenterally (intramuscularly or subcutaneously), or intranasally.¹ A common oral regimen is 1 mg/d; parental regimens vary widely, but might include a 1-mg injection once a week for 8 weeks, then once a month for life.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety. A Pakistani study randomized 73 patients with depression and “low normal” B₁₂ levels (190-300 pg/mL) to an antidepressant only (equivalent to imipramine 100-250 mg/d or fluoxetine 20-40 mg/d) or an antidepressant plus parenteral B₁₂ (1000 mcg once a week).⁴ At 3 months follow-up, 100% of the treatment group showed at least a 20% reduction in their Hamilton Depression Rating Scale (HAM-D) score, compared to 69% in the control arm (P<.001).⁴

A Swedish study analyzed the effects of several B vitamins, including 0.5 mg/d of B₁₂ vs placebo on mood in 65 celiac patients on a gluten-free diet who had borderline/low normal B₁₂ levels (>191 pg/mL).⁵ Patients who scored low on a measure of psychological well-being at the beginning of the study and who received B₁₂ experienced significant improvements in anxiety and depressed mood compared to those who received placebo.

Our patient

Because neurologic and psychiatric symptoms require assured compliance and urgent treatment, our patient received vitamin B₁₂ parenterally as cyanocobalamin 1 mg/mL. She was given this dosage intramuscularly once a day for 5 days, then once a week for 4 weeks. She will continue to receive it once a month indefinitely.

The patient was advised that if she wished to switch to oral therapy, she could do so after several months of parenteral treatment, as long as she had close follow-up with frequent B₁₂ measurements to assure that she was absorbing oral therapy. Her anxiety and mood symptoms resolved within one month, and her disequilibrium was almost entirely resolved within 3 months of treatment.

THE TAKEAWAY

Although more common in older patients, vitamin B₁₂ deficiency can also affect younger patients. “Low normal” B₁₂ levels (200-400 pg/mL) may affect psychological well-being.

Consider testing serum B₁₂ and a second biomarker—such as homocysteine or methylmalonic acid, if indicated—in patients who present with depressed mood, anxiety, cognitive symptoms, and/or fatigue. Vitamin B₁₂ supplementation can be administered orally and has no major adverse effects.

THE CASE

A 21-year-old college student was referred to us by the counseling center at our university for a psychiatric evaluation after 11 psychotherapy sessions over 3 months had failed to reduce her feelings of anxiety and panic.

During our evaluation, the patient described feeling “not quite right” for many months. She had been experiencing mental fogginess, fatigue, and worsening concentration/ memory. Her anxiety, which had been gradually increasing, was the result of being unsure about her gait. She first noticed this while walking down some bleachers; she felt dizzy, was afraid of falling, and couldn’t walk down without assistance. All episodes of “panic” occurred in situations where she experienced disequilibrium, unsteady gait, and fear of falling. She grew fearful of driving or going anywhere without assistance.

The patient had celiac disease that was well controlled with a gluten-free diet. She had no personal or family psychiatric history and no history of substance abuse.

THE DIAGNOSIS

Physical exam and lab studies, including a complete blood count, comprehensive metabolic panel, and thyrotropin and folate levels, were normal. Her homocysteine level was 11.8 μmol/L (reference range, 5.4-11.9 μmol/L) and vitamin B₁₂ level was 292 pg/mL (reference range, 200-1100 pg/mL). Her lab report included a note that read, “Although the reference range for vitamin B₁₂ is 200 to 1100 pg/mL, it has been reported that between 5% and 10% of patients with values between 200 and 400 pg/mL may experience neuropsychiatric and hematologic abnormalities due to occult B₁₂ deficiency; <1% of patients with values >400 pg/mL will have symptoms.”

Based on this vitamin B₁₂ level, the patient’s symptoms, and her borderline high homocysteine level, we diagnosed vitamin B₁₂ deficiency.

DISCUSSION

There are no recommendations by the US Preventive Services Task Force or any other major US medical society for routine vitamin B₁₂ screening.¹ In Canada, the Medical Services Commission of the British Columbia Ministry of Health recommends B₁₂ screening for patients who present with macrocytic anemia or unexplained neurologic symptoms (eg, paresthesia, numbness, poor motor coordination, memory lapses, or cognitive or personality changes).²

Vitamin B₁₂ deficiency can be caused by numerous conditions, including those that cause malabsorption (such as gastric bypass). It can also be caused by diseases such as human immunodeficiency virus infection or Crohn’s disease, long-term adherence to a vegetarian or vegan diet, or by any other lack of dietary intake.¹ The condition can cause hematologic-related signs and symptoms such as megaloblastic anemia, fatigue, and syncope. It also can have neurologic manifestations, including paresthesia, weakness, motor disturbances (including gait abnormalities), vision loss, and a wide range of cognitive and behavioral changes.¹ Anemia is uncommon because since 1998, the US Food and Drug Administration has required fortification of all enriched grain and cereal products with folic acid; thus, vitamin B₁₂ deficiency may proceed without anemia revealing its presence.¹

A controversial topic. Vitamin B₁₂ deficiency is a complicated and controversial subject. Specifically, there is uncertainty about the clinical importance of lower serum levels of vitamin B₁₂ (200-400 pg/mL), their impact on well-being, and the need for treatment. In addition to measuring a patient’s serum B₁₂ level, testing a second biomarker (such as homocysteine or methylmalonic acid) can be helpful in establishing a diagnosis of B₁₂ deficiency.¹ Levels of each of these are elevated in patients with B₁₂ deficiency.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety.

Although vitamin B₁₂ deficiency has been well studied in older patients,³ little has been published about the condition in young adults. National Health and Nutrition Examination Survey (NHANES) data from 2000 to 2004 shows that almost 40% of people ages 19 to 30 years have a B₁₂ level <400 pg/mL.¹ How many of these individuals are at risk of complications of B₁₂ deficiency is unknown.

B₁₂ supplementation might improve depression, anxiety

B₁₂ supplementation is inexpensive and has no significant adverse effects.¹ It can be administered orally, parenterally (intramuscularly or subcutaneously), or intranasally.¹ A common oral regimen is 1 mg/d; parental regimens vary widely, but might include a 1-mg injection once a week for 8 weeks, then once a month for life.¹

Some evidence suggests B₁₂ supplementation may improve symptoms of depression and anxiety. A Pakistani study randomized 73 patients with depression and “low normal” B₁₂ levels (190-300 pg/mL) to an antidepressant only (equivalent to imipramine 100-250 mg/d or fluoxetine 20-40 mg/d) or an antidepressant plus parenteral B₁₂ (1000 mcg once a week).⁴ At 3 months follow-up, 100% of the treatment group showed at least a 20% reduction in their Hamilton Depression Rating Scale (HAM-D) score, compared to 69% in the control arm (P<.001).⁴

A Swedish study analyzed the effects of several B vitamins, including 0.5 mg/d of B₁₂ vs placebo on mood in 65 celiac patients on a gluten-free diet who had borderline/low normal B₁₂ levels (>191 pg/mL).⁵ Patients who scored low on a measure of psychological well-being at the beginning of the study and who received B₁₂ experienced significant improvements in anxiety and depressed mood compared to those who received placebo.

Our patient

Because neurologic and psychiatric symptoms require assured compliance and urgent treatment, our patient received vitamin B₁₂ parenterally as cyanocobalamin 1 mg/mL. She was given this dosage intramuscularly once a day for 5 days, then once a week for 4 weeks. She will continue to receive it once a month indefinitely.

The patient was advised that if she wished to switch to oral therapy, she could do so after several months of parenteral treatment, as long as she had close follow-up with frequent B₁₂ measurements to assure that she was absorbing oral therapy. Her anxiety and mood symptoms resolved within one month, and her disequilibrium was almost entirely resolved within 3 months of treatment.

THE TAKEAWAY

Although more common in older patients, vitamin B₁₂ deficiency can also affect younger patients. “Low normal” B₁₂ levels (200-400 pg/mL) may affect psychological well-being.

Consider testing serum B₁₂ and a second biomarker—such as homocysteine or methylmalonic acid, if indicated—in patients who present with depressed mood, anxiety, cognitive symptoms, and/or fatigue. Vitamin B₁₂ supplementation can be administered orally and has no major adverse effects.

EVIDENCE SUMMARY

A meta-analysis of 13 randomized, placebo or standard of care-controlled statin trials (113,148 patients, 81% without diabetes at enrollment, mean ages 55-76 years) found that statin therapy increased the incidence of diabetes by 9% over 4 years (odds ratio [OR]=1.09; 95% confidence interval [CI], 1.02-1.17), or one additional case per 255 patients.¹ The increased risk was similar for lipophilic (pravastatin, rosuvastatin) and hydrophilic (atorvastatin, simvastatin, lovastatin) statins, although the analysis wasn’t adjusted for doses used.

In a meta-regression analysis, baseline body mass index or percentage change in low-density lipoprotein cholesterol didn’t appear to confer additional risk. The risk of diabetes with statins was generally higher in studies with older patients (data given graphically).

Higher statin doses mean higher risk

A meta-analysis of 5 placebo and standard-of-care randomized controlled trials (39,612 patients, 83% without diabetes at enrollment, mean age 58-64 years) found that the risk of diabetes was higher with higher-dose statins.² Therapy with atorvastatin 80 mg or simvastatin 40 to 80 mg was defined as intensive. Treatment with simvastatin 20 to 40 mg, atorvastatin 10 mg, or pravastatin 40 mg was defined as moderate.

At a mean follow-up of 4.9 years, intensive statin therapy was associated with a higher risk of developing diabetes than moderate therapy (OR=1.12; 95% CI, 1.04-1.22) with 2 additional cases of diabetes per 1000 patient-years in the intensive therapy group. The authors noted significant heterogeneity between trials with regard to major cardiovascular events.

Similar results were found in a subsequent population-based cohort study of 471,250 nondiabetic patients older than 66 years who were newly prescribed a statin.³ The study authors used the incidence of new diabetes in patients taking pravastatin as the baseline, since it had been associated with reduced rates of diabetes in a large cardiovascular prevention trial.⁴ Without adjusting for dose, patients were at significantly higher risk of diabetes if prescribed atorvastatin (hazard ratio [HR]=1.22; 95% CI, 1.15-1.29), rosuvastatin (HR=1.18; 95% CI, 1.10-1.26), or simvastatin (HR=1.10; 95% CI, 1.04-1.17) compared with pravastatin. The risk with fluvastatin and lovastatin was similar to pravastatin.

A subanalysis that compared moderate- and high-dose statin therapy with low-dose therapy (atorvastatin <20 mg, rosuvastatin <10 mg, simvastatin <80 mg, or any dose of fluvastatin, lovastatin, or pravastatin) found a 22% increased risk of diabetes (HR=1.22; 95% CI, 1.19-1.26) for moderate-dose therapy (atorvastatin 20-79 mg, rosuvastatin 10-39 mg, or simvastatin >80 mg) and a 30% increased risk (HR=1.3; 95% CI, 1.2-1.4) for high-dose therapy (atorvastatin ≥80 mg or rosuvastatin ≥40 mg).

A cohort trial also shows increased diabetes risk

A smaller subsequent cohort trial based on data from Taiwan National Health Insurance records compared 8412 nondiabetic adult patients (mean age 63 years) taking statins with 33,648 age- and risk-matched controls not taking statins over a mean duration of 7.2 years.⁵ Statin use was associated with a 15% higher risk of developing diabetes (HR=1.15; 95% CI, 1.08-1.22).

RECOMMENDATIONS

The 2013 American College of Cardiology/American Heart Association guidelines for lipid-lowering therapy recommend that patients taking statins be screened for diabetes according to current screening recommendations.⁶ The guidelines advise encouraging patients who develop diabetes while on statin therapy to adhere to a heart-healthy dietary pattern, engage in physical activity, achieve and maintain a healthy body weight, cease tobacco use, and continue statin therapy to reduce the risk of cardiovascular events.

EVIDENCE SUMMARY

A meta-analysis of 13 randomized, placebo or standard of care-controlled statin trials (113,148 patients, 81% without diabetes at enrollment, mean ages 55-76 years) found that statin therapy increased the incidence of diabetes by 9% over 4 years (odds ratio [OR]=1.09; 95% confidence interval [CI], 1.02-1.17), or one additional case per 255 patients.¹ The increased risk was similar for lipophilic (pravastatin, rosuvastatin) and hydrophilic (atorvastatin, simvastatin, lovastatin) statins, although the analysis wasn’t adjusted for doses used.

In a meta-regression analysis, baseline body mass index or percentage change in low-density lipoprotein cholesterol didn’t appear to confer additional risk. The risk of diabetes with statins was generally higher in studies with older patients (data given graphically).

Higher statin doses mean higher risk

A meta-analysis of 5 placebo and standard-of-care randomized controlled trials (39,612 patients, 83% without diabetes at enrollment, mean age 58-64 years) found that the risk of diabetes was higher with higher-dose statins.² Therapy with atorvastatin 80 mg or simvastatin 40 to 80 mg was defined as intensive. Treatment with simvastatin 20 to 40 mg, atorvastatin 10 mg, or pravastatin 40 mg was defined as moderate.

At a mean follow-up of 4.9 years, intensive statin therapy was associated with a higher risk of developing diabetes than moderate therapy (OR=1.12; 95% CI, 1.04-1.22) with 2 additional cases of diabetes per 1000 patient-years in the intensive therapy group. The authors noted significant heterogeneity between trials with regard to major cardiovascular events.

Similar results were found in a subsequent population-based cohort study of 471,250 nondiabetic patients older than 66 years who were newly prescribed a statin.³ The study authors used the incidence of new diabetes in patients taking pravastatin as the baseline, since it had been associated with reduced rates of diabetes in a large cardiovascular prevention trial.⁴ Without adjusting for dose, patients were at significantly higher risk of diabetes if prescribed atorvastatin (hazard ratio [HR]=1.22; 95% CI, 1.15-1.29), rosuvastatin (HR=1.18; 95% CI, 1.10-1.26), or simvastatin (HR=1.10; 95% CI, 1.04-1.17) compared with pravastatin. The risk with fluvastatin and lovastatin was similar to pravastatin.

A subanalysis that compared moderate- and high-dose statin therapy with low-dose therapy (atorvastatin <20 mg, rosuvastatin <10 mg, simvastatin <80 mg, or any dose of fluvastatin, lovastatin, or pravastatin) found a 22% increased risk of diabetes (HR=1.22; 95% CI, 1.19-1.26) for moderate-dose therapy (atorvastatin 20-79 mg, rosuvastatin 10-39 mg, or simvastatin >80 mg) and a 30% increased risk (HR=1.3; 95% CI, 1.2-1.4) for high-dose therapy (atorvastatin ≥80 mg or rosuvastatin ≥40 mg).

A cohort trial also shows increased diabetes risk

A smaller subsequent cohort trial based on data from Taiwan National Health Insurance records compared 8412 nondiabetic adult patients (mean age 63 years) taking statins with 33,648 age- and risk-matched controls not taking statins over a mean duration of 7.2 years.⁵ Statin use was associated with a 15% higher risk of developing diabetes (HR=1.15; 95% CI, 1.08-1.22).

RECOMMENDATIONS

The 2013 American College of Cardiology/American Heart Association guidelines for lipid-lowering therapy recommend that patients taking statins be screened for diabetes according to current screening recommendations.⁶ The guidelines advise encouraging patients who develop diabetes while on statin therapy to adhere to a heart-healthy dietary pattern, engage in physical activity, achieve and maintain a healthy body weight, cease tobacco use, and continue statin therapy to reduce the risk of cardiovascular events.

EVIDENCE SUMMARY

A meta-analysis of 13 randomized, placebo or standard of care-controlled statin trials (113,148 patients, 81% without diabetes at enrollment, mean ages 55-76 years) found that statin therapy increased the incidence of diabetes by 9% over 4 years (odds ratio [OR]=1.09; 95% confidence interval [CI], 1.02-1.17), or one additional case per 255 patients.¹ The increased risk was similar for lipophilic (pravastatin, rosuvastatin) and hydrophilic (atorvastatin, simvastatin, lovastatin) statins, although the analysis wasn’t adjusted for doses used.

In a meta-regression analysis, baseline body mass index or percentage change in low-density lipoprotein cholesterol didn’t appear to confer additional risk. The risk of diabetes with statins was generally higher in studies with older patients (data given graphically).

Higher statin doses mean higher risk

A meta-analysis of 5 placebo and standard-of-care randomized controlled trials (39,612 patients, 83% without diabetes at enrollment, mean age 58-64 years) found that the risk of diabetes was higher with higher-dose statins.² Therapy with atorvastatin 80 mg or simvastatin 40 to 80 mg was defined as intensive. Treatment with simvastatin 20 to 40 mg, atorvastatin 10 mg, or pravastatin 40 mg was defined as moderate.

At a mean follow-up of 4.9 years, intensive statin therapy was associated with a higher risk of developing diabetes than moderate therapy (OR=1.12; 95% CI, 1.04-1.22) with 2 additional cases of diabetes per 1000 patient-years in the intensive therapy group. The authors noted significant heterogeneity between trials with regard to major cardiovascular events.

Similar results were found in a subsequent population-based cohort study of 471,250 nondiabetic patients older than 66 years who were newly prescribed a statin.³ The study authors used the incidence of new diabetes in patients taking pravastatin as the baseline, since it had been associated with reduced rates of diabetes in a large cardiovascular prevention trial.⁴ Without adjusting for dose, patients were at significantly higher risk of diabetes if prescribed atorvastatin (hazard ratio [HR]=1.22; 95% CI, 1.15-1.29), rosuvastatin (HR=1.18; 95% CI, 1.10-1.26), or simvastatin (HR=1.10; 95% CI, 1.04-1.17) compared with pravastatin. The risk with fluvastatin and lovastatin was similar to pravastatin.

A subanalysis that compared moderate- and high-dose statin therapy with low-dose therapy (atorvastatin <20 mg, rosuvastatin <10 mg, simvastatin <80 mg, or any dose of fluvastatin, lovastatin, or pravastatin) found a 22% increased risk of diabetes (HR=1.22; 95% CI, 1.19-1.26) for moderate-dose therapy (atorvastatin 20-79 mg, rosuvastatin 10-39 mg, or simvastatin >80 mg) and a 30% increased risk (HR=1.3; 95% CI, 1.2-1.4) for high-dose therapy (atorvastatin ≥80 mg or rosuvastatin ≥40 mg).

A cohort trial also shows increased diabetes risk

A smaller subsequent cohort trial based on data from Taiwan National Health Insurance records compared 8412 nondiabetic adult patients (mean age 63 years) taking statins with 33,648 age- and risk-matched controls not taking statins over a mean duration of 7.2 years.⁵ Statin use was associated with a 15% higher risk of developing diabetes (HR=1.15; 95% CI, 1.08-1.22).

RECOMMENDATIONS

The 2013 American College of Cardiology/American Heart Association guidelines for lipid-lowering therapy recommend that patients taking statins be screened for diabetes according to current screening recommendations.⁶ The guidelines advise encouraging patients who develop diabetes while on statin therapy to adhere to a heart-healthy dietary pattern, engage in physical activity, achieve and maintain a healthy body weight, cease tobacco use, and continue statin therapy to reduce the risk of cardiovascular events.

THE CASE

One month after moving into her mother’s apartment, a 27-year-old woman sought care at our clinic for fatigue, headache, blurred vision, nausea, and morning vomiting. She had weakness and difficulty sleeping, but denied any fever, rashes, neck stiffness, recent travel, trauma, or tobacco or illicit drug use. She did, however, have a 6-year history of migraines. Her physical exam was normal. She was sent home with a prescription for tramadol 50 mg bid for her headaches.

The patient subsequently went to the emergency department 3 times for the same complaints; none of the treatments she received there (mostly acetaminophen with codeine) relieved her symptoms. Three weeks later she returned to our clinic. She was distressed that the symptoms hadn’t gone away, and noted that her family was now experiencing similar symptoms.

Her temperature was 98.1°F (36.7°C), blood pressure was 131/88 mm Hg, pulse  was 85 beats/min, and respiratory rate was 18 breaths/min. Physical and neurologic exams were normal.

THE DIAGNOSIS

Although most of the patient’s lab test results were within normal ranges, her carboxyhemoglobin (COHb) level was 4.2%. COHb levels of >2% to 3% in nonsmokers or >9% to 10% in smokers suggest carbon monoxide (CO) poisoning.^1,2 Based on this finding and our patient’s symptoms, we diagnosed unintentional CO poisoning. We recommended that she and her mother vacate the apartment and have it inspected.

DISCUSSION

CO is the leading cause of poisoning mortality in the United States, and causes half of all fatal poisonings worldwide.^1,3,4 It is a colorless, odorless, and tasteless gas that is produced by the incomplete combustion of carbon-based products, such as coal or gas.^5,6 Exposure can occur from car exhaust fumes, faulty room heaters, and other sources (TABLE 1).⁶ The incidence of CO poisoning is higher during the winter months and after natural disasters. Individuals who have a lowered oxygen capacity, such as older adults, pregnant women (and their fetuses), infants, and patients with anemia, cardiovascular disease, or cerebrovascular disease, are more susceptible to CO poisoning.^5,6

COHb, a stable complex of CO that forms in red blood cells when CO is inhaled, impairs oxygen delivery and peripheral utilization, resulting in cellular hypoxia.¹ Signs and symptoms of CO poisoning are nonspecific and require a high degree of clinical suspicion for early diagnosis and treatment. Although cherry-red lips, peripheral cyanosis, and retinal hemorrhages are often described as “classic” symptoms of CO poisoning, these are rarely seen.⁶ The most common symptoms are actually headache (90%), dizziness (82%), and weakness (53%).⁷ Other symptoms include nausea, vomiting, confusion, visual disturbances, loss of consciousness, angina, seizure, and fatigue.^6,7 Symptoms of chronic CO poisoning may differ from those of acute poisoning and can include chronic fatigue, neuropathy, and memory deficit.⁸

The differential diagnosis for CO poisoning includes flu-like syndrome/influenza/other viral illnesses, migraine or tension headaches, depression, transient ischemic attack, encephalitis, coronary artery disease, gastroenteritis or food poisoning, seizures, and dysrhythmias.^1,4 Lab testing for COHb can help narrow the diagnosis. CO poisoning can be classified as mild, moderate, or severe based on COHb levels and the patient’s signs and symptoms (TABLE 2).⁶ However, COHb level is a poor predictor of clinical presentation and should not be used to dictate management.^2,7

Oxygen therapy is the recommended treatment

Early treatment with supplemental oxygen is recommended to reduce the length of time red blood cells are exposed to CO.¹ A COHb level >25% is the criterion for hyperbaric oxygen therapy.^1,3 Patients should receive treatment until their symptoms become less intense.

The 3 most common symptoms of CO poisoning are headache, dizziness, and weakness.

Delayed neuropsychiatric sequelae (DNS) can occur in up to one-third of patients with acute CO poisoning more than a month after apparent recovery.^1,6,9 DNS symptoms include cognitive changes, emotional lability, visual disturbances, disorientation, depression, dementia, psychotic behavior, parkinsonism, amnesia, and incontinence.^1,6,9 Approximately 50% to 75% of patients with DNS recover spontaneously within a year with symptomatic treatment.^1,6,9

Our patient

After recommending that our patient (and her mother) leave the apartment and have it inspected, we later learned that the fire department was unable to determine the source of the CO. A CO detector was installed and our patient was advised to keep the windows in the apartment open to allow for adequate oxygen flow. One month later she returned to our clinic and reported that her symptoms resolved; serum COHb was negative upon repeat lab tests.

THE TAKEAWAY

Patients who present with headaches, dizziness and/or fatigue should be evaluated for CO poisoning. The patient’s environmental history should be reviewed carefully, especially because CO poisoning is more common during the winter months. Oxygen therapy is the mainstay of treatment. Up to one-third of patients with acute poisoning may develop delayed neuropsychiatric sequelae, including cognitive changes, emotional lability, visual disturbances, disorientation, and depression, that may resolve within one year.

THE CASE

One month after moving into her mother’s apartment, a 27-year-old woman sought care at our clinic for fatigue, headache, blurred vision, nausea, and morning vomiting. She had weakness and difficulty sleeping, but denied any fever, rashes, neck stiffness, recent travel, trauma, or tobacco or illicit drug use. She did, however, have a 6-year history of migraines. Her physical exam was normal. She was sent home with a prescription for tramadol 50 mg bid for her headaches.

The patient subsequently went to the emergency department 3 times for the same complaints; none of the treatments she received there (mostly acetaminophen with codeine) relieved her symptoms. Three weeks later she returned to our clinic. She was distressed that the symptoms hadn’t gone away, and noted that her family was now experiencing similar symptoms.

Her temperature was 98.1°F (36.7°C), blood pressure was 131/88 mm Hg, pulse  was 85 beats/min, and respiratory rate was 18 breaths/min. Physical and neurologic exams were normal.

THE DIAGNOSIS

Although most of the patient’s lab test results were within normal ranges, her carboxyhemoglobin (COHb) level was 4.2%. COHb levels of >2% to 3% in nonsmokers or >9% to 10% in smokers suggest carbon monoxide (CO) poisoning.^1,2 Based on this finding and our patient’s symptoms, we diagnosed unintentional CO poisoning. We recommended that she and her mother vacate the apartment and have it inspected.

DISCUSSION

CO is the leading cause of poisoning mortality in the United States, and causes half of all fatal poisonings worldwide.^1,3,4 It is a colorless, odorless, and tasteless gas that is produced by the incomplete combustion of carbon-based products, such as coal or gas.^5,6 Exposure can occur from car exhaust fumes, faulty room heaters, and other sources (TABLE 1).⁶ The incidence of CO poisoning is higher during the winter months and after natural disasters. Individuals who have a lowered oxygen capacity, such as older adults, pregnant women (and their fetuses), infants, and patients with anemia, cardiovascular disease, or cerebrovascular disease, are more susceptible to CO poisoning.^5,6

COHb, a stable complex of CO that forms in red blood cells when CO is inhaled, impairs oxygen delivery and peripheral utilization, resulting in cellular hypoxia.¹ Signs and symptoms of CO poisoning are nonspecific and require a high degree of clinical suspicion for early diagnosis and treatment. Although cherry-red lips, peripheral cyanosis, and retinal hemorrhages are often described as “classic” symptoms of CO poisoning, these are rarely seen.⁶ The most common symptoms are actually headache (90%), dizziness (82%), and weakness (53%).⁷ Other symptoms include nausea, vomiting, confusion, visual disturbances, loss of consciousness, angina, seizure, and fatigue.^6,7 Symptoms of chronic CO poisoning may differ from those of acute poisoning and can include chronic fatigue, neuropathy, and memory deficit.⁸

The differential diagnosis for CO poisoning includes flu-like syndrome/influenza/other viral illnesses, migraine or tension headaches, depression, transient ischemic attack, encephalitis, coronary artery disease, gastroenteritis or food poisoning, seizures, and dysrhythmias.^1,4 Lab testing for COHb can help narrow the diagnosis. CO poisoning can be classified as mild, moderate, or severe based on COHb levels and the patient’s signs and symptoms (TABLE 2).⁶ However, COHb level is a poor predictor of clinical presentation and should not be used to dictate management.^2,7

Oxygen therapy is the recommended treatment

Early treatment with supplemental oxygen is recommended to reduce the length of time red blood cells are exposed to CO.¹ A COHb level >25% is the criterion for hyperbaric oxygen therapy.^1,3 Patients should receive treatment until their symptoms become less intense.

The 3 most common symptoms of CO poisoning are headache, dizziness, and weakness.

Delayed neuropsychiatric sequelae (DNS) can occur in up to one-third of patients with acute CO poisoning more than a month after apparent recovery.^1,6,9 DNS symptoms include cognitive changes, emotional lability, visual disturbances, disorientation, depression, dementia, psychotic behavior, parkinsonism, amnesia, and incontinence.^1,6,9 Approximately 50% to 75% of patients with DNS recover spontaneously within a year with symptomatic treatment.^1,6,9

Our patient

After recommending that our patient (and her mother) leave the apartment and have it inspected, we later learned that the fire department was unable to determine the source of the CO. A CO detector was installed and our patient was advised to keep the windows in the apartment open to allow for adequate oxygen flow. One month later she returned to our clinic and reported that her symptoms resolved; serum COHb was negative upon repeat lab tests.

THE TAKEAWAY

Patients who present with headaches, dizziness and/or fatigue should be evaluated for CO poisoning. The patient’s environmental history should be reviewed carefully, especially because CO poisoning is more common during the winter months. Oxygen therapy is the mainstay of treatment. Up to one-third of patients with acute poisoning may develop delayed neuropsychiatric sequelae, including cognitive changes, emotional lability, visual disturbances, disorientation, and depression, that may resolve within one year.

THE CASE

One month after moving into her mother’s apartment, a 27-year-old woman sought care at our clinic for fatigue, headache, blurred vision, nausea, and morning vomiting. She had weakness and difficulty sleeping, but denied any fever, rashes, neck stiffness, recent travel, trauma, or tobacco or illicit drug use. She did, however, have a 6-year history of migraines. Her physical exam was normal. She was sent home with a prescription for tramadol 50 mg bid for her headaches.

The patient subsequently went to the emergency department 3 times for the same complaints; none of the treatments she received there (mostly acetaminophen with codeine) relieved her symptoms. Three weeks later she returned to our clinic. She was distressed that the symptoms hadn’t gone away, and noted that her family was now experiencing similar symptoms.

Her temperature was 98.1°F (36.7°C), blood pressure was 131/88 mm Hg, pulse  was 85 beats/min, and respiratory rate was 18 breaths/min. Physical and neurologic exams were normal.

THE DIAGNOSIS

Although most of the patient’s lab test results were within normal ranges, her carboxyhemoglobin (COHb) level was 4.2%. COHb levels of >2% to 3% in nonsmokers or >9% to 10% in smokers suggest carbon monoxide (CO) poisoning.^1,2 Based on this finding and our patient’s symptoms, we diagnosed unintentional CO poisoning. We recommended that she and her mother vacate the apartment and have it inspected.

DISCUSSION

CO is the leading cause of poisoning mortality in the United States, and causes half of all fatal poisonings worldwide.^1,3,4 It is a colorless, odorless, and tasteless gas that is produced by the incomplete combustion of carbon-based products, such as coal or gas.^5,6 Exposure can occur from car exhaust fumes, faulty room heaters, and other sources (TABLE 1).⁶ The incidence of CO poisoning is higher during the winter months and after natural disasters. Individuals who have a lowered oxygen capacity, such as older adults, pregnant women (and their fetuses), infants, and patients with anemia, cardiovascular disease, or cerebrovascular disease, are more susceptible to CO poisoning.^5,6

COHb, a stable complex of CO that forms in red blood cells when CO is inhaled, impairs oxygen delivery and peripheral utilization, resulting in cellular hypoxia.¹ Signs and symptoms of CO poisoning are nonspecific and require a high degree of clinical suspicion for early diagnosis and treatment. Although cherry-red lips, peripheral cyanosis, and retinal hemorrhages are often described as “classic” symptoms of CO poisoning, these are rarely seen.⁶ The most common symptoms are actually headache (90%), dizziness (82%), and weakness (53%).⁷ Other symptoms include nausea, vomiting, confusion, visual disturbances, loss of consciousness, angina, seizure, and fatigue.^6,7 Symptoms of chronic CO poisoning may differ from those of acute poisoning and can include chronic fatigue, neuropathy, and memory deficit.⁸

The differential diagnosis for CO poisoning includes flu-like syndrome/influenza/other viral illnesses, migraine or tension headaches, depression, transient ischemic attack, encephalitis, coronary artery disease, gastroenteritis or food poisoning, seizures, and dysrhythmias.^1,4 Lab testing for COHb can help narrow the diagnosis. CO poisoning can be classified as mild, moderate, or severe based on COHb levels and the patient’s signs and symptoms (TABLE 2).⁶ However, COHb level is a poor predictor of clinical presentation and should not be used to dictate management.^2,7

Oxygen therapy is the recommended treatment

Early treatment with supplemental oxygen is recommended to reduce the length of time red blood cells are exposed to CO.¹ A COHb level >25% is the criterion for hyperbaric oxygen therapy.^1,3 Patients should receive treatment until their symptoms become less intense.

The 3 most common symptoms of CO poisoning are headache, dizziness, and weakness.

Delayed neuropsychiatric sequelae (DNS) can occur in up to one-third of patients with acute CO poisoning more than a month after apparent recovery.^1,6,9 DNS symptoms include cognitive changes, emotional lability, visual disturbances, disorientation, depression, dementia, psychotic behavior, parkinsonism, amnesia, and incontinence.^1,6,9 Approximately 50% to 75% of patients with DNS recover spontaneously within a year with symptomatic treatment.^1,6,9

Our patient

After recommending that our patient (and her mother) leave the apartment and have it inspected, we later learned that the fire department was unable to determine the source of the CO. A CO detector was installed and our patient was advised to keep the windows in the apartment open to allow for adequate oxygen flow. One month later she returned to our clinic and reported that her symptoms resolved; serum COHb was negative upon repeat lab tests.

THE TAKEAWAY

Patients who present with headaches, dizziness and/or fatigue should be evaluated for CO poisoning. The patient’s environmental history should be reviewed carefully, especially because CO poisoning is more common during the winter months. Oxygen therapy is the mainstay of treatment. Up to one-third of patients with acute poisoning may develop delayed neuropsychiatric sequelae, including cognitive changes, emotional lability, visual disturbances, disorientation, and depression, that may resolve within one year.