Applied Evidence

CASE › Dennis D, age 63, was recently diagnosed with diabetes. his glycated hemoglobin (HbA1c) is 7.8%, his blood pressure (BP) is mildly elevated (145/95 mm Hg), and his body mass index (BMI) is 28.5, but his low-density lipoprotein (LDL) cholesterol is 100 mg/dl, his high-density lipoprotein (HDL) cholesterol is 52 mg/dL, and he has no history of cardiovascular disease (CVD). After an unsuccessful attempt to treat him with lifestyle modification, it is time to initiate diabetes therapy.

Other than an alpha-blocker for benign prostatic hyperplasia and a prostaglandin for glaucoma, Mr. D takes no other medications. You prescribe metformin 500 mg twice daily and consider what else to add to keep his diabetes well controlled. Should you prescribe an antihypertensive? And, despite the patient’s normal lipid levels, should he begin taking a statin?

The challenge faced by primary care physicians? Using the latest recommendations in a way that addresses the entire clinical picture, including the patient's age and overall health status, priorities, and preferences.Type 2 diabetes has been extensively studied in rigorous randomized controlled trials (RCTs). While studies have provided ample evidence in support of optimal treatment, differing interpretations of the findings are reflected in consensus guidelines developed by expert panels that don’t always see eye to eye on what diabetes treatment targets should be and how best to prevent micro- and macrovascular complications.

What’s more, recommendations continue to be updated as new data emerge. In February 2014, the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 8) revised its target for patients with diabetes to <140/90 mm Hg (from <130/80 mm Hg).¹ This is likely to lead to revisions in other leading consensus guidelines, as well.

Thus, primary care physicians managing the care of patients with diabetes face the challenge of using the latest recommendations in a manner that addresses the entire clinical picture, considering each patient’s age and overall health status, priorities, and preferences. We developed this evidence-based review and guide- line summary with that in mind.

HbA1c target: How low should you go?


The Diabetes Control and Complications Trial (DCCT), published nearly 20 years ago, studied patients with type 1 diabetes, and found that intensive insulin therapy (HbA1c ≤6%) delayed the onset of retinopathy, nephropathy, and neuropathy.² However, there was an important adverse effect of such intensive therapy: Patients in this group suffered from severe hypoglycemic episodes 3 times more frequently than those in the usual care group. Nonetheless, the microvascular benefits of intensive control observed in those with type 1 diabetes were thought to be similar for patients with type 2 diabetes.

The United Kingdom Prospective Diabetes Study (UKPDS), published in 1999, was the first major study to investigate targets for glucose control in patients with type 2 diabetes.³ Participants treated intensively (mean HbA1c goal, 7%) had a 25% reduction in microvascular complications, including the need for retinal photocoagulation, com- pared with those on standard control (mean HbA1c, 7.9%). There was also a nonsignificant trend toward a reduction in macrovascular complications in the intensive therapy group, but no difference in overall mortality rate.³

A 10-year follow-up of the UKPDS showed that while baseline differences in HbA1c between the 2 groups were lost by one year, reductions in microvascular complications continued to occur in the intensive treatment group.⁴ Reductions in myocardial infarction (MI) and death emerged over time, a possible legacy effect (ie, the result of intense treatment early in the course of the disease).

The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, published in 2008, studied patients at risk for CVD, defined by either a prior history of CVD or ≥2 other cardiovascular risk factors.⁵ Participants, all of whom had poorly controlled type 2 diabetes (mean HbA1c, 8.1%), were randomized to either intensive treatment (HbA1c goal, <6%) or standard therapy (HbA1c goal, 7%-7.9%). The study was discontinued after a mean follow-up of 3.5 years, when those in the intensive therapy group were found to have a higher mortality rate.⁵

The rate of nonfatal MI reported by the ACCORD trial was lower in the intensive therapy group, however, and participants in this group also had delayed onset of microalbuminuria.⁶ No differences were seen in serum creatinine concentrations, advanced nephropathy, diabetic eye complications, or nonfatal stroke. Five-year follow up confirmed an increased mortality rate in the intensive therapy group,⁷ the result of severe hypoglycemia.⁸

A less stringent HbA1c target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects.The Veterans Affairs Diabetes Trial (VADT) randomized patients with poorly controlled type 2 diabetes to intensive or standard therapy.⁹ At 6 months, the intensive therapy group’s HbA1c averaged 6.9%, compared with 8.4% for the standard therapy group. Except for a delay in the progression of albuminuria, no significant effects of intensive therapy were found: Rates of other microvascular complications, major cardiovascular events, and death were similar.⁹ It should be noted that the VADT involved fewer participants and shorter follow-up than the other trials cited (TABLE 1),^3-10 which may have affected its findings.

The Action in Diabetes and Vascular Disease (ADVANCE) trial, which included participants with either a history of major CVD or ≥1 other CVD risk factors, compared an intensive control group (mean HbA1c, 6.5%) with a standard care group (mean HbA1c, 7.3%)—with mixed results.¹⁰ Microalbuminuria occurred less frequently in the intensive therapy group, but hypoglycemia and hospitalization increased. No reduction in death from any cause, in cardiovascular death, or in major macrovascular events was found.

How to proceed? What the experts recommend

In updated standards for the medical care of diabetes released in January 2013,¹¹ the American Diabetes Association (ADA) calls for an HbA1c goal <7% for most nonpregnant adults with type 2 diabetes. This is in line with the 2012 International Diabetes Federation (IDF) guideline.¹²

The 2011 guideline from the American Association of Clinical Endocrinologists (AACE),¹³ however, recommends tighter control—an HbA1c of ≤6.5% for most patients. For patients with diabetes of short duration, a long life expectancy, and no significant history of CVD, the AACE believes that this more aggressive goal has the potential to further reduce the risk of microvascular complications.

A less stringent target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects. That would apply to those with a history of severe hypoglycemia, a limited life expectancy, advanced micro- or macrovascular complications, or extensive comorbid conditions, as well as to any patient for whom stricter control is difficult to attain even with intensive therapy.¹³

Setting a BP target

In 2003, the 7th report of the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 7) recommended a target BP <130/80 mm Hg for diabetes patients.¹⁴ Most major diabetes guidelines, including those of the AACE¹³ and IDF,¹² echoed this recommendation. As noted earlier, JNC 8, published earlier this year, loosened the recommendation to <140/90 mm Hg.¹ Although evidence has shown that treatment to a systolic BP <150 mm Hg improves cardiovascular and cerebrovascular outcomes for patients with diabetes,¹⁵ no RCTs have addressed whether more intensive treatment to achieve a systolic BP <140 mm Hg provides further benefit.

The BP of participants in the UKPDS has been examined, with patients with tighter control (<150/85 mm Hg) compared with those with less stringent control (<180/105 mm Hg). The tight control group showed a significant reduction in both death and complications related to diabetes, progression of diabetic retinopathy, and deterioration in visual acuity.¹⁵ Further investigation found that each 10 mm Hg reduction in systolic pressure was associated with a risk reduction of 15% for death related to diabetes, 12% for diabetes-related complications, 11% for MI, and 13% for microvascular complications.¹⁶

The ACCORD trial randomized participants to more intensive control (systolic BP <120 mm Hg, with a mean of 119.3) or standard therapy (systolic BP <140 mm Hg, mean 133.).¹⁷ After 4.7 years, no difference was found in the rates of MI, stroke, or death. However, a significant increase in the rate of serious adverse effects from antihypertensive treatment (including hypotension, syncope, bradycardia, hypokalemia, angioedema, and renal failure) occurred in the intensive control group.¹⁷

A subgroup analysis of patients with type 2 diabetes enrolled in the International Verapamil SR-Trandolapril Study (INVEST) evaluated systolic BP control and cardiovascular outcomes in those with preexisting coronary artery disease.¹⁸ Participants were categorized as having tight control if their systolic BP <130 mm Hg; usual control, if systolic pressure was between 130 and <140 mm Hg; and uncontrolled, if systolic BP ≥140 mm Hg. Those in the usual control group had lower risks of death, nonfatal MI, and stroke compared with those in the uncontrolled group, but little difference was found between patients in the usual control and tight control groups. The studies are summarized in TABLE 2.^15-18

Interpreting the results: The experts disagree


The ADA recommends that patients with diabetes and hypertension be treated to a goal <140 mm Hg systolic and <80 mm Hg diastolic pressure¹¹—more lenient than the recommendations of either the AACE or the IDF. It is not clear whether these recommendations will change, however, given the recent JNC 8 report.¹ A lower systolic target may be appropriate for certain patients, if it can be achieved without undue adverse effects from antihypertensive medication. Older patients in particular may be at risk for orthostasis or falls as a result of more aggressive treatment.

CASE › Mr. D’s most recent BP is 145/95. Given that his goal is <140/90, you elect to start lisinopril 10 mg daily, advise him to monitor his BP at home, and refer him to a dietician to discuss the Dietary Approaches to Stop Hypertension diet.

Lipid levels: When to add statin therapy


Like glucose and BP control, lipid control and, concomitantly, the benefit of statin therapy for patients with type 2 diabetes has been studied extensively (TABLE 3).^19-24

The updated standards released by the ADA in January 2013 recommend statin therapy, regardless of LDL level, for patients who have diabetes and known CVD.The Scandinavian Simvastatin Survival Study (4S) recruited participants with a history of MI or angina, and included a small diabetes subgroup.¹⁹ Participants were randomized to simvastatin 20 mg daily, with blinded titration up to 40 mg/d, or placebo. Among those with diabetes, patients on simvastatin had a 55% reduction in risk for major coronary heart disease events and a 43% reduction in total mortality. The risk reduction did not depend on baseline levels of total cholesterol, LDL cholesterol, HDL cholesterol, or triglycerides.

Cholesterol and Recurrent Events (CARE), which studied participants with a history of MI 3 to 20 months prior to the start of the study and also included a diabetes subgroup, had a similar outcome.²⁰ Compared with placebo, treatment with pravastatin 40 mg/d reduced the risk of both coronary events and revascularization procedures by 25%.

The Heart Protection Study randomized patients with either diabetes or a history of occlusive arterial disease to receive simvastatin 40 mg daily or placebo.²¹ In the treatment group, the risk of major vascular events was reduced in patients with diabetes by 27%. Improvements were seen in patients with LDL cholesterol levels both above and below 116 mg/dL.

Multiple studies have evaluated the benefits of atorvastatin for patients with diabetes. All have demonstrated a significant reduction in the risk of MI and death in those on statin therapy. The Treating to New Targets study showed a 25% reduction in major cardiovascular events in those treated with 80 mg atorvastatin daily (mean LDL, 77 mg/dL) vs those treated with 10 mg of the drug (mean LDL, 86 mg/dL).²² The Anglo-Scandinavian Cardiac Outcomes Trial–Lipid-Lowering Arm (ASCOT-LLA)²³ and the Collaborative Atorvastatin Diabetes Study (CARDS)²⁴ were both terminated early due to the magnitude of benefit seen with statin therapy. In contrast to LDL, evidence for non-LDL treatment goals is lacking in the diabetes literature. Also, there is little evidence to support nonstatin cholesterol-lowering therapy for the management of diabetes patients.

Statin use is widely recommended

In 2008, the ADA and the American College of Cardiology Foundation (ACCF) produced a joint consensus statement regarding lipoprotein management for patients with diabetes and multiple CVD risk factors.²⁵ Target LDL was recommended at <100 mg/dL for moderately high-risk primary prevention patients, including those with diabetes. For patients with diabetes and ≥1 other risk factors, the ADA/ACCF recommended an LDL goal <70 mg/dL. The 2011 AACE guideline has the same treatment goals,¹³ while the 2012 IDF guidelines are more aggressive.¹² For primary prevention, the AACE endorses an LDL goal <80 mg/dL, and <70 mg/dL for those with known CVD.¹³

The updated standards released by the ADA in January 2013 recommend statin therapy regardless of LDL level for patients who have diabetes and known CVD, as well as for those ages 40 years and older who do not have CVD but have ≥1 other risk factors. Specific risk factors include hypertension, dyslipidemia, albuminuria, and a family history of CVD.¹¹

The latest statin guideline. In November 2013, the American College of Cardiology and American Heart Association (ACC/AHA) published a new guideline for the treatment of cholesterol to reduce cardiovascular risk,²⁶ but said nothing for or against specific LDL or non-HDL cholesterol targets. The ACC/AHA recommends that all patients who have diabetes and are between the ages of 40 and 75 years be treated with a moderate dose of a statin—a target supported with strong (strength of recommendation: A) evidence.

Patients with diabetes and an estimated 10-year risk of CVD >7.5% should be considered for high-intensity statin therapy, according to the ACC/AHA.²⁶ For patients younger than 40 or older than 75, the decision to initiate statin therapy should be made by weighing the potential cardiovascular benefits, the risk of adverse effects, and the potential for drug-drug-interactions, as well as patient preference.

CASE › You discuss the need for moderate-dose statin therapy with Mr. D. He is hesitant at first, referring to a coworker who had “leg cramps” when he was taking a statin. You emphasize the importance of prevention in the care of his diabetes and convince the patient to begin a trial of atorvastatin 40 mg daily.

The ACC and AHA published a new guideline for the treatment of cholesterol to reduce cardiovascular risk, but said nothing for or against specific LDL or non-HDL cholesterol targets.You warn Mr. D of the possibility of an allergic reaction, rash, or cough from lisinopril and loose stools from metformin, and advise him to call if he develops muscle cramps that could be associated with the statin. Finally, you stress the importance of lifestyle modification, including diet and weight loss, and schedule a follow-up visit in 3 months.

At Mr. D’s next visit, you will check his HbA1c and BP. If his HbA1c is still >7.0%, you may increase the dose of metformin or add a sulfonylurea. The dose of lisinopril could be increased if the patient’s BP continues to be elevated. There will be no need to recheck Mr. D’s cholesterol levels, however, because the purpose of the statin therapy is to improve overall outcomes, rather than to achieve a target goal.

CORRESPONDENCE
Kathryn M. Harmes, MD, Department of Family Medicine, University of Michigan Medical School, 1150 West Medical Center Drive, M7300 Med Sci I, SPC 5625, Ann Arbor, MI 48109-5625; [email protected]

CASE › Dennis D, age 63, was recently diagnosed with diabetes. his glycated hemoglobin (HbA1c) is 7.8%, his blood pressure (BP) is mildly elevated (145/95 mm Hg), and his body mass index (BMI) is 28.5, but his low-density lipoprotein (LDL) cholesterol is 100 mg/dl, his high-density lipoprotein (HDL) cholesterol is 52 mg/dL, and he has no history of cardiovascular disease (CVD). After an unsuccessful attempt to treat him with lifestyle modification, it is time to initiate diabetes therapy.

Other than an alpha-blocker for benign prostatic hyperplasia and a prostaglandin for glaucoma, Mr. D takes no other medications. You prescribe metformin 500 mg twice daily and consider what else to add to keep his diabetes well controlled. Should you prescribe an antihypertensive? And, despite the patient’s normal lipid levels, should he begin taking a statin?

The challenge faced by primary care physicians? Using the latest recommendations in a way that addresses the entire clinical picture, including the patient's age and overall health status, priorities, and preferences.Type 2 diabetes has been extensively studied in rigorous randomized controlled trials (RCTs). While studies have provided ample evidence in support of optimal treatment, differing interpretations of the findings are reflected in consensus guidelines developed by expert panels that don’t always see eye to eye on what diabetes treatment targets should be and how best to prevent micro- and macrovascular complications.

What’s more, recommendations continue to be updated as new data emerge. In February 2014, the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 8) revised its target for patients with diabetes to <140/90 mm Hg (from <130/80 mm Hg).¹ This is likely to lead to revisions in other leading consensus guidelines, as well.

Thus, primary care physicians managing the care of patients with diabetes face the challenge of using the latest recommendations in a manner that addresses the entire clinical picture, considering each patient’s age and overall health status, priorities, and preferences. We developed this evidence-based review and guide- line summary with that in mind.

HbA1c target: How low should you go?


The Diabetes Control and Complications Trial (DCCT), published nearly 20 years ago, studied patients with type 1 diabetes, and found that intensive insulin therapy (HbA1c ≤6%) delayed the onset of retinopathy, nephropathy, and neuropathy.² However, there was an important adverse effect of such intensive therapy: Patients in this group suffered from severe hypoglycemic episodes 3 times more frequently than those in the usual care group. Nonetheless, the microvascular benefits of intensive control observed in those with type 1 diabetes were thought to be similar for patients with type 2 diabetes.

The United Kingdom Prospective Diabetes Study (UKPDS), published in 1999, was the first major study to investigate targets for glucose control in patients with type 2 diabetes.³ Participants treated intensively (mean HbA1c goal, 7%) had a 25% reduction in microvascular complications, including the need for retinal photocoagulation, com- pared with those on standard control (mean HbA1c, 7.9%). There was also a nonsignificant trend toward a reduction in macrovascular complications in the intensive therapy group, but no difference in overall mortality rate.³

A 10-year follow-up of the UKPDS showed that while baseline differences in HbA1c between the 2 groups were lost by one year, reductions in microvascular complications continued to occur in the intensive treatment group.⁴ Reductions in myocardial infarction (MI) and death emerged over time, a possible legacy effect (ie, the result of intense treatment early in the course of the disease).

The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, published in 2008, studied patients at risk for CVD, defined by either a prior history of CVD or ≥2 other cardiovascular risk factors.⁵ Participants, all of whom had poorly controlled type 2 diabetes (mean HbA1c, 8.1%), were randomized to either intensive treatment (HbA1c goal, <6%) or standard therapy (HbA1c goal, 7%-7.9%). The study was discontinued after a mean follow-up of 3.5 years, when those in the intensive therapy group were found to have a higher mortality rate.⁵

The rate of nonfatal MI reported by the ACCORD trial was lower in the intensive therapy group, however, and participants in this group also had delayed onset of microalbuminuria.⁶ No differences were seen in serum creatinine concentrations, advanced nephropathy, diabetic eye complications, or nonfatal stroke. Five-year follow up confirmed an increased mortality rate in the intensive therapy group,⁷ the result of severe hypoglycemia.⁸

A less stringent HbA1c target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects.The Veterans Affairs Diabetes Trial (VADT) randomized patients with poorly controlled type 2 diabetes to intensive or standard therapy.⁹ At 6 months, the intensive therapy group’s HbA1c averaged 6.9%, compared with 8.4% for the standard therapy group. Except for a delay in the progression of albuminuria, no significant effects of intensive therapy were found: Rates of other microvascular complications, major cardiovascular events, and death were similar.⁹ It should be noted that the VADT involved fewer participants and shorter follow-up than the other trials cited (TABLE 1),^3-10 which may have affected its findings.

The Action in Diabetes and Vascular Disease (ADVANCE) trial, which included participants with either a history of major CVD or ≥1 other CVD risk factors, compared an intensive control group (mean HbA1c, 6.5%) with a standard care group (mean HbA1c, 7.3%)—with mixed results.¹⁰ Microalbuminuria occurred less frequently in the intensive therapy group, but hypoglycemia and hospitalization increased. No reduction in death from any cause, in cardiovascular death, or in major macrovascular events was found.

How to proceed? What the experts recommend

In updated standards for the medical care of diabetes released in January 2013,¹¹ the American Diabetes Association (ADA) calls for an HbA1c goal <7% for most nonpregnant adults with type 2 diabetes. This is in line with the 2012 International Diabetes Federation (IDF) guideline.¹²

The 2011 guideline from the American Association of Clinical Endocrinologists (AACE),¹³ however, recommends tighter control—an HbA1c of ≤6.5% for most patients. For patients with diabetes of short duration, a long life expectancy, and no significant history of CVD, the AACE believes that this more aggressive goal has the potential to further reduce the risk of microvascular complications.

A less stringent target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects. That would apply to those with a history of severe hypoglycemia, a limited life expectancy, advanced micro- or macrovascular complications, or extensive comorbid conditions, as well as to any patient for whom stricter control is difficult to attain even with intensive therapy.¹³

Setting a BP target

In 2003, the 7th report of the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 7) recommended a target BP <130/80 mm Hg for diabetes patients.¹⁴ Most major diabetes guidelines, including those of the AACE¹³ and IDF,¹² echoed this recommendation. As noted earlier, JNC 8, published earlier this year, loosened the recommendation to <140/90 mm Hg.¹ Although evidence has shown that treatment to a systolic BP <150 mm Hg improves cardiovascular and cerebrovascular outcomes for patients with diabetes,¹⁵ no RCTs have addressed whether more intensive treatment to achieve a systolic BP <140 mm Hg provides further benefit.

The BP of participants in the UKPDS has been examined, with patients with tighter control (<150/85 mm Hg) compared with those with less stringent control (<180/105 mm Hg). The tight control group showed a significant reduction in both death and complications related to diabetes, progression of diabetic retinopathy, and deterioration in visual acuity.¹⁵ Further investigation found that each 10 mm Hg reduction in systolic pressure was associated with a risk reduction of 15% for death related to diabetes, 12% for diabetes-related complications, 11% for MI, and 13% for microvascular complications.¹⁶

The ACCORD trial randomized participants to more intensive control (systolic BP <120 mm Hg, with a mean of 119.3) or standard therapy (systolic BP <140 mm Hg, mean 133.).¹⁷ After 4.7 years, no difference was found in the rates of MI, stroke, or death. However, a significant increase in the rate of serious adverse effects from antihypertensive treatment (including hypotension, syncope, bradycardia, hypokalemia, angioedema, and renal failure) occurred in the intensive control group.¹⁷

A subgroup analysis of patients with type 2 diabetes enrolled in the International Verapamil SR-Trandolapril Study (INVEST) evaluated systolic BP control and cardiovascular outcomes in those with preexisting coronary artery disease.¹⁸ Participants were categorized as having tight control if their systolic BP <130 mm Hg; usual control, if systolic pressure was between 130 and <140 mm Hg; and uncontrolled, if systolic BP ≥140 mm Hg. Those in the usual control group had lower risks of death, nonfatal MI, and stroke compared with those in the uncontrolled group, but little difference was found between patients in the usual control and tight control groups. The studies are summarized in TABLE 2.^15-18

Interpreting the results: The experts disagree


The ADA recommends that patients with diabetes and hypertension be treated to a goal <140 mm Hg systolic and <80 mm Hg diastolic pressure¹¹—more lenient than the recommendations of either the AACE or the IDF. It is not clear whether these recommendations will change, however, given the recent JNC 8 report.¹ A lower systolic target may be appropriate for certain patients, if it can be achieved without undue adverse effects from antihypertensive medication. Older patients in particular may be at risk for orthostasis or falls as a result of more aggressive treatment.

CASE › Mr. D’s most recent BP is 145/95. Given that his goal is <140/90, you elect to start lisinopril 10 mg daily, advise him to monitor his BP at home, and refer him to a dietician to discuss the Dietary Approaches to Stop Hypertension diet.

Lipid levels: When to add statin therapy


Like glucose and BP control, lipid control and, concomitantly, the benefit of statin therapy for patients with type 2 diabetes has been studied extensively (TABLE 3).^19-24

The updated standards released by the ADA in January 2013 recommend statin therapy, regardless of LDL level, for patients who have diabetes and known CVD.The Scandinavian Simvastatin Survival Study (4S) recruited participants with a history of MI or angina, and included a small diabetes subgroup.¹⁹ Participants were randomized to simvastatin 20 mg daily, with blinded titration up to 40 mg/d, or placebo. Among those with diabetes, patients on simvastatin had a 55% reduction in risk for major coronary heart disease events and a 43% reduction in total mortality. The risk reduction did not depend on baseline levels of total cholesterol, LDL cholesterol, HDL cholesterol, or triglycerides.

Cholesterol and Recurrent Events (CARE), which studied participants with a history of MI 3 to 20 months prior to the start of the study and also included a diabetes subgroup, had a similar outcome.²⁰ Compared with placebo, treatment with pravastatin 40 mg/d reduced the risk of both coronary events and revascularization procedures by 25%.

The Heart Protection Study randomized patients with either diabetes or a history of occlusive arterial disease to receive simvastatin 40 mg daily or placebo.²¹ In the treatment group, the risk of major vascular events was reduced in patients with diabetes by 27%. Improvements were seen in patients with LDL cholesterol levels both above and below 116 mg/dL.

Multiple studies have evaluated the benefits of atorvastatin for patients with diabetes. All have demonstrated a significant reduction in the risk of MI and death in those on statin therapy. The Treating to New Targets study showed a 25% reduction in major cardiovascular events in those treated with 80 mg atorvastatin daily (mean LDL, 77 mg/dL) vs those treated with 10 mg of the drug (mean LDL, 86 mg/dL).²² The Anglo-Scandinavian Cardiac Outcomes Trial–Lipid-Lowering Arm (ASCOT-LLA)²³ and the Collaborative Atorvastatin Diabetes Study (CARDS)²⁴ were both terminated early due to the magnitude of benefit seen with statin therapy. In contrast to LDL, evidence for non-LDL treatment goals is lacking in the diabetes literature. Also, there is little evidence to support nonstatin cholesterol-lowering therapy for the management of diabetes patients.

Statin use is widely recommended

In 2008, the ADA and the American College of Cardiology Foundation (ACCF) produced a joint consensus statement regarding lipoprotein management for patients with diabetes and multiple CVD risk factors.²⁵ Target LDL was recommended at <100 mg/dL for moderately high-risk primary prevention patients, including those with diabetes. For patients with diabetes and ≥1 other risk factors, the ADA/ACCF recommended an LDL goal <70 mg/dL. The 2011 AACE guideline has the same treatment goals,¹³ while the 2012 IDF guidelines are more aggressive.¹² For primary prevention, the AACE endorses an LDL goal <80 mg/dL, and <70 mg/dL for those with known CVD.¹³

The updated standards released by the ADA in January 2013 recommend statin therapy regardless of LDL level for patients who have diabetes and known CVD, as well as for those ages 40 years and older who do not have CVD but have ≥1 other risk factors. Specific risk factors include hypertension, dyslipidemia, albuminuria, and a family history of CVD.¹¹

The latest statin guideline. In November 2013, the American College of Cardiology and American Heart Association (ACC/AHA) published a new guideline for the treatment of cholesterol to reduce cardiovascular risk,²⁶ but said nothing for or against specific LDL or non-HDL cholesterol targets. The ACC/AHA recommends that all patients who have diabetes and are between the ages of 40 and 75 years be treated with a moderate dose of a statin—a target supported with strong (strength of recommendation: A) evidence.

Patients with diabetes and an estimated 10-year risk of CVD >7.5% should be considered for high-intensity statin therapy, according to the ACC/AHA.²⁶ For patients younger than 40 or older than 75, the decision to initiate statin therapy should be made by weighing the potential cardiovascular benefits, the risk of adverse effects, and the potential for drug-drug-interactions, as well as patient preference.

CASE › You discuss the need for moderate-dose statin therapy with Mr. D. He is hesitant at first, referring to a coworker who had “leg cramps” when he was taking a statin. You emphasize the importance of prevention in the care of his diabetes and convince the patient to begin a trial of atorvastatin 40 mg daily.

The ACC and AHA published a new guideline for the treatment of cholesterol to reduce cardiovascular risk, but said nothing for or against specific LDL or non-HDL cholesterol targets.You warn Mr. D of the possibility of an allergic reaction, rash, or cough from lisinopril and loose stools from metformin, and advise him to call if he develops muscle cramps that could be associated with the statin. Finally, you stress the importance of lifestyle modification, including diet and weight loss, and schedule a follow-up visit in 3 months.

At Mr. D’s next visit, you will check his HbA1c and BP. If his HbA1c is still >7.0%, you may increase the dose of metformin or add a sulfonylurea. The dose of lisinopril could be increased if the patient’s BP continues to be elevated. There will be no need to recheck Mr. D’s cholesterol levels, however, because the purpose of the statin therapy is to improve overall outcomes, rather than to achieve a target goal.

CORRESPONDENCE
Kathryn M. Harmes, MD, Department of Family Medicine, University of Michigan Medical School, 1150 West Medical Center Drive, M7300 Med Sci I, SPC 5625, Ann Arbor, MI 48109-5625; [email protected]

CASE › Dennis D, age 63, was recently diagnosed with diabetes. his glycated hemoglobin (HbA1c) is 7.8%, his blood pressure (BP) is mildly elevated (145/95 mm Hg), and his body mass index (BMI) is 28.5, but his low-density lipoprotein (LDL) cholesterol is 100 mg/dl, his high-density lipoprotein (HDL) cholesterol is 52 mg/dL, and he has no history of cardiovascular disease (CVD). After an unsuccessful attempt to treat him with lifestyle modification, it is time to initiate diabetes therapy.

Other than an alpha-blocker for benign prostatic hyperplasia and a prostaglandin for glaucoma, Mr. D takes no other medications. You prescribe metformin 500 mg twice daily and consider what else to add to keep his diabetes well controlled. Should you prescribe an antihypertensive? And, despite the patient’s normal lipid levels, should he begin taking a statin?

The challenge faced by primary care physicians? Using the latest recommendations in a way that addresses the entire clinical picture, including the patient's age and overall health status, priorities, and preferences.Type 2 diabetes has been extensively studied in rigorous randomized controlled trials (RCTs). While studies have provided ample evidence in support of optimal treatment, differing interpretations of the findings are reflected in consensus guidelines developed by expert panels that don’t always see eye to eye on what diabetes treatment targets should be and how best to prevent micro- and macrovascular complications.

What’s more, recommendations continue to be updated as new data emerge. In February 2014, the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 8) revised its target for patients with diabetes to <140/90 mm Hg (from <130/80 mm Hg).¹ This is likely to lead to revisions in other leading consensus guidelines, as well.

Thus, primary care physicians managing the care of patients with diabetes face the challenge of using the latest recommendations in a manner that addresses the entire clinical picture, considering each patient’s age and overall health status, priorities, and preferences. We developed this evidence-based review and guide- line summary with that in mind.

HbA1c target: How low should you go?


The Diabetes Control and Complications Trial (DCCT), published nearly 20 years ago, studied patients with type 1 diabetes, and found that intensive insulin therapy (HbA1c ≤6%) delayed the onset of retinopathy, nephropathy, and neuropathy.² However, there was an important adverse effect of such intensive therapy: Patients in this group suffered from severe hypoglycemic episodes 3 times more frequently than those in the usual care group. Nonetheless, the microvascular benefits of intensive control observed in those with type 1 diabetes were thought to be similar for patients with type 2 diabetes.

The United Kingdom Prospective Diabetes Study (UKPDS), published in 1999, was the first major study to investigate targets for glucose control in patients with type 2 diabetes.³ Participants treated intensively (mean HbA1c goal, 7%) had a 25% reduction in microvascular complications, including the need for retinal photocoagulation, com- pared with those on standard control (mean HbA1c, 7.9%). There was also a nonsignificant trend toward a reduction in macrovascular complications in the intensive therapy group, but no difference in overall mortality rate.³

A 10-year follow-up of the UKPDS showed that while baseline differences in HbA1c between the 2 groups were lost by one year, reductions in microvascular complications continued to occur in the intensive treatment group.⁴ Reductions in myocardial infarction (MI) and death emerged over time, a possible legacy effect (ie, the result of intense treatment early in the course of the disease).

The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, published in 2008, studied patients at risk for CVD, defined by either a prior history of CVD or ≥2 other cardiovascular risk factors.⁵ Participants, all of whom had poorly controlled type 2 diabetes (mean HbA1c, 8.1%), were randomized to either intensive treatment (HbA1c goal, <6%) or standard therapy (HbA1c goal, 7%-7.9%). The study was discontinued after a mean follow-up of 3.5 years, when those in the intensive therapy group were found to have a higher mortality rate.⁵

The rate of nonfatal MI reported by the ACCORD trial was lower in the intensive therapy group, however, and participants in this group also had delayed onset of microalbuminuria.⁶ No differences were seen in serum creatinine concentrations, advanced nephropathy, diabetic eye complications, or nonfatal stroke. Five-year follow up confirmed an increased mortality rate in the intensive therapy group,⁷ the result of severe hypoglycemia.⁸

A less stringent HbA1c target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects.The Veterans Affairs Diabetes Trial (VADT) randomized patients with poorly controlled type 2 diabetes to intensive or standard therapy.⁹ At 6 months, the intensive therapy group’s HbA1c averaged 6.9%, compared with 8.4% for the standard therapy group. Except for a delay in the progression of albuminuria, no significant effects of intensive therapy were found: Rates of other microvascular complications, major cardiovascular events, and death were similar.⁹ It should be noted that the VADT involved fewer participants and shorter follow-up than the other trials cited (TABLE 1),^3-10 which may have affected its findings.

The Action in Diabetes and Vascular Disease (ADVANCE) trial, which included participants with either a history of major CVD or ≥1 other CVD risk factors, compared an intensive control group (mean HbA1c, 6.5%) with a standard care group (mean HbA1c, 7.3%)—with mixed results.¹⁰ Microalbuminuria occurred less frequently in the intensive therapy group, but hypoglycemia and hospitalization increased. No reduction in death from any cause, in cardiovascular death, or in major macrovascular events was found.

How to proceed? What the experts recommend

In updated standards for the medical care of diabetes released in January 2013,¹¹ the American Diabetes Association (ADA) calls for an HbA1c goal <7% for most nonpregnant adults with type 2 diabetes. This is in line with the 2012 International Diabetes Federation (IDF) guideline.¹²

The 2011 guideline from the American Association of Clinical Endocrinologists (AACE),¹³ however, recommends tighter control—an HbA1c of ≤6.5% for most patients. For patients with diabetes of short duration, a long life expectancy, and no significant history of CVD, the AACE believes that this more aggressive goal has the potential to further reduce the risk of microvascular complications.

A less stringent target (eg, <8%) may be more appropriate for patients with a higher risk of adverse effects. That would apply to those with a history of severe hypoglycemia, a limited life expectancy, advanced micro- or macrovascular complications, or extensive comorbid conditions, as well as to any patient for whom stricter control is difficult to attain even with intensive therapy.¹³

Setting a BP target

In 2003, the 7th report of the Joint Committee on Prevention, Evaluation, and Treatment of High Blood Pressure (JCN 7) recommended a target BP <130/80 mm Hg for diabetes patients.¹⁴ Most major diabetes guidelines, including those of the AACE¹³ and IDF,¹² echoed this recommendation. As noted earlier, JNC 8, published earlier this year, loosened the recommendation to <140/90 mm Hg.¹ Although evidence has shown that treatment to a systolic BP <150 mm Hg improves cardiovascular and cerebrovascular outcomes for patients with diabetes,¹⁵ no RCTs have addressed whether more intensive treatment to achieve a systolic BP <140 mm Hg provides further benefit.

The BP of participants in the UKPDS has been examined, with patients with tighter control (<150/85 mm Hg) compared with those with less stringent control (<180/105 mm Hg). The tight control group showed a significant reduction in both death and complications related to diabetes, progression of diabetic retinopathy, and deterioration in visual acuity.¹⁵ Further investigation found that each 10 mm Hg reduction in systolic pressure was associated with a risk reduction of 15% for death related to diabetes, 12% for diabetes-related complications, 11% for MI, and 13% for microvascular complications.¹⁶

The ACCORD trial randomized participants to more intensive control (systolic BP <120 mm Hg, with a mean of 119.3) or standard therapy (systolic BP <140 mm Hg, mean 133.).¹⁷ After 4.7 years, no difference was found in the rates of MI, stroke, or death. However, a significant increase in the rate of serious adverse effects from antihypertensive treatment (including hypotension, syncope, bradycardia, hypokalemia, angioedema, and renal failure) occurred in the intensive control group.¹⁷

A subgroup analysis of patients with type 2 diabetes enrolled in the International Verapamil SR-Trandolapril Study (INVEST) evaluated systolic BP control and cardiovascular outcomes in those with preexisting coronary artery disease.¹⁸ Participants were categorized as having tight control if their systolic BP <130 mm Hg; usual control, if systolic pressure was between 130 and <140 mm Hg; and uncontrolled, if systolic BP ≥140 mm Hg. Those in the usual control group had lower risks of death, nonfatal MI, and stroke compared with those in the uncontrolled group, but little difference was found between patients in the usual control and tight control groups. The studies are summarized in TABLE 2.^15-18

Interpreting the results: The experts disagree


The ADA recommends that patients with diabetes and hypertension be treated to a goal <140 mm Hg systolic and <80 mm Hg diastolic pressure¹¹—more lenient than the recommendations of either the AACE or the IDF. It is not clear whether these recommendations will change, however, given the recent JNC 8 report.¹ A lower systolic target may be appropriate for certain patients, if it can be achieved without undue adverse effects from antihypertensive medication. Older patients in particular may be at risk for orthostasis or falls as a result of more aggressive treatment.

CASE › Mr. D’s most recent BP is 145/95. Given that his goal is <140/90, you elect to start lisinopril 10 mg daily, advise him to monitor his BP at home, and refer him to a dietician to discuss the Dietary Approaches to Stop Hypertension diet.

Lipid levels: When to add statin therapy


Like glucose and BP control, lipid control and, concomitantly, the benefit of statin therapy for patients with type 2 diabetes has been studied extensively (TABLE 3).^19-24

The updated standards released by the ADA in January 2013 recommend statin therapy, regardless of LDL level, for patients who have diabetes and known CVD.The Scandinavian Simvastatin Survival Study (4S) recruited participants with a history of MI or angina, and included a small diabetes subgroup.¹⁹ Participants were randomized to simvastatin 20 mg daily, with blinded titration up to 40 mg/d, or placebo. Among those with diabetes, patients on simvastatin had a 55% reduction in risk for major coronary heart disease events and a 43% reduction in total mortality. The risk reduction did not depend on baseline levels of total cholesterol, LDL cholesterol, HDL cholesterol, or triglycerides.

Cholesterol and Recurrent Events (CARE), which studied participants with a history of MI 3 to 20 months prior to the start of the study and also included a diabetes subgroup, had a similar outcome.²⁰ Compared with placebo, treatment with pravastatin 40 mg/d reduced the risk of both coronary events and revascularization procedures by 25%.

The Heart Protection Study randomized patients with either diabetes or a history of occlusive arterial disease to receive simvastatin 40 mg daily or placebo.²¹ In the treatment group, the risk of major vascular events was reduced in patients with diabetes by 27%. Improvements were seen in patients with LDL cholesterol levels both above and below 116 mg/dL.

Multiple studies have evaluated the benefits of atorvastatin for patients with diabetes. All have demonstrated a significant reduction in the risk of MI and death in those on statin therapy. The Treating to New Targets study showed a 25% reduction in major cardiovascular events in those treated with 80 mg atorvastatin daily (mean LDL, 77 mg/dL) vs those treated with 10 mg of the drug (mean LDL, 86 mg/dL).²² The Anglo-Scandinavian Cardiac Outcomes Trial–Lipid-Lowering Arm (ASCOT-LLA)²³ and the Collaborative Atorvastatin Diabetes Study (CARDS)²⁴ were both terminated early due to the magnitude of benefit seen with statin therapy. In contrast to LDL, evidence for non-LDL treatment goals is lacking in the diabetes literature. Also, there is little evidence to support nonstatin cholesterol-lowering therapy for the management of diabetes patients.

Statin use is widely recommended

In 2008, the ADA and the American College of Cardiology Foundation (ACCF) produced a joint consensus statement regarding lipoprotein management for patients with diabetes and multiple CVD risk factors.²⁵ Target LDL was recommended at <100 mg/dL for moderately high-risk primary prevention patients, including those with diabetes. For patients with diabetes and ≥1 other risk factors, the ADA/ACCF recommended an LDL goal <70 mg/dL. The 2011 AACE guideline has the same treatment goals,¹³ while the 2012 IDF guidelines are more aggressive.¹² For primary prevention, the AACE endorses an LDL goal <80 mg/dL, and <70 mg/dL for those with known CVD.¹³

The updated standards released by the ADA in January 2013 recommend statin therapy regardless of LDL level for patients who have diabetes and known CVD, as well as for those ages 40 years and older who do not have CVD but have ≥1 other risk factors. Specific risk factors include hypertension, dyslipidemia, albuminuria, and a family history of CVD.¹¹

The latest statin guideline. In November 2013, the American College of Cardiology and American Heart Association (ACC/AHA) published a new guideline for the treatment of cholesterol to reduce cardiovascular risk,²⁶ but said nothing for or against specific LDL or non-HDL cholesterol targets. The ACC/AHA recommends that all patients who have diabetes and are between the ages of 40 and 75 years be treated with a moderate dose of a statin—a target supported with strong (strength of recommendation: A) evidence.

Patients with diabetes and an estimated 10-year risk of CVD >7.5% should be considered for high-intensity statin therapy, according to the ACC/AHA.²⁶ For patients younger than 40 or older than 75, the decision to initiate statin therapy should be made by weighing the potential cardiovascular benefits, the risk of adverse effects, and the potential for drug-drug-interactions, as well as patient preference.

CASE › You discuss the need for moderate-dose statin therapy with Mr. D. He is hesitant at first, referring to a coworker who had “leg cramps” when he was taking a statin. You emphasize the importance of prevention in the care of his diabetes and convince the patient to begin a trial of atorvastatin 40 mg daily.

The ACC and AHA published a new guideline for the treatment of cholesterol to reduce cardiovascular risk, but said nothing for or against specific LDL or non-HDL cholesterol targets.You warn Mr. D of the possibility of an allergic reaction, rash, or cough from lisinopril and loose stools from metformin, and advise him to call if he develops muscle cramps that could be associated with the statin. Finally, you stress the importance of lifestyle modification, including diet and weight loss, and schedule a follow-up visit in 3 months.

At Mr. D’s next visit, you will check his HbA1c and BP. If his HbA1c is still >7.0%, you may increase the dose of metformin or add a sulfonylurea. The dose of lisinopril could be increased if the patient’s BP continues to be elevated. There will be no need to recheck Mr. D’s cholesterol levels, however, because the purpose of the statin therapy is to improve overall outcomes, rather than to achieve a target goal.

CORRESPONDENCE
Kathryn M. Harmes, MD, Department of Family Medicine, University of Michigan Medical School, 1150 West Medical Center Drive, M7300 Med Sci I, SPC 5625, Ann Arbor, MI 48109-5625; [email protected]

For the first 15 years of the epidemic, human immunodeficiency virus and acquired immunodeficiency syndrome (HIV/AIDs) was uniformly fatal. Between 1981 and 1996, approximately 362,000 people in the United States succumbed to the disease.¹ That began to change in the mid 1990s, though, when highly active antiretroviral therapy (HAART) came into routine use. From that point forward, HIV became a chronic, manageable disease for most patients; an estimated 1.2 million people in the United States are now living with HIV infection.²

Unfortunately, the number of new infections continues to grow. There are more than 50,000 new infections in the United States each year,² and an estimated approximately 200,000 people have it but are undiagnosed, leading to further spread of the disease.³ The Office of National AIDS Policy has issued a National HIV/AIDS Strategy that seeks to reduce new infections by 25% in 2015, in part by identifying people with the disease who do not know their HIV status.⁴

But screening still has not gotten the uptake by clinicians that health officials would like.

Lack of awareness by physicians? Or an unwillingness of patients?

In 2006, the Centers for Disease Control and Prevention (CDC) began recommending routine HIV screening for individuals between the ages of 13 and 64, with patients given the ability to opt out of such testing.⁵ That same year, the CDC also removed some prior barriers to testing, such as requiring written consent and pretest counseling. But as of 2009, fewer than 50% of US adults had ever been tested for HIV⁶—possibly the result of physicians being unaware of the guidelines, patients being unwilling to be tested, and/or reimbursement issues.

Conflicting recommendations may have played a role. When the CDC released its 2006 recommendations, the United States Preventive Services Task Force (USPSTF) felt there was insufficient evidence to support routine HIV screening and issued a grade C recommendation. At that time, the USPSTF recommended that only high-risk individuals and pregnant women be tested (A recommendation, meaning there was high certainty that the net benefit was substantial).

However, in April 2013, based on new evidence regarding the clinical and public health benefits of early identification of HIV infection and subsequent treatment, the USPSTF updated its recommendations. The USPSTF now encourages clinicians to screen all adolescents and adults age 15 to 65 years for HIV (A recommendation).⁷ Shortly thereafter, the American Academy of Family Physicians (AAFP) also endorsed routine HIV screening, although the AAFP calls for such screening to begin at age 18.⁸

Insurance now covers it… A USPSTF A recommendation carries significant health policy implications because the Affordable Care Act requires private and public health insurance plans to cover preventive services recommended by USPSTF.⁹

Integrating screening into your practice

Serologic tests have come a long way. The first HIV antibody test was an enzyme immunoassay (EIA) that was introduced in 1985 and used mainly to screen the blood supply. This first-generation EIA identified only immunoglobulin G (IgG) antibodies to HIV type 1 (HIV-1). More sensitive and specific second- and third-generation EIAs have since been developed to detect both IgG and IgM antibodies, as well as antibodies to HIV-2. The third-generation assays also can detect antibodies as soon as 3 weeks after infection.

Fourth-generation assays allow for reliable detection of HIV infection within about 2 weeks of infection.The fourth-generation EIAs were approved by the US Food and Drug Administration (FDA) in 2010¹⁰ and are the first step in the CDC’s current HIV diagnostic testing algorithm. These tests can detect HIV-1/HIV-2 IgG and IgM antibodies and also p24 antigen, which is present within 7 days of the appearance of HIV RNA.¹¹ The fourth-generation assay allows for reliable detection within about 2 weeks of infection (FIGURE 1).¹⁰

Rapid HIV tests are also an option.¹² These tests can detect IgG and IgM antibodies in samples of saliva, whole blood, serum, and plasma. Results of rapid tests usually are available in 20 to 30 minutes and allow physicians to give patients the results while they are still in the office. In 2013 the FDA approved a combination p24 antigen/antibody rapid HIV assay that according to the manufacturer can detect infection earlier than other currently available rapid tests.¹³

When rapid tests are most useful. Rapid tests can be particularly useful for testing women presenting in labor who have not been screened for HIV as part of prenatal care. They also can be used to determine the need for postexposure prophylaxis in the event of a needlestick injury. According to manufacturer’s data, the sensitivity of rapid tests ranges from 99.3% to 100% and specificity from 99.7% to 99.9%.¹² However, in real-world experience these numbers have been slightly lower.¹² By comparison, the sensitivity and specificity of the fourth-generation EIAs are 99.4% and 99.5%, respectively.¹⁴

The downside... A disadvantage of rapid HIV testing is that under current FDA-approval status and CDC guidance, tests performed on oral fluid must have serologic confirmation. In addition, patients tested during the “window period” of seroconversion (after infection occurs but before antibodies are detectable) will test negative with rapid HIV tests and must be reminded that repeat testing should be done within 4 to 6 weeks of their last potential exposure to the virus. In high-prevalence settings such as urban emergency departments (EDs), rapid HIV tests have detected a significant number of new infections.¹⁵ However, ED physicians and urgent care providers have been reluctant to perform HIV tests due to the lack of follow-up for most patients treated in these settings.

Over-the-counter (OTC) tests. Approved by the FDA in 2012, the OraQuick In-Home HIV Test is the only available OTC test for use at home. Patients can go to the company’s Web site at www.oraquick.com to learn more about HIV and testing, and the company offers 24-hour phone support. It’s not clear how many patients are taking advantage of this home testing option. The test costs approximately $40 and several studies suggest that this price may deter patients from using it.¹⁶ In addition, it is not clear how patients who test positive using an OTC test will access medical care or get appropriate medical follow-up.

New testing algorithm eliminates Western blot

Historically, a patient with a reactive (positive) EIA result would undergo the Western blot assay as a confirmatory test. Although the Western blot for HIV is highly specific (99.7%), it tests only for the IgG antibody. This could lead to a false negative test in a patient in whom IgG seroconversion has not yet occurred. Additionally, the time for HIV confirmation with the Western blot often is one week or longer.

When speaking with a patient who tests positive on a rapid test, consider using the phrase "preliminary positive" results.Recently, the CDC has made available for public comment a diagnostic algorithm that removes the Western blot as a recommended test (FIGURE 2).¹⁷ This algorithm replaces the Western blot with an assay to differentiate HIV-1 and HIV-2 antibodies. Patients for whom this test is negative should undergo additional testing for HIV RNA to determine if HIV-1 is present. Positive HIV RNA would indicate acute or more recent infection. Studies suggest that this new algorithm is better than the existing algorithm at detecting HIV infections, and many reference labs have already adapted it.^17,18

Choosing your words carefully when giving patients their results

Patients can be given the results of a rapid HIV test during their visit, but a positive result on a rapid test should be confirmed by serologic testing. When speaking with a patient who tests positive on a rapid test, consider using the phrase “preliminary positive” results. This allows the patient to more easily process the results, knowing that a confirmatory blood draw will be done. State laws vary regarding how patients can receive HIV test results. Most states allow negative serologic test results to be given over the telephone (or electronically). For positive tests, it is preferable to give these results at a face-to-face consultation so that you can ensure the patient will have access to medical care. For more on HIV testing and lab reporting laws by state, see http://www.cdc.gov/hiv/policies/law/states/index.html.

CORRESPONDENCE
Jeffrey T. Kirchner, DO, FAAFP, AAHIVS, Family and Community Medicine, Lancaster General Hospital, 555 N. Duke Street #3555, Lancaster, PA 17602; [email protected]

For the first 15 years of the epidemic, human immunodeficiency virus and acquired immunodeficiency syndrome (HIV/AIDs) was uniformly fatal. Between 1981 and 1996, approximately 362,000 people in the United States succumbed to the disease.¹ That began to change in the mid 1990s, though, when highly active antiretroviral therapy (HAART) came into routine use. From that point forward, HIV became a chronic, manageable disease for most patients; an estimated 1.2 million people in the United States are now living with HIV infection.²

Unfortunately, the number of new infections continues to grow. There are more than 50,000 new infections in the United States each year,² and an estimated approximately 200,000 people have it but are undiagnosed, leading to further spread of the disease.³ The Office of National AIDS Policy has issued a National HIV/AIDS Strategy that seeks to reduce new infections by 25% in 2015, in part by identifying people with the disease who do not know their HIV status.⁴

But screening still has not gotten the uptake by clinicians that health officials would like.

Lack of awareness by physicians? Or an unwillingness of patients?

In 2006, the Centers for Disease Control and Prevention (CDC) began recommending routine HIV screening for individuals between the ages of 13 and 64, with patients given the ability to opt out of such testing.⁵ That same year, the CDC also removed some prior barriers to testing, such as requiring written consent and pretest counseling. But as of 2009, fewer than 50% of US adults had ever been tested for HIV⁶—possibly the result of physicians being unaware of the guidelines, patients being unwilling to be tested, and/or reimbursement issues.

Conflicting recommendations may have played a role. When the CDC released its 2006 recommendations, the United States Preventive Services Task Force (USPSTF) felt there was insufficient evidence to support routine HIV screening and issued a grade C recommendation. At that time, the USPSTF recommended that only high-risk individuals and pregnant women be tested (A recommendation, meaning there was high certainty that the net benefit was substantial).

However, in April 2013, based on new evidence regarding the clinical and public health benefits of early identification of HIV infection and subsequent treatment, the USPSTF updated its recommendations. The USPSTF now encourages clinicians to screen all adolescents and adults age 15 to 65 years for HIV (A recommendation).⁷ Shortly thereafter, the American Academy of Family Physicians (AAFP) also endorsed routine HIV screening, although the AAFP calls for such screening to begin at age 18.⁸

Insurance now covers it… A USPSTF A recommendation carries significant health policy implications because the Affordable Care Act requires private and public health insurance plans to cover preventive services recommended by USPSTF.⁹

Integrating screening into your practice

Serologic tests have come a long way. The first HIV antibody test was an enzyme immunoassay (EIA) that was introduced in 1985 and used mainly to screen the blood supply. This first-generation EIA identified only immunoglobulin G (IgG) antibodies to HIV type 1 (HIV-1). More sensitive and specific second- and third-generation EIAs have since been developed to detect both IgG and IgM antibodies, as well as antibodies to HIV-2. The third-generation assays also can detect antibodies as soon as 3 weeks after infection.

Fourth-generation assays allow for reliable detection of HIV infection within about 2 weeks of infection.The fourth-generation EIAs were approved by the US Food and Drug Administration (FDA) in 2010¹⁰ and are the first step in the CDC’s current HIV diagnostic testing algorithm. These tests can detect HIV-1/HIV-2 IgG and IgM antibodies and also p24 antigen, which is present within 7 days of the appearance of HIV RNA.¹¹ The fourth-generation assay allows for reliable detection within about 2 weeks of infection (FIGURE 1).¹⁰

Rapid HIV tests are also an option.¹² These tests can detect IgG and IgM antibodies in samples of saliva, whole blood, serum, and plasma. Results of rapid tests usually are available in 20 to 30 minutes and allow physicians to give patients the results while they are still in the office. In 2013 the FDA approved a combination p24 antigen/antibody rapid HIV assay that according to the manufacturer can detect infection earlier than other currently available rapid tests.¹³

When rapid tests are most useful. Rapid tests can be particularly useful for testing women presenting in labor who have not been screened for HIV as part of prenatal care. They also can be used to determine the need for postexposure prophylaxis in the event of a needlestick injury. According to manufacturer’s data, the sensitivity of rapid tests ranges from 99.3% to 100% and specificity from 99.7% to 99.9%.¹² However, in real-world experience these numbers have been slightly lower.¹² By comparison, the sensitivity and specificity of the fourth-generation EIAs are 99.4% and 99.5%, respectively.¹⁴

The downside... A disadvantage of rapid HIV testing is that under current FDA-approval status and CDC guidance, tests performed on oral fluid must have serologic confirmation. In addition, patients tested during the “window period” of seroconversion (after infection occurs but before antibodies are detectable) will test negative with rapid HIV tests and must be reminded that repeat testing should be done within 4 to 6 weeks of their last potential exposure to the virus. In high-prevalence settings such as urban emergency departments (EDs), rapid HIV tests have detected a significant number of new infections.¹⁵ However, ED physicians and urgent care providers have been reluctant to perform HIV tests due to the lack of follow-up for most patients treated in these settings.

Over-the-counter (OTC) tests. Approved by the FDA in 2012, the OraQuick In-Home HIV Test is the only available OTC test for use at home. Patients can go to the company’s Web site at www.oraquick.com to learn more about HIV and testing, and the company offers 24-hour phone support. It’s not clear how many patients are taking advantage of this home testing option. The test costs approximately $40 and several studies suggest that this price may deter patients from using it.¹⁶ In addition, it is not clear how patients who test positive using an OTC test will access medical care or get appropriate medical follow-up.

New testing algorithm eliminates Western blot

Historically, a patient with a reactive (positive) EIA result would undergo the Western blot assay as a confirmatory test. Although the Western blot for HIV is highly specific (99.7%), it tests only for the IgG antibody. This could lead to a false negative test in a patient in whom IgG seroconversion has not yet occurred. Additionally, the time for HIV confirmation with the Western blot often is one week or longer.

When speaking with a patient who tests positive on a rapid test, consider using the phrase "preliminary positive" results.Recently, the CDC has made available for public comment a diagnostic algorithm that removes the Western blot as a recommended test (FIGURE 2).¹⁷ This algorithm replaces the Western blot with an assay to differentiate HIV-1 and HIV-2 antibodies. Patients for whom this test is negative should undergo additional testing for HIV RNA to determine if HIV-1 is present. Positive HIV RNA would indicate acute or more recent infection. Studies suggest that this new algorithm is better than the existing algorithm at detecting HIV infections, and many reference labs have already adapted it.^17,18

Choosing your words carefully when giving patients their results

Patients can be given the results of a rapid HIV test during their visit, but a positive result on a rapid test should be confirmed by serologic testing. When speaking with a patient who tests positive on a rapid test, consider using the phrase “preliminary positive” results. This allows the patient to more easily process the results, knowing that a confirmatory blood draw will be done. State laws vary regarding how patients can receive HIV test results. Most states allow negative serologic test results to be given over the telephone (or electronically). For positive tests, it is preferable to give these results at a face-to-face consultation so that you can ensure the patient will have access to medical care. For more on HIV testing and lab reporting laws by state, see http://www.cdc.gov/hiv/policies/law/states/index.html.

CORRESPONDENCE
Jeffrey T. Kirchner, DO, FAAFP, AAHIVS, Family and Community Medicine, Lancaster General Hospital, 555 N. Duke Street #3555, Lancaster, PA 17602; [email protected]

For the first 15 years of the epidemic, human immunodeficiency virus and acquired immunodeficiency syndrome (HIV/AIDs) was uniformly fatal. Between 1981 and 1996, approximately 362,000 people in the United States succumbed to the disease.¹ That began to change in the mid 1990s, though, when highly active antiretroviral therapy (HAART) came into routine use. From that point forward, HIV became a chronic, manageable disease for most patients; an estimated 1.2 million people in the United States are now living with HIV infection.²

Unfortunately, the number of new infections continues to grow. There are more than 50,000 new infections in the United States each year,² and an estimated approximately 200,000 people have it but are undiagnosed, leading to further spread of the disease.³ The Office of National AIDS Policy has issued a National HIV/AIDS Strategy that seeks to reduce new infections by 25% in 2015, in part by identifying people with the disease who do not know their HIV status.⁴

But screening still has not gotten the uptake by clinicians that health officials would like.

Lack of awareness by physicians? Or an unwillingness of patients?

In 2006, the Centers for Disease Control and Prevention (CDC) began recommending routine HIV screening for individuals between the ages of 13 and 64, with patients given the ability to opt out of such testing.⁵ That same year, the CDC also removed some prior barriers to testing, such as requiring written consent and pretest counseling. But as of 2009, fewer than 50% of US adults had ever been tested for HIV⁶—possibly the result of physicians being unaware of the guidelines, patients being unwilling to be tested, and/or reimbursement issues.

Conflicting recommendations may have played a role. When the CDC released its 2006 recommendations, the United States Preventive Services Task Force (USPSTF) felt there was insufficient evidence to support routine HIV screening and issued a grade C recommendation. At that time, the USPSTF recommended that only high-risk individuals and pregnant women be tested (A recommendation, meaning there was high certainty that the net benefit was substantial).

However, in April 2013, based on new evidence regarding the clinical and public health benefits of early identification of HIV infection and subsequent treatment, the USPSTF updated its recommendations. The USPSTF now encourages clinicians to screen all adolescents and adults age 15 to 65 years for HIV (A recommendation).⁷ Shortly thereafter, the American Academy of Family Physicians (AAFP) also endorsed routine HIV screening, although the AAFP calls for such screening to begin at age 18.⁸

Insurance now covers it… A USPSTF A recommendation carries significant health policy implications because the Affordable Care Act requires private and public health insurance plans to cover preventive services recommended by USPSTF.⁹

Integrating screening into your practice

Serologic tests have come a long way. The first HIV antibody test was an enzyme immunoassay (EIA) that was introduced in 1985 and used mainly to screen the blood supply. This first-generation EIA identified only immunoglobulin G (IgG) antibodies to HIV type 1 (HIV-1). More sensitive and specific second- and third-generation EIAs have since been developed to detect both IgG and IgM antibodies, as well as antibodies to HIV-2. The third-generation assays also can detect antibodies as soon as 3 weeks after infection.

Fourth-generation assays allow for reliable detection of HIV infection within about 2 weeks of infection.The fourth-generation EIAs were approved by the US Food and Drug Administration (FDA) in 2010¹⁰ and are the first step in the CDC’s current HIV diagnostic testing algorithm. These tests can detect HIV-1/HIV-2 IgG and IgM antibodies and also p24 antigen, which is present within 7 days of the appearance of HIV RNA.¹¹ The fourth-generation assay allows for reliable detection within about 2 weeks of infection (FIGURE 1).¹⁰

Rapid HIV tests are also an option.¹² These tests can detect IgG and IgM antibodies in samples of saliva, whole blood, serum, and plasma. Results of rapid tests usually are available in 20 to 30 minutes and allow physicians to give patients the results while they are still in the office. In 2013 the FDA approved a combination p24 antigen/antibody rapid HIV assay that according to the manufacturer can detect infection earlier than other currently available rapid tests.¹³

When rapid tests are most useful. Rapid tests can be particularly useful for testing women presenting in labor who have not been screened for HIV as part of prenatal care. They also can be used to determine the need for postexposure prophylaxis in the event of a needlestick injury. According to manufacturer’s data, the sensitivity of rapid tests ranges from 99.3% to 100% and specificity from 99.7% to 99.9%.¹² However, in real-world experience these numbers have been slightly lower.¹² By comparison, the sensitivity and specificity of the fourth-generation EIAs are 99.4% and 99.5%, respectively.¹⁴

The downside... A disadvantage of rapid HIV testing is that under current FDA-approval status and CDC guidance, tests performed on oral fluid must have serologic confirmation. In addition, patients tested during the “window period” of seroconversion (after infection occurs but before antibodies are detectable) will test negative with rapid HIV tests and must be reminded that repeat testing should be done within 4 to 6 weeks of their last potential exposure to the virus. In high-prevalence settings such as urban emergency departments (EDs), rapid HIV tests have detected a significant number of new infections.¹⁵ However, ED physicians and urgent care providers have been reluctant to perform HIV tests due to the lack of follow-up for most patients treated in these settings.

Over-the-counter (OTC) tests. Approved by the FDA in 2012, the OraQuick In-Home HIV Test is the only available OTC test for use at home. Patients can go to the company’s Web site at www.oraquick.com to learn more about HIV and testing, and the company offers 24-hour phone support. It’s not clear how many patients are taking advantage of this home testing option. The test costs approximately $40 and several studies suggest that this price may deter patients from using it.¹⁶ In addition, it is not clear how patients who test positive using an OTC test will access medical care or get appropriate medical follow-up.

New testing algorithm eliminates Western blot

Historically, a patient with a reactive (positive) EIA result would undergo the Western blot assay as a confirmatory test. Although the Western blot for HIV is highly specific (99.7%), it tests only for the IgG antibody. This could lead to a false negative test in a patient in whom IgG seroconversion has not yet occurred. Additionally, the time for HIV confirmation with the Western blot often is one week or longer.

When speaking with a patient who tests positive on a rapid test, consider using the phrase "preliminary positive" results.Recently, the CDC has made available for public comment a diagnostic algorithm that removes the Western blot as a recommended test (FIGURE 2).¹⁷ This algorithm replaces the Western blot with an assay to differentiate HIV-1 and HIV-2 antibodies. Patients for whom this test is negative should undergo additional testing for HIV RNA to determine if HIV-1 is present. Positive HIV RNA would indicate acute or more recent infection. Studies suggest that this new algorithm is better than the existing algorithm at detecting HIV infections, and many reference labs have already adapted it.^17,18

Choosing your words carefully when giving patients their results

Patients can be given the results of a rapid HIV test during their visit, but a positive result on a rapid test should be confirmed by serologic testing. When speaking with a patient who tests positive on a rapid test, consider using the phrase “preliminary positive” results. This allows the patient to more easily process the results, knowing that a confirmatory blood draw will be done. State laws vary regarding how patients can receive HIV test results. Most states allow negative serologic test results to be given over the telephone (or electronically). For positive tests, it is preferable to give these results at a face-to-face consultation so that you can ensure the patient will have access to medical care. For more on HIV testing and lab reporting laws by state, see http://www.cdc.gov/hiv/policies/law/states/index.html.

CORRESPONDENCE
Jeffrey T. Kirchner, DO, FAAFP, AAHIVS, Family and Community Medicine, Lancaster General Hospital, 555 N. Duke Street #3555, Lancaster, PA 17602; [email protected]

When evaluating a patient with a history of chest pain, palpitations, syncope, and/or new-onset seizures, an electrocardiogram (EKG) may be the key to identifying a potentially life-threatening condition. Here we present 4 cases in which EKG findings were the clue to underlying medical conditions that, if left untreated, could be fatal. Because each of these conditions may not have associated findings on a physical exam, early recognition of these EKG findings can be lifesaving. 

CASE 1 › A 15-year-old boy suddenly collapses while walking, and bystanders report seizure-like activity. The patient doesn’t remember the event. Vital signs and physical exam are normal, and his blood glucose level is 86 mg/dL (normal: 70-100 mg/dL). He doesn’t take any medications and denies illicit drug use or recent illness.

What EKG abnormality (FIGURE 1) likely explains the cause of the patient’s collapse?

The EKG abnormality and diagnosis. The patient’s EKG revealed a prolonged QT interval (FIGURE 1, BRACKETS). His QTc (QT interval corrected for heart rate) was .470 seconds, which is at the high end of the normal range for his age and gender.¹ The patient had no family history of syncope, sudden cardiac death (SCD), or seizure disorder. Evaluation uncovered a calcium level of 4.4 mEq/L (normal: 4.5-5.5 mEq/L) and a phosphate level of 7.8 mg/dL (normal: 2.4-4.1 mg/dL).

This patient had a low parathyroid hormone from primary hypoparathyroidism. His conduction abnormality was treated with both oral calcium and vitamin D supplements.

Etiology and epidemiology. A prolonged QT interval may be the result of a primary long QT syndrome (LQTS) or an acquired condition from electrolyte imbalance, medication effect, or toxin exposure.

In the United States, the incidence of a genetic mutation that causes LQTS is 1 in 2500 people.² Patients with LQTS usually remain asymptomatic unless the QT interval is further prolonged by a condition or medication. There are several hundred congenital LQTS subtypes based on specific ion channel defects; the most common is LQTS1, with an inherited defect in the KCNQ1 gene, which regulates the slow potassium ion channel.

Acquired LQTS is much more common than congenital LQTS.³ Many drugs have been linked to an increased risk of LQTS, including certain antiarrhythmics, antibiotics, and antipsychotics (TABLE 1).⁴ In addition, electrolyte disturbances such as hypokalemia, hypocalcemia, and hypomagnesemia can be etiologic factors.

Be aware that an acquired LQTS may mask an underlying congenital LQTS. Therefore, patients in whom the offending agent or condition is corrected should still have a follow-up EKG. Screening family members for LQTS is worthwhile, even in those without symptoms.

Clinical features. Patients with symptomatic LQTS may have dizziness, palpitations, and syncope. SCD also is possible. These signs and symptoms may be triggered by strong emotions (in LQTS2) or physical activity (in LQTS1). They likely are caused by torsades de pointes and ventricular fibrillation. A brief aura may precede these arrhythmias, and patients may experience urinary or fecal incontinence.⁵

A common error in measuring the QT interval occurs when physicians inadvertently include a U-wave in the measurement.The key to making a diagnosis of LQTS is correctly measuring the QT interval. The QT interval is measured from the beginning of the Q-wave to the end of the T-wave as measured from the intersection of a line tangent to the downslope of the T-wave and the isoelectric line. This can be difficult to determine in EKGs showing bundle branch block or an irregular rhythm, such as atrial fibrillation (AF).^6,7 A common error in measuring the QT interval occurs when clinicians inadvertently include a U-wave in the measurement.¹ Some EKG machines may provide QT interval and QTc measurements. Normal QT intervals are ≤.450 seconds for men and ≤.470 seconds for women.⁸

It is essential to confirm the QT interval by using the Bazett formula (QTc equals the QT in seconds divided by the square root of the RR interval in seconds) for all patients with a history that suggests a possible arrhythmia.

First rule out or treat any causes of acquired LQTS by taking a medication history and evaluating the patient's electrolytes.Our patient had hypocalcemia, which on an EKG can cause T-wave widening with a normal ST segment, rather than a normal T-wave with a long ST segment, as is typically seen in LQTS. This distinction may be difficult to discern and should not preclude the search for either an acquired prolonged QTc or an underlying LQTS.⁹

Treatment. First rule out or treat any causes of acquired LQTS by taking a careful medication history and evaluating the patient’s electrolytes. Once these have been addressed, a beta-blocker is first-line therapy for symptomatic patients.⁵

Unfortunately, up to 20% of individuals treated with beta-blockers may continue to have syncope.⁵ For these patients, options include a left cardiac sympathetic denervation (LCSD) or placement of an automatic implantable cardioverter-defibrillator (AICD). An LCSD involves removal of the left-sided stellate and/or thoracic ganglia. This procedure can be used instead of, or in addition to, beta-blockers. If the patient’s syncope persists, AICDs are an option. AICDs can be lifesaving, but patients run the risk of adverse effects that include inappropriate shocks and infection.¹⁰ As the result of these therapies, mortality associated with LQTS has dropped to approximately 1%.¹¹

CASE 2 › A 14-year-old boy has a syncopal episode while at rest. A similar event occurred 3 years earlier; at that time, an echocardiogram and EKG were normal. For 2 days, he’s had a cough and low-grade fever. His temperature is 102ºF and he has a productive cough. Based on this EKG (FIGURE 2), what is the likely diagnosis? What is the significance of his fever?

The EKG abnormality and diagnosis. This patient’s EKG showed a type 1 Brugada pattern (FIGURE 2, ARROWS), which strongly supported the diagnosis of Brugada syndrome (BS). BS is an inherited condition caused by a genetic defect in cardiac ion channel function that leads to characteristic EKG changes and a predisposition to ventricular fibrillation.¹² In this case, the fever likely unmasked these EKG findings.

The patient was transferred to a local hospital for treatment of community-acquired pneumonia, and ultimately received an AICD.

Etiology and epidemiology. BS was first described in 1992¹³ and is a major cause of SCD, responsible for up to 4% of all cases of SCD, and 20% of cases of patients without structural heart disease.14 BS is more common in men, and the mean age of diagnosis is 40 to 45.^15-18

Mutations in at least 17 cardiac ion channel genes have been linked to BS.¹⁹ The SCN5A gene—a cardiac sodium channel—is the most commonly implicated, but accounts for only 11% to 24% of all BS cases.¹⁵

Clinical features. Patients with BS may present with syncope, nocturnal agonal respirations, or ventricular arrhythmias.¹² EKG findings include partial or complete right bundle branch block (RBBB) and ST segment elevation in right precordial leads V1 to V3.¹² There are 2 Brugada EKG patterns, a type 1 cloved pattern as seen in our patient’s EKG and a type 2 saddleback pattern.20 EKG findings are dynamic over time and may alternate between normal, type 1, and type 2.²⁰ Factors that modulate EKG appearance include fever, intoxication, vagal tone, electrolyte imbalance, and sodium channel blockade.^12,20

Diagnosis requires a type 1 Brugada pattern on EKG plus a family history of BS, documented ventricular arrhythmia, or arrhythmia-related symptoms such as syncope.¹² Patients with a type 2 Brugada pattern may undergo electrophysiology (EP) testing with Class I antiarrhythmic drugs to induce a diagnostic type 1 Brugada pattern.^12,21 Patients who have Brugada EKG findings but none of the other diagnostic criteria are considered to have a Brugada pattern (rather than Brugada syndrome).¹²

The most concerning outcome of BS is ventricular fibrillation. The estimated annual rate of cardiac events is 7.7% among patients who have experienced an aborted SCD, 1.9% among those who have experienced syncope, and 0.5% in asymptomatic patients.¹⁸

Treatment. The only effective treatment for BS is placement of an AICD; however, complications of AICD placement cause significant morbidity.⁶ Ten years after AICD placement, 37% of patients experienced inappropriate shocks and 29% experienced lead failure.²² Recent modifications in device programming and the addition of remote monitoring have decreased complication rates.^12,22

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal.The decision to place an AICD is based on the patient’s prior symptoms, EKG findings, and other factors. Recent guidelines recommend an AICD for all patients with a type 1 Brugada pattern (spontaneous or induced) who also have had an aborted SCD, syncope, or documented ventricular arrhythmia.¹²

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal. Asymptomatic patients with type 1 Brugada findings should undergo further EP testing, and should receive AICD only upon demonstration of inducible ventricular arrhythmia.¹²

TABLE 2
Arrhythmias associated with Wolff-Parkinson-White syndrome²³

^{EKG, electrocardiogram; N/A, not applicable; PSVT, paroxysmal supraventricular tachycardia.}

CASE 3 › A 21-year-old man with no medical history presents with sudden onset of lightheadedness followed by syncope. He denies any chest pain or other associated symptoms. At the time of evaluation, he is asymptomatic. His EKG (FIGURE 3) is diagnostic of what syndrome?

The EKG abnormality and diagnosis. The patient had a classic presentation for Wolff-Parkinson-White (WPW) syndrome, a common congenital disorder that alters normal cardiac conduction. He described 2 past instances of unexplained light-headedness and palpitations. Subsequent EP studies demonstrated that the patient had an accessory atrioventricular (AV) tract, causing electrical activity in the heart to bypass the AV node, resulting in a delta wave on EKG (FIGURE 3, GREEN ARROWS).

The patient opted for ablation therapy, which successfully eliminated the delta wave on EKG. Five years later he has had no recurrences.

Epidemiology. The prevalence of WPW syndrome is .1% to 3%.²³ Accessory AV tracts are found in men twice as often as in women. Only half of individuals with confirmed tracts develop a tachyarrhythmia. The estimated risk of sudden death due to WPW syndrome is .5% to 4%.²⁴

Pathophysiology. Normally cardiac conduction originates from the sinus node and travels to the AV node, where conduction is slowed, and then proceeds to the His-Purkinje system, and finally to the rest of the ventricular myocardium. In WPW syndrome, ventricular depolarization occurs first by an accessory AV tract called the bundle of Kent, followed shortly thereafter by the His-Purkinje system. This sequence of depolarization is what leads to the EKG findings characteristic of WPW syndrome: a PR interval <.12 seconds, presence of a delta wave, widened QRS complex (>.12 seconds), and repolarization changes seen as ST segment and T-wave changes discordant to (opposite direction) the delta wave and QRS complex (FIGURE 3, RED ARROW).

Factors that influence electrical conduction through the bundle of Kent include cardioactive medications, physiological stress, circulating catecholamines, coronary ischemia, and aging. The end result is a propensity for the heart to convert to one of 4 arrhythmias: paroxysmal supraventricular tachycardia (PSVT), AF, atrial flutter, or ventricular fibrillation (TABLE 2).²³

The most common arrhythmia in WPW syndrome is PSVT.²³ This rhythm is induced by the formation of a reentry circuit—a pattern in which the heart’s electrical signal loops back on itself—involving the normal conduction pathway and the bundle of Kent. Reentry progressing down the His-Purkinje system and traveling up the bundle of Kent is referred to as orthodromic (anterograde) PSVT. Antidromic (retrograde) PSVT is due to a reentry circuit conducting from the bundle of Kent to the ventricles, and then retrograde through the His-Purkinje system and AV node to the atria.

Clinical features. Under normal circumstances, patients with WPW syndrome are asymptomatic. As was the case with our patient, individuals who develop one of the 4 characteristic arrhythmias can experience light-headedness and syncope.

Treatment. An unstable patient who is experiencing PSVT, AF, or atrial flutter should receive synchronized cardioversion; those experiencing ventricular fibrillation should receive defibrillation (TABLE 2).²³ For stable patients, therapy is tailored to the type of arrhythmia. Calcium channel blockers, beta-blockers, and adenosine might be appropriate for patients with orthodromic PSVT but should be avoided in patients with antidromic PSVT, AF, or atrial flutter because these medications block AV node conduction and thus facilitate conduction down the bundle of Kent, which can result in potentially unstable arrhythmias. In general, the longer an arrhythmia has been present, the less effective the pharmacologic intervention because of the increasing sympathetic tone.

An unstable patient who is experiencing PSVT, AF, or atrial flutter associated with Wolff-Parkinson-White syndrome should receive synchronized cardioversion.Preventive long-term therapies for WPW patients who have experienced arrhythmia include antiarrhythmic medications or ablative procedures. Long-term antiarrhythmic therapy often is reserved for older, more sedentary individuals with less frequent arrhythmias that are not life-threatening. Radiofrequency ablation is a popular option, with long-term success rates as high as 95% and complication rates <1%.²³ Patients in whom a WPW pattern is identified incidentally on EKG should be referred to cardiology for EP studies and risk stratification.²⁵

CASE 4 › A 61-year-old woman has an episode of substernal exertional chest pressure that lasted approximately 2 hours but resolved before she arrived at her physician’s office. She also experienced mild nausea. She has no history of coronary artery disease but says that she has experienced similar episodes of chest pressure. What abnormality is seen on her EKG (FIGURE 4)? What is the most likely cause of her symptoms?

The EKG abnormality and diagnosis. Although classically associated with syncope, hypertrophic cardiomyopathy (HCM) often presents similarly to acute coronary syndromes, with chest pain and dyspnea on exertion.²⁶ This patient had no history of cardiac disease or family history of SCD or cardiomyopathy; however, her EKG showed changes indicating left ventricular hypertrophy (LVH), which is consistent with HCM (FIGURE 4, ARROWS). Echocardiography identified myocardial hypertrophy, normal left ventricular ejection fraction, but severe left ventricular outflow obstruction and mild diastolic dysfunction. She was treated with metoprolol and verapamil.

Etiology and epidemiology. Hypertrophic cardiomyopathy is an autosomal dominant intrinsic myocardial disorder resulting in LVH that is commonly associated with SCD during extreme physical activity.^26,27 The prevalence of HCM is approximately 1 in 500.²⁶ Although it can present at any age, it is the most common cause of SCD in young people (under age 30), responsible for 33% of deaths during athletic events.²⁸

TABLE 3
4 diagnoses and what you'll see on EKG

^{EKG, electrocardiogram; RBBB, right bundle branch block.}

Clinical features. The severity of HCM ranges from asymptomatic to fatal. Symptoms of HCM include chest pain, dyspnea, and syncope. The disorder causes morbidity and mortality in at least one of 3 ways: ventricular tachyarrhythmias (often in younger patients), heart failure (from left ventricular outflow obstruction), and/or thromboembolism.²⁷

Although classically associated with syncope, hypertrophic cardiomyopathy often presents similarly to acute coronary syndromes.Although echocardiography typically is used to make the diagnosis,²⁷ an EKG often is the initial screening tool. EKG changes are seen in 75% to 95% of affected patients and include the presence of Q-waves and increased voltages related to LVH.^27,29 Infarct-like patterns may be present before wall thickening on echocardiogram. Abnormal Q-waves are found in 20% to 50% of HCM patients, and are more common in younger patients. Konno et al³⁰ have shown that Q-waves >3 mm in depth and/or >.040 seconds in duration in at least 2 leads other than aVR is specific (90%) in identifying carriers of HCM genes before they develop clinical symptoms.

Ambulatory monitoring may be useful for risk stratifying HCM patients; those with nonsustained ventricular tachycardia (NSVT) are at higher risk of SCD. Holter monitoring is recommended in initial evaluation because evidence of ventricular tachyarrhythmias may warrant AICD placement.²⁶

Treatment. The risk of SCD in HCM is approximately 1%, but higher in those with a family history of SCD, syncope, NSVT, hypotension during exercise, or severe LVH (left ventricle thickness >30 mm).²⁶ AICDs are recommended for HCM patients with prior cardiac arrest, patients with ≥2 of these risk factors, or patients with one risk factor who have experienced syncope related to arrhythmia.²⁶

For patients who are symptomatic but have <2 risk factors, beta-blockers are firstline therapy.²⁶ Verapamil is used as a second line treatment. Both beta-blockers and calcium channel blockers reduce dyspnea, palpitations, and chest pain.²⁷

For hypertrophic cardiomyopathy patients who are symptomatic but have <2 risk factors, beta-blockers are first-line therapy.For patients who don’t respond to medical therapy, septal reduction therapy may be performed, either by septal myectomy or alcohol septal ablation.²⁷ It is also important to consider genetic screening and counseling for the family.

A summary of all 4 diagnoses described in this article, their associated EKG findings, and their pathophysiology appears in TABLE 3.

CORRESPONDENCE
Samir Haydar, DO, MPH, FACEP, Tufts University School of Medicine, Maine Medical Center, Department of Emergency Medicine, 47 Bramhall St., Portland, ME 04103; [email protected]

When evaluating a patient with a history of chest pain, palpitations, syncope, and/or new-onset seizures, an electrocardiogram (EKG) may be the key to identifying a potentially life-threatening condition. Here we present 4 cases in which EKG findings were the clue to underlying medical conditions that, if left untreated, could be fatal. Because each of these conditions may not have associated findings on a physical exam, early recognition of these EKG findings can be lifesaving. 

CASE 1 › A 15-year-old boy suddenly collapses while walking, and bystanders report seizure-like activity. The patient doesn’t remember the event. Vital signs and physical exam are normal, and his blood glucose level is 86 mg/dL (normal: 70-100 mg/dL). He doesn’t take any medications and denies illicit drug use or recent illness.

What EKG abnormality (FIGURE 1) likely explains the cause of the patient’s collapse?

The EKG abnormality and diagnosis. The patient’s EKG revealed a prolonged QT interval (FIGURE 1, BRACKETS). His QTc (QT interval corrected for heart rate) was .470 seconds, which is at the high end of the normal range for his age and gender.¹ The patient had no family history of syncope, sudden cardiac death (SCD), or seizure disorder. Evaluation uncovered a calcium level of 4.4 mEq/L (normal: 4.5-5.5 mEq/L) and a phosphate level of 7.8 mg/dL (normal: 2.4-4.1 mg/dL).

This patient had a low parathyroid hormone from primary hypoparathyroidism. His conduction abnormality was treated with both oral calcium and vitamin D supplements.

Etiology and epidemiology. A prolonged QT interval may be the result of a primary long QT syndrome (LQTS) or an acquired condition from electrolyte imbalance, medication effect, or toxin exposure.

In the United States, the incidence of a genetic mutation that causes LQTS is 1 in 2500 people.² Patients with LQTS usually remain asymptomatic unless the QT interval is further prolonged by a condition or medication. There are several hundred congenital LQTS subtypes based on specific ion channel defects; the most common is LQTS1, with an inherited defect in the KCNQ1 gene, which regulates the slow potassium ion channel.

Acquired LQTS is much more common than congenital LQTS.³ Many drugs have been linked to an increased risk of LQTS, including certain antiarrhythmics, antibiotics, and antipsychotics (TABLE 1).⁴ In addition, electrolyte disturbances such as hypokalemia, hypocalcemia, and hypomagnesemia can be etiologic factors.

Be aware that an acquired LQTS may mask an underlying congenital LQTS. Therefore, patients in whom the offending agent or condition is corrected should still have a follow-up EKG. Screening family members for LQTS is worthwhile, even in those without symptoms.

Clinical features. Patients with symptomatic LQTS may have dizziness, palpitations, and syncope. SCD also is possible. These signs and symptoms may be triggered by strong emotions (in LQTS2) or physical activity (in LQTS1). They likely are caused by torsades de pointes and ventricular fibrillation. A brief aura may precede these arrhythmias, and patients may experience urinary or fecal incontinence.⁵

A common error in measuring the QT interval occurs when physicians inadvertently include a U-wave in the measurement.The key to making a diagnosis of LQTS is correctly measuring the QT interval. The QT interval is measured from the beginning of the Q-wave to the end of the T-wave as measured from the intersection of a line tangent to the downslope of the T-wave and the isoelectric line. This can be difficult to determine in EKGs showing bundle branch block or an irregular rhythm, such as atrial fibrillation (AF).^6,7 A common error in measuring the QT interval occurs when clinicians inadvertently include a U-wave in the measurement.¹ Some EKG machines may provide QT interval and QTc measurements. Normal QT intervals are ≤.450 seconds for men and ≤.470 seconds for women.⁸

It is essential to confirm the QT interval by using the Bazett formula (QTc equals the QT in seconds divided by the square root of the RR interval in seconds) for all patients with a history that suggests a possible arrhythmia.

First rule out or treat any causes of acquired LQTS by taking a medication history and evaluating the patient's electrolytes.Our patient had hypocalcemia, which on an EKG can cause T-wave widening with a normal ST segment, rather than a normal T-wave with a long ST segment, as is typically seen in LQTS. This distinction may be difficult to discern and should not preclude the search for either an acquired prolonged QTc or an underlying LQTS.⁹

Treatment. First rule out or treat any causes of acquired LQTS by taking a careful medication history and evaluating the patient’s electrolytes. Once these have been addressed, a beta-blocker is first-line therapy for symptomatic patients.⁵

Unfortunately, up to 20% of individuals treated with beta-blockers may continue to have syncope.⁵ For these patients, options include a left cardiac sympathetic denervation (LCSD) or placement of an automatic implantable cardioverter-defibrillator (AICD). An LCSD involves removal of the left-sided stellate and/or thoracic ganglia. This procedure can be used instead of, or in addition to, beta-blockers. If the patient’s syncope persists, AICDs are an option. AICDs can be lifesaving, but patients run the risk of adverse effects that include inappropriate shocks and infection.¹⁰ As the result of these therapies, mortality associated with LQTS has dropped to approximately 1%.¹¹

CASE 2 › A 14-year-old boy has a syncopal episode while at rest. A similar event occurred 3 years earlier; at that time, an echocardiogram and EKG were normal. For 2 days, he’s had a cough and low-grade fever. His temperature is 102ºF and he has a productive cough. Based on this EKG (FIGURE 2), what is the likely diagnosis? What is the significance of his fever?

The EKG abnormality and diagnosis. This patient’s EKG showed a type 1 Brugada pattern (FIGURE 2, ARROWS), which strongly supported the diagnosis of Brugada syndrome (BS). BS is an inherited condition caused by a genetic defect in cardiac ion channel function that leads to characteristic EKG changes and a predisposition to ventricular fibrillation.¹² In this case, the fever likely unmasked these EKG findings.

The patient was transferred to a local hospital for treatment of community-acquired pneumonia, and ultimately received an AICD.

Etiology and epidemiology. BS was first described in 1992¹³ and is a major cause of SCD, responsible for up to 4% of all cases of SCD, and 20% of cases of patients without structural heart disease.14 BS is more common in men, and the mean age of diagnosis is 40 to 45.^15-18

Mutations in at least 17 cardiac ion channel genes have been linked to BS.¹⁹ The SCN5A gene—a cardiac sodium channel—is the most commonly implicated, but accounts for only 11% to 24% of all BS cases.¹⁵

Clinical features. Patients with BS may present with syncope, nocturnal agonal respirations, or ventricular arrhythmias.¹² EKG findings include partial or complete right bundle branch block (RBBB) and ST segment elevation in right precordial leads V1 to V3.¹² There are 2 Brugada EKG patterns, a type 1 cloved pattern as seen in our patient’s EKG and a type 2 saddleback pattern.20 EKG findings are dynamic over time and may alternate between normal, type 1, and type 2.²⁰ Factors that modulate EKG appearance include fever, intoxication, vagal tone, electrolyte imbalance, and sodium channel blockade.^12,20

Diagnosis requires a type 1 Brugada pattern on EKG plus a family history of BS, documented ventricular arrhythmia, or arrhythmia-related symptoms such as syncope.¹² Patients with a type 2 Brugada pattern may undergo electrophysiology (EP) testing with Class I antiarrhythmic drugs to induce a diagnostic type 1 Brugada pattern.^12,21 Patients who have Brugada EKG findings but none of the other diagnostic criteria are considered to have a Brugada pattern (rather than Brugada syndrome).¹²

The most concerning outcome of BS is ventricular fibrillation. The estimated annual rate of cardiac events is 7.7% among patients who have experienced an aborted SCD, 1.9% among those who have experienced syncope, and 0.5% in asymptomatic patients.¹⁸

Treatment. The only effective treatment for BS is placement of an AICD; however, complications of AICD placement cause significant morbidity.⁶ Ten years after AICD placement, 37% of patients experienced inappropriate shocks and 29% experienced lead failure.²² Recent modifications in device programming and the addition of remote monitoring have decreased complication rates.^12,22

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal.The decision to place an AICD is based on the patient’s prior symptoms, EKG findings, and other factors. Recent guidelines recommend an AICD for all patients with a type 1 Brugada pattern (spontaneous or induced) who also have had an aborted SCD, syncope, or documented ventricular arrhythmia.¹²

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal. Asymptomatic patients with type 1 Brugada findings should undergo further EP testing, and should receive AICD only upon demonstration of inducible ventricular arrhythmia.¹²

TABLE 2
Arrhythmias associated with Wolff-Parkinson-White syndrome²³

^{EKG, electrocardiogram; N/A, not applicable; PSVT, paroxysmal supraventricular tachycardia.}

CASE 3 › A 21-year-old man with no medical history presents with sudden onset of lightheadedness followed by syncope. He denies any chest pain or other associated symptoms. At the time of evaluation, he is asymptomatic. His EKG (FIGURE 3) is diagnostic of what syndrome?

The EKG abnormality and diagnosis. The patient had a classic presentation for Wolff-Parkinson-White (WPW) syndrome, a common congenital disorder that alters normal cardiac conduction. He described 2 past instances of unexplained light-headedness and palpitations. Subsequent EP studies demonstrated that the patient had an accessory atrioventricular (AV) tract, causing electrical activity in the heart to bypass the AV node, resulting in a delta wave on EKG (FIGURE 3, GREEN ARROWS).

The patient opted for ablation therapy, which successfully eliminated the delta wave on EKG. Five years later he has had no recurrences.

Epidemiology. The prevalence of WPW syndrome is .1% to 3%.²³ Accessory AV tracts are found in men twice as often as in women. Only half of individuals with confirmed tracts develop a tachyarrhythmia. The estimated risk of sudden death due to WPW syndrome is .5% to 4%.²⁴

Pathophysiology. Normally cardiac conduction originates from the sinus node and travels to the AV node, where conduction is slowed, and then proceeds to the His-Purkinje system, and finally to the rest of the ventricular myocardium. In WPW syndrome, ventricular depolarization occurs first by an accessory AV tract called the bundle of Kent, followed shortly thereafter by the His-Purkinje system. This sequence of depolarization is what leads to the EKG findings characteristic of WPW syndrome: a PR interval <.12 seconds, presence of a delta wave, widened QRS complex (>.12 seconds), and repolarization changes seen as ST segment and T-wave changes discordant to (opposite direction) the delta wave and QRS complex (FIGURE 3, RED ARROW).

Factors that influence electrical conduction through the bundle of Kent include cardioactive medications, physiological stress, circulating catecholamines, coronary ischemia, and aging. The end result is a propensity for the heart to convert to one of 4 arrhythmias: paroxysmal supraventricular tachycardia (PSVT), AF, atrial flutter, or ventricular fibrillation (TABLE 2).²³

The most common arrhythmia in WPW syndrome is PSVT.²³ This rhythm is induced by the formation of a reentry circuit—a pattern in which the heart’s electrical signal loops back on itself—involving the normal conduction pathway and the bundle of Kent. Reentry progressing down the His-Purkinje system and traveling up the bundle of Kent is referred to as orthodromic (anterograde) PSVT. Antidromic (retrograde) PSVT is due to a reentry circuit conducting from the bundle of Kent to the ventricles, and then retrograde through the His-Purkinje system and AV node to the atria.

Clinical features. Under normal circumstances, patients with WPW syndrome are asymptomatic. As was the case with our patient, individuals who develop one of the 4 characteristic arrhythmias can experience light-headedness and syncope.

Treatment. An unstable patient who is experiencing PSVT, AF, or atrial flutter should receive synchronized cardioversion; those experiencing ventricular fibrillation should receive defibrillation (TABLE 2).²³ For stable patients, therapy is tailored to the type of arrhythmia. Calcium channel blockers, beta-blockers, and adenosine might be appropriate for patients with orthodromic PSVT but should be avoided in patients with antidromic PSVT, AF, or atrial flutter because these medications block AV node conduction and thus facilitate conduction down the bundle of Kent, which can result in potentially unstable arrhythmias. In general, the longer an arrhythmia has been present, the less effective the pharmacologic intervention because of the increasing sympathetic tone.

An unstable patient who is experiencing PSVT, AF, or atrial flutter associated with Wolff-Parkinson-White syndrome should receive synchronized cardioversion.Preventive long-term therapies for WPW patients who have experienced arrhythmia include antiarrhythmic medications or ablative procedures. Long-term antiarrhythmic therapy often is reserved for older, more sedentary individuals with less frequent arrhythmias that are not life-threatening. Radiofrequency ablation is a popular option, with long-term success rates as high as 95% and complication rates <1%.²³ Patients in whom a WPW pattern is identified incidentally on EKG should be referred to cardiology for EP studies and risk stratification.²⁵

CASE 4 › A 61-year-old woman has an episode of substernal exertional chest pressure that lasted approximately 2 hours but resolved before she arrived at her physician’s office. She also experienced mild nausea. She has no history of coronary artery disease but says that she has experienced similar episodes of chest pressure. What abnormality is seen on her EKG (FIGURE 4)? What is the most likely cause of her symptoms?

The EKG abnormality and diagnosis. Although classically associated with syncope, hypertrophic cardiomyopathy (HCM) often presents similarly to acute coronary syndromes, with chest pain and dyspnea on exertion.²⁶ This patient had no history of cardiac disease or family history of SCD or cardiomyopathy; however, her EKG showed changes indicating left ventricular hypertrophy (LVH), which is consistent with HCM (FIGURE 4, ARROWS). Echocardiography identified myocardial hypertrophy, normal left ventricular ejection fraction, but severe left ventricular outflow obstruction and mild diastolic dysfunction. She was treated with metoprolol and verapamil.

Etiology and epidemiology. Hypertrophic cardiomyopathy is an autosomal dominant intrinsic myocardial disorder resulting in LVH that is commonly associated with SCD during extreme physical activity.^26,27 The prevalence of HCM is approximately 1 in 500.²⁶ Although it can present at any age, it is the most common cause of SCD in young people (under age 30), responsible for 33% of deaths during athletic events.²⁸

TABLE 3
4 diagnoses and what you'll see on EKG

^{EKG, electrocardiogram; RBBB, right bundle branch block.}

Clinical features. The severity of HCM ranges from asymptomatic to fatal. Symptoms of HCM include chest pain, dyspnea, and syncope. The disorder causes morbidity and mortality in at least one of 3 ways: ventricular tachyarrhythmias (often in younger patients), heart failure (from left ventricular outflow obstruction), and/or thromboembolism.²⁷

Although classically associated with syncope, hypertrophic cardiomyopathy often presents similarly to acute coronary syndromes.Although echocardiography typically is used to make the diagnosis,²⁷ an EKG often is the initial screening tool. EKG changes are seen in 75% to 95% of affected patients and include the presence of Q-waves and increased voltages related to LVH.^27,29 Infarct-like patterns may be present before wall thickening on echocardiogram. Abnormal Q-waves are found in 20% to 50% of HCM patients, and are more common in younger patients. Konno et al³⁰ have shown that Q-waves >3 mm in depth and/or >.040 seconds in duration in at least 2 leads other than aVR is specific (90%) in identifying carriers of HCM genes before they develop clinical symptoms.

Ambulatory monitoring may be useful for risk stratifying HCM patients; those with nonsustained ventricular tachycardia (NSVT) are at higher risk of SCD. Holter monitoring is recommended in initial evaluation because evidence of ventricular tachyarrhythmias may warrant AICD placement.²⁶

Treatment. The risk of SCD in HCM is approximately 1%, but higher in those with a family history of SCD, syncope, NSVT, hypotension during exercise, or severe LVH (left ventricle thickness >30 mm).²⁶ AICDs are recommended for HCM patients with prior cardiac arrest, patients with ≥2 of these risk factors, or patients with one risk factor who have experienced syncope related to arrhythmia.²⁶

For patients who are symptomatic but have <2 risk factors, beta-blockers are firstline therapy.²⁶ Verapamil is used as a second line treatment. Both beta-blockers and calcium channel blockers reduce dyspnea, palpitations, and chest pain.²⁷

For hypertrophic cardiomyopathy patients who are symptomatic but have <2 risk factors, beta-blockers are first-line therapy.For patients who don’t respond to medical therapy, septal reduction therapy may be performed, either by septal myectomy or alcohol septal ablation.²⁷ It is also important to consider genetic screening and counseling for the family.

A summary of all 4 diagnoses described in this article, their associated EKG findings, and their pathophysiology appears in TABLE 3.

CORRESPONDENCE
Samir Haydar, DO, MPH, FACEP, Tufts University School of Medicine, Maine Medical Center, Department of Emergency Medicine, 47 Bramhall St., Portland, ME 04103; [email protected]

When evaluating a patient with a history of chest pain, palpitations, syncope, and/or new-onset seizures, an electrocardiogram (EKG) may be the key to identifying a potentially life-threatening condition. Here we present 4 cases in which EKG findings were the clue to underlying medical conditions that, if left untreated, could be fatal. Because each of these conditions may not have associated findings on a physical exam, early recognition of these EKG findings can be lifesaving. 

CASE 1 › A 15-year-old boy suddenly collapses while walking, and bystanders report seizure-like activity. The patient doesn’t remember the event. Vital signs and physical exam are normal, and his blood glucose level is 86 mg/dL (normal: 70-100 mg/dL). He doesn’t take any medications and denies illicit drug use or recent illness.

What EKG abnormality (FIGURE 1) likely explains the cause of the patient’s collapse?

The EKG abnormality and diagnosis. The patient’s EKG revealed a prolonged QT interval (FIGURE 1, BRACKETS). His QTc (QT interval corrected for heart rate) was .470 seconds, which is at the high end of the normal range for his age and gender.¹ The patient had no family history of syncope, sudden cardiac death (SCD), or seizure disorder. Evaluation uncovered a calcium level of 4.4 mEq/L (normal: 4.5-5.5 mEq/L) and a phosphate level of 7.8 mg/dL (normal: 2.4-4.1 mg/dL).

This patient had a low parathyroid hormone from primary hypoparathyroidism. His conduction abnormality was treated with both oral calcium and vitamin D supplements.

Etiology and epidemiology. A prolonged QT interval may be the result of a primary long QT syndrome (LQTS) or an acquired condition from electrolyte imbalance, medication effect, or toxin exposure.

In the United States, the incidence of a genetic mutation that causes LQTS is 1 in 2500 people.² Patients with LQTS usually remain asymptomatic unless the QT interval is further prolonged by a condition or medication. There are several hundred congenital LQTS subtypes based on specific ion channel defects; the most common is LQTS1, with an inherited defect in the KCNQ1 gene, which regulates the slow potassium ion channel.

Acquired LQTS is much more common than congenital LQTS.³ Many drugs have been linked to an increased risk of LQTS, including certain antiarrhythmics, antibiotics, and antipsychotics (TABLE 1).⁴ In addition, electrolyte disturbances such as hypokalemia, hypocalcemia, and hypomagnesemia can be etiologic factors.

Be aware that an acquired LQTS may mask an underlying congenital LQTS. Therefore, patients in whom the offending agent or condition is corrected should still have a follow-up EKG. Screening family members for LQTS is worthwhile, even in those without symptoms.

Clinical features. Patients with symptomatic LQTS may have dizziness, palpitations, and syncope. SCD also is possible. These signs and symptoms may be triggered by strong emotions (in LQTS2) or physical activity (in LQTS1). They likely are caused by torsades de pointes and ventricular fibrillation. A brief aura may precede these arrhythmias, and patients may experience urinary or fecal incontinence.⁵

A common error in measuring the QT interval occurs when physicians inadvertently include a U-wave in the measurement.The key to making a diagnosis of LQTS is correctly measuring the QT interval. The QT interval is measured from the beginning of the Q-wave to the end of the T-wave as measured from the intersection of a line tangent to the downslope of the T-wave and the isoelectric line. This can be difficult to determine in EKGs showing bundle branch block or an irregular rhythm, such as atrial fibrillation (AF).^6,7 A common error in measuring the QT interval occurs when clinicians inadvertently include a U-wave in the measurement.¹ Some EKG machines may provide QT interval and QTc measurements. Normal QT intervals are ≤.450 seconds for men and ≤.470 seconds for women.⁸

It is essential to confirm the QT interval by using the Bazett formula (QTc equals the QT in seconds divided by the square root of the RR interval in seconds) for all patients with a history that suggests a possible arrhythmia.

First rule out or treat any causes of acquired LQTS by taking a medication history and evaluating the patient's electrolytes.Our patient had hypocalcemia, which on an EKG can cause T-wave widening with a normal ST segment, rather than a normal T-wave with a long ST segment, as is typically seen in LQTS. This distinction may be difficult to discern and should not preclude the search for either an acquired prolonged QTc or an underlying LQTS.⁹

Treatment. First rule out or treat any causes of acquired LQTS by taking a careful medication history and evaluating the patient’s electrolytes. Once these have been addressed, a beta-blocker is first-line therapy for symptomatic patients.⁵

Unfortunately, up to 20% of individuals treated with beta-blockers may continue to have syncope.⁵ For these patients, options include a left cardiac sympathetic denervation (LCSD) or placement of an automatic implantable cardioverter-defibrillator (AICD). An LCSD involves removal of the left-sided stellate and/or thoracic ganglia. This procedure can be used instead of, or in addition to, beta-blockers. If the patient’s syncope persists, AICDs are an option. AICDs can be lifesaving, but patients run the risk of adverse effects that include inappropriate shocks and infection.¹⁰ As the result of these therapies, mortality associated with LQTS has dropped to approximately 1%.¹¹

CASE 2 › A 14-year-old boy has a syncopal episode while at rest. A similar event occurred 3 years earlier; at that time, an echocardiogram and EKG were normal. For 2 days, he’s had a cough and low-grade fever. His temperature is 102ºF and he has a productive cough. Based on this EKG (FIGURE 2), what is the likely diagnosis? What is the significance of his fever?

The EKG abnormality and diagnosis. This patient’s EKG showed a type 1 Brugada pattern (FIGURE 2, ARROWS), which strongly supported the diagnosis of Brugada syndrome (BS). BS is an inherited condition caused by a genetic defect in cardiac ion channel function that leads to characteristic EKG changes and a predisposition to ventricular fibrillation.¹² In this case, the fever likely unmasked these EKG findings.

The patient was transferred to a local hospital for treatment of community-acquired pneumonia, and ultimately received an AICD.

Etiology and epidemiology. BS was first described in 1992¹³ and is a major cause of SCD, responsible for up to 4% of all cases of SCD, and 20% of cases of patients without structural heart disease.14 BS is more common in men, and the mean age of diagnosis is 40 to 45.^15-18

Mutations in at least 17 cardiac ion channel genes have been linked to BS.¹⁹ The SCN5A gene—a cardiac sodium channel—is the most commonly implicated, but accounts for only 11% to 24% of all BS cases.¹⁵

Clinical features. Patients with BS may present with syncope, nocturnal agonal respirations, or ventricular arrhythmias.¹² EKG findings include partial or complete right bundle branch block (RBBB) and ST segment elevation in right precordial leads V1 to V3.¹² There are 2 Brugada EKG patterns, a type 1 cloved pattern as seen in our patient’s EKG and a type 2 saddleback pattern.20 EKG findings are dynamic over time and may alternate between normal, type 1, and type 2.²⁰ Factors that modulate EKG appearance include fever, intoxication, vagal tone, electrolyte imbalance, and sodium channel blockade.^12,20

Diagnosis requires a type 1 Brugada pattern on EKG plus a family history of BS, documented ventricular arrhythmia, or arrhythmia-related symptoms such as syncope.¹² Patients with a type 2 Brugada pattern may undergo electrophysiology (EP) testing with Class I antiarrhythmic drugs to induce a diagnostic type 1 Brugada pattern.^12,21 Patients who have Brugada EKG findings but none of the other diagnostic criteria are considered to have a Brugada pattern (rather than Brugada syndrome).¹²

The most concerning outcome of BS is ventricular fibrillation. The estimated annual rate of cardiac events is 7.7% among patients who have experienced an aborted SCD, 1.9% among those who have experienced syncope, and 0.5% in asymptomatic patients.¹⁸

Treatment. The only effective treatment for BS is placement of an AICD; however, complications of AICD placement cause significant morbidity.⁶ Ten years after AICD placement, 37% of patients experienced inappropriate shocks and 29% experienced lead failure.²² Recent modifications in device programming and the addition of remote monitoring have decreased complication rates.^12,22

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal.The decision to place an AICD is based on the patient’s prior symptoms, EKG findings, and other factors. Recent guidelines recommend an AICD for all patients with a type 1 Brugada pattern (spontaneous or induced) who also have had an aborted SCD, syncope, or documented ventricular arrhythmia.¹²

Management of asymptomatic patients with type 1 Brugada pattern remains controversial because the rate of cardiac events is low, although such events can be fatal. Asymptomatic patients with type 1 Brugada findings should undergo further EP testing, and should receive AICD only upon demonstration of inducible ventricular arrhythmia.¹²

TABLE 2
Arrhythmias associated with Wolff-Parkinson-White syndrome²³

^{EKG, electrocardiogram; N/A, not applicable; PSVT, paroxysmal supraventricular tachycardia.}

CASE 3 › A 21-year-old man with no medical history presents with sudden onset of lightheadedness followed by syncope. He denies any chest pain or other associated symptoms. At the time of evaluation, he is asymptomatic. His EKG (FIGURE 3) is diagnostic of what syndrome?

The EKG abnormality and diagnosis. The patient had a classic presentation for Wolff-Parkinson-White (WPW) syndrome, a common congenital disorder that alters normal cardiac conduction. He described 2 past instances of unexplained light-headedness and palpitations. Subsequent EP studies demonstrated that the patient had an accessory atrioventricular (AV) tract, causing electrical activity in the heart to bypass the AV node, resulting in a delta wave on EKG (FIGURE 3, GREEN ARROWS).

The patient opted for ablation therapy, which successfully eliminated the delta wave on EKG. Five years later he has had no recurrences.

Epidemiology. The prevalence of WPW syndrome is .1% to 3%.²³ Accessory AV tracts are found in men twice as often as in women. Only half of individuals with confirmed tracts develop a tachyarrhythmia. The estimated risk of sudden death due to WPW syndrome is .5% to 4%.²⁴

Pathophysiology. Normally cardiac conduction originates from the sinus node and travels to the AV node, where conduction is slowed, and then proceeds to the His-Purkinje system, and finally to the rest of the ventricular myocardium. In WPW syndrome, ventricular depolarization occurs first by an accessory AV tract called the bundle of Kent, followed shortly thereafter by the His-Purkinje system. This sequence of depolarization is what leads to the EKG findings characteristic of WPW syndrome: a PR interval <.12 seconds, presence of a delta wave, widened QRS complex (>.12 seconds), and repolarization changes seen as ST segment and T-wave changes discordant to (opposite direction) the delta wave and QRS complex (FIGURE 3, RED ARROW).

Factors that influence electrical conduction through the bundle of Kent include cardioactive medications, physiological stress, circulating catecholamines, coronary ischemia, and aging. The end result is a propensity for the heart to convert to one of 4 arrhythmias: paroxysmal supraventricular tachycardia (PSVT), AF, atrial flutter, or ventricular fibrillation (TABLE 2).²³

The most common arrhythmia in WPW syndrome is PSVT.²³ This rhythm is induced by the formation of a reentry circuit—a pattern in which the heart’s electrical signal loops back on itself—involving the normal conduction pathway and the bundle of Kent. Reentry progressing down the His-Purkinje system and traveling up the bundle of Kent is referred to as orthodromic (anterograde) PSVT. Antidromic (retrograde) PSVT is due to a reentry circuit conducting from the bundle of Kent to the ventricles, and then retrograde through the His-Purkinje system and AV node to the atria.

Clinical features. Under normal circumstances, patients with WPW syndrome are asymptomatic. As was the case with our patient, individuals who develop one of the 4 characteristic arrhythmias can experience light-headedness and syncope.

Treatment. An unstable patient who is experiencing PSVT, AF, or atrial flutter should receive synchronized cardioversion; those experiencing ventricular fibrillation should receive defibrillation (TABLE 2).²³ For stable patients, therapy is tailored to the type of arrhythmia. Calcium channel blockers, beta-blockers, and adenosine might be appropriate for patients with orthodromic PSVT but should be avoided in patients with antidromic PSVT, AF, or atrial flutter because these medications block AV node conduction and thus facilitate conduction down the bundle of Kent, which can result in potentially unstable arrhythmias. In general, the longer an arrhythmia has been present, the less effective the pharmacologic intervention because of the increasing sympathetic tone.

An unstable patient who is experiencing PSVT, AF, or atrial flutter associated with Wolff-Parkinson-White syndrome should receive synchronized cardioversion.Preventive long-term therapies for WPW patients who have experienced arrhythmia include antiarrhythmic medications or ablative procedures. Long-term antiarrhythmic therapy often is reserved for older, more sedentary individuals with less frequent arrhythmias that are not life-threatening. Radiofrequency ablation is a popular option, with long-term success rates as high as 95% and complication rates <1%.²³ Patients in whom a WPW pattern is identified incidentally on EKG should be referred to cardiology for EP studies and risk stratification.²⁵

CASE 4 › A 61-year-old woman has an episode of substernal exertional chest pressure that lasted approximately 2 hours but resolved before she arrived at her physician’s office. She also experienced mild nausea. She has no history of coronary artery disease but says that she has experienced similar episodes of chest pressure. What abnormality is seen on her EKG (FIGURE 4)? What is the most likely cause of her symptoms?

The EKG abnormality and diagnosis. Although classically associated with syncope, hypertrophic cardiomyopathy (HCM) often presents similarly to acute coronary syndromes, with chest pain and dyspnea on exertion.²⁶ This patient had no history of cardiac disease or family history of SCD or cardiomyopathy; however, her EKG showed changes indicating left ventricular hypertrophy (LVH), which is consistent with HCM (FIGURE 4, ARROWS). Echocardiography identified myocardial hypertrophy, normal left ventricular ejection fraction, but severe left ventricular outflow obstruction and mild diastolic dysfunction. She was treated with metoprolol and verapamil.

Etiology and epidemiology. Hypertrophic cardiomyopathy is an autosomal dominant intrinsic myocardial disorder resulting in LVH that is commonly associated with SCD during extreme physical activity.^26,27 The prevalence of HCM is approximately 1 in 500.²⁶ Although it can present at any age, it is the most common cause of SCD in young people (under age 30), responsible for 33% of deaths during athletic events.²⁸

TABLE 3
4 diagnoses and what you'll see on EKG

^{EKG, electrocardiogram; RBBB, right bundle branch block.}

Clinical features. The severity of HCM ranges from asymptomatic to fatal. Symptoms of HCM include chest pain, dyspnea, and syncope. The disorder causes morbidity and mortality in at least one of 3 ways: ventricular tachyarrhythmias (often in younger patients), heart failure (from left ventricular outflow obstruction), and/or thromboembolism.²⁷

Although classically associated with syncope, hypertrophic cardiomyopathy often presents similarly to acute coronary syndromes.Although echocardiography typically is used to make the diagnosis,²⁷ an EKG often is the initial screening tool. EKG changes are seen in 75% to 95% of affected patients and include the presence of Q-waves and increased voltages related to LVH.^27,29 Infarct-like patterns may be present before wall thickening on echocardiogram. Abnormal Q-waves are found in 20% to 50% of HCM patients, and are more common in younger patients. Konno et al³⁰ have shown that Q-waves >3 mm in depth and/or >.040 seconds in duration in at least 2 leads other than aVR is specific (90%) in identifying carriers of HCM genes before they develop clinical symptoms.

Ambulatory monitoring may be useful for risk stratifying HCM patients; those with nonsustained ventricular tachycardia (NSVT) are at higher risk of SCD. Holter monitoring is recommended in initial evaluation because evidence of ventricular tachyarrhythmias may warrant AICD placement.²⁶

Treatment. The risk of SCD in HCM is approximately 1%, but higher in those with a family history of SCD, syncope, NSVT, hypotension during exercise, or severe LVH (left ventricle thickness >30 mm).²⁶ AICDs are recommended for HCM patients with prior cardiac arrest, patients with ≥2 of these risk factors, or patients with one risk factor who have experienced syncope related to arrhythmia.²⁶

For patients who are symptomatic but have <2 risk factors, beta-blockers are firstline therapy.²⁶ Verapamil is used as a second line treatment. Both beta-blockers and calcium channel blockers reduce dyspnea, palpitations, and chest pain.²⁷

For hypertrophic cardiomyopathy patients who are symptomatic but have <2 risk factors, beta-blockers are first-line therapy.For patients who don’t respond to medical therapy, septal reduction therapy may be performed, either by septal myectomy or alcohol septal ablation.²⁷ It is also important to consider genetic screening and counseling for the family.

A summary of all 4 diagnoses described in this article, their associated EKG findings, and their pathophysiology appears in TABLE 3.

CORRESPONDENCE
Samir Haydar, DO, MPH, FACEP, Tufts University School of Medicine, Maine Medical Center, Department of Emergency Medicine, 47 Bramhall St., Portland, ME 04103; [email protected]

CASE › A 6-year-old African girl, developing and growing appropriately for age, was brought to our clinic by her father with the chief complaint of “eating the textbooks at school.” The child had eaten paper for years, the father said; he never thought it unusual until her teacher brought it to his attention. The father reported that his daughter had met all developmental milestones and was up to date with her immunizations. When asked why she ate paper, the child responded, “I don’t know.”

The child was diagnosed with pica and, because we were concerned that she was eating other nonnutritive foods, we ordered hematologic studies. Her lead level (2 mcg/dL) was within the normal range; her hemoglobin/hematocrit was 10.4 g/dL/32.3%. Iron therapy was started. At follow-up 4 weeks later, the child’s paper-eating behavior had resolved.

The word pica comes from the Latin word for magpie, a bird with a reputation for eating practically anything. The Diagnostic and Statistical Manual of Mental Disorders, 5th edition, defines pica as persistent eating of nonnutritive substances for at least 1 month that is inappropriate to developmental level and not part of a culturally supported or socially normative practice.¹

Case reports on paper pica are few, but numerous reports describe other forms of the behavior, including eating ice; dirt, soil, and clay; starch; burnt matches; cardboard; hair; laundry detergent; chalk; soap; firecrackers; and metal artifacts such as coins.^2-16

Pica has been described in the literature as “underreported” and “unrecognized.” Its true prevalence is difficult to assess because most people don’t report it and the methodology of data collection varies among populations, as does the definition of pica. According to some estimates, more than 50% of children ages 18 to 36 months seek and ingest nonfood items. The practice reportedly decreases as a child ages, but an estimated 10% of children older than 12 years may engage in it.¹⁷

Pica has been reported since antiquity. Many medical and anthropological studies refer to the practice of geophagia, or dirt eating, which is prevalent in Africa and among small children and women, particularly women who are native to the southern United States, African-American, or pregnant.^5-10,18,19

Pica often occurs in people with developmental disabilities such as autism and is considered a psychiatric condition in that context.^{3,11,15,20-31} However, because many forms of pica, especially geophagia, aren’t associated with mental health issues, researchers disagree about whether to consider it an abnormal behavior. A 2000 workshop on pica organized by the Agency for Toxic Substances and Disease Registry concluded that geophagia is not an abnormal behavior.¹⁷ One of the most compelling arguments for this view is that dirt eating is far too common around the world to be considered abnormal, and dirt is held in some cultures to have therapeutic powers.^7,13,24

Adverse outcomes linked to pica

Pica is associated with adverse outcomes, however. A study by the Agency for Healthcare Research and Quality found that despite an overall decline in hospitalizations for eating disorders, hospitalizations for pica have risen.²⁵ From 1999 to 2009, pica-related hospitalizations jumped 93%, although the overall number of patients hospitalized for the condition remains small (964 in 1999 to 2000, 1862 in 2008-2009).

Adverse effects of pica include potassium abnormalities and GI conditions ranging from abdominal pain to perforation, blockage, and colon ischemia.Documented adverse effects of pica include potassium abnormalities and gastrointestinal conditions ranging from irritation and abdominal pain to perforation, blockage, and colon ischemia.^3,11,26-29 Reported bidirectional effects (which both result from and contribute to pica) include iron deficiency, parasitic infections, and heavy metal exposure—particularly lead, mercury, and arsenic.^{4,6,9,20,30-38}

Diagnosis: Focus on history and selective testing

Pica is a clinical diagnosis, confirmed by the patient’s history, not any single laboratory test. Providers should ask about pica behavior or unusual cravings in certain high-risk groups: pregnant women, particularly women from the southern United States, immigrants or refugees, and children and adults with autism or other developmental disabilities.^18,22

Testing should be based on the type of pica behavior. Because various forms of pica are commonly associated with iron-deficiency anemia, obtain serum hemoglobin and hematocrit levels along with iron levels if necessary in patients who report cravings for unusual substances. Pica in pregnancy is a sign of iron deficiency, but it also may signal iron deficiency in patients who aren’t pregnant. In one study of 262 nonpregnant adults with iron-deficiency anemia, 45% reported pica behaviors; of these, 87.3% reported eating ice.³⁴

Check serum lead levels in children who engage in geophagia since dirt may contain lead. Because ingestion of soil or clay is associated with soil-borne parasitic infections, also consider testing for ova and parasites if clinically indicated. Patients who eat paper may be exposed to mercury poisoning, so a serum mercury level is advisable.

Management: Prevention and behavior modification are key

Treatment for pica varies by patient and the specific behavior. Management approaches are primarily preventive, educational, and directed toward behavior modification.

Prevention. Residential facilities and primary care offices that care for people with developmental disabilities may screen for pica by means of prevalence surveys, direct observation, stool checks, review of medical history records, and interviews with caregivers.

Residential facilities can create a pica-safe environment by training staff in pica prevention, instituting regular on-site monitoring to ensure that no dangerous objects are available, and developing procedures to guide staff behavior, such as safe disposal of rubber gloves.²² Parents and caregivers of young children or children with developmental disabilities who don’t live in residential facilities should be aware of pica and monitor what their children are ingesting.

Behavior modification. Behavior-based approaches have proved effective for treating pica in developmentally disabled patients. Applied behavioral analysis “was found to have the most robust empirical support to treat this behavior.”³⁹ Patients found to have pica may be referred for further assessment to a behavior specialist or a psychologist with experience in treating the condition.^22,39

A review of 26 studies found that, in 25 studies, behavioral therapy reduced pica behavior by 80% or more.²³ Behavioral treatments included reinforcement procedures alone, response reduction procedures alone, and combined reinforcement and response reduction procedures. Reinforcement shapes behavior by controlling the consequences of the behavior using a combination of rewards and punishments.²³ Response reduction, or blocking, involves obstructing every attempt to eat inedible items.²²

Ask about pica behavior or unusual cravings in high-risk groups, such as pregnant women and children and adults with autism.Treatments that combined reinforcement and response reduction showed good efficacy.²³ An example of the combined approach would be to stop the patient from eating nonnutritive items while redirecting him to eat food instead.²²

Supplementation. Iron supplementation has decreased or even reversed pica in patients whose clinical symptoms and behavior were associated with iron deficiency.^35,40

Medications. Successful treatment with selective serotonin reuptake inhibitors (escitalopram), atypical neuroleptics (olanzapine), and attention-deficit/hyperactivity disorder medications (methylphenidate) has been reported in some patients, but case reports are few, and the evidence for the drugs’ efficacy is limited.^41-43

Be alert for pica. Primary care physicians need to be aware of pica and proactively seek information about cravings or behaviors suggesting the condition from patients in high-risk populations—pregnant women, children, immigrants and refugees, people with developmental disabilities—or their caregivers. Once pica is identified, clinicians should undertake appropriate laboratory investigation and behavior modification attempts.

CORRESPONDENCE
Ranit Mishori, MD, MHS, Department of Family Medicine, Georgetown University School of Medicine, Pre-Clinical Building, GB-01D, 3900 Reservoir Road, NW, Washington, DC 20007; [email protected]

CASE › A 6-year-old African girl, developing and growing appropriately for age, was brought to our clinic by her father with the chief complaint of “eating the textbooks at school.” The child had eaten paper for years, the father said; he never thought it unusual until her teacher brought it to his attention. The father reported that his daughter had met all developmental milestones and was up to date with her immunizations. When asked why she ate paper, the child responded, “I don’t know.”

The child was diagnosed with pica and, because we were concerned that she was eating other nonnutritive foods, we ordered hematologic studies. Her lead level (2 mcg/dL) was within the normal range; her hemoglobin/hematocrit was 10.4 g/dL/32.3%. Iron therapy was started. At follow-up 4 weeks later, the child’s paper-eating behavior had resolved.

The word pica comes from the Latin word for magpie, a bird with a reputation for eating practically anything. The Diagnostic and Statistical Manual of Mental Disorders, 5th edition, defines pica as persistent eating of nonnutritive substances for at least 1 month that is inappropriate to developmental level and not part of a culturally supported or socially normative practice.¹

Case reports on paper pica are few, but numerous reports describe other forms of the behavior, including eating ice; dirt, soil, and clay; starch; burnt matches; cardboard; hair; laundry detergent; chalk; soap; firecrackers; and metal artifacts such as coins.^2-16

Pica has been described in the literature as “underreported” and “unrecognized.” Its true prevalence is difficult to assess because most people don’t report it and the methodology of data collection varies among populations, as does the definition of pica. According to some estimates, more than 50% of children ages 18 to 36 months seek and ingest nonfood items. The practice reportedly decreases as a child ages, but an estimated 10% of children older than 12 years may engage in it.¹⁷

Pica has been reported since antiquity. Many medical and anthropological studies refer to the practice of geophagia, or dirt eating, which is prevalent in Africa and among small children and women, particularly women who are native to the southern United States, African-American, or pregnant.^5-10,18,19

Pica often occurs in people with developmental disabilities such as autism and is considered a psychiatric condition in that context.^{3,11,15,20-31} However, because many forms of pica, especially geophagia, aren’t associated with mental health issues, researchers disagree about whether to consider it an abnormal behavior. A 2000 workshop on pica organized by the Agency for Toxic Substances and Disease Registry concluded that geophagia is not an abnormal behavior.¹⁷ One of the most compelling arguments for this view is that dirt eating is far too common around the world to be considered abnormal, and dirt is held in some cultures to have therapeutic powers.^7,13,24

Adverse outcomes linked to pica

Pica is associated with adverse outcomes, however. A study by the Agency for Healthcare Research and Quality found that despite an overall decline in hospitalizations for eating disorders, hospitalizations for pica have risen.²⁵ From 1999 to 2009, pica-related hospitalizations jumped 93%, although the overall number of patients hospitalized for the condition remains small (964 in 1999 to 2000, 1862 in 2008-2009).

Adverse effects of pica include potassium abnormalities and GI conditions ranging from abdominal pain to perforation, blockage, and colon ischemia.Documented adverse effects of pica include potassium abnormalities and gastrointestinal conditions ranging from irritation and abdominal pain to perforation, blockage, and colon ischemia.^3,11,26-29 Reported bidirectional effects (which both result from and contribute to pica) include iron deficiency, parasitic infections, and heavy metal exposure—particularly lead, mercury, and arsenic.^{4,6,9,20,30-38}

Diagnosis: Focus on history and selective testing

Pica is a clinical diagnosis, confirmed by the patient’s history, not any single laboratory test. Providers should ask about pica behavior or unusual cravings in certain high-risk groups: pregnant women, particularly women from the southern United States, immigrants or refugees, and children and adults with autism or other developmental disabilities.^18,22

Testing should be based on the type of pica behavior. Because various forms of pica are commonly associated with iron-deficiency anemia, obtain serum hemoglobin and hematocrit levels along with iron levels if necessary in patients who report cravings for unusual substances. Pica in pregnancy is a sign of iron deficiency, but it also may signal iron deficiency in patients who aren’t pregnant. In one study of 262 nonpregnant adults with iron-deficiency anemia, 45% reported pica behaviors; of these, 87.3% reported eating ice.³⁴

Check serum lead levels in children who engage in geophagia since dirt may contain lead. Because ingestion of soil or clay is associated with soil-borne parasitic infections, also consider testing for ova and parasites if clinically indicated. Patients who eat paper may be exposed to mercury poisoning, so a serum mercury level is advisable.

Management: Prevention and behavior modification are key

Treatment for pica varies by patient and the specific behavior. Management approaches are primarily preventive, educational, and directed toward behavior modification.

Prevention. Residential facilities and primary care offices that care for people with developmental disabilities may screen for pica by means of prevalence surveys, direct observation, stool checks, review of medical history records, and interviews with caregivers.

Residential facilities can create a pica-safe environment by training staff in pica prevention, instituting regular on-site monitoring to ensure that no dangerous objects are available, and developing procedures to guide staff behavior, such as safe disposal of rubber gloves.²² Parents and caregivers of young children or children with developmental disabilities who don’t live in residential facilities should be aware of pica and monitor what their children are ingesting.

Behavior modification. Behavior-based approaches have proved effective for treating pica in developmentally disabled patients. Applied behavioral analysis “was found to have the most robust empirical support to treat this behavior.”³⁹ Patients found to have pica may be referred for further assessment to a behavior specialist or a psychologist with experience in treating the condition.^22,39

A review of 26 studies found that, in 25 studies, behavioral therapy reduced pica behavior by 80% or more.²³ Behavioral treatments included reinforcement procedures alone, response reduction procedures alone, and combined reinforcement and response reduction procedures. Reinforcement shapes behavior by controlling the consequences of the behavior using a combination of rewards and punishments.²³ Response reduction, or blocking, involves obstructing every attempt to eat inedible items.²²

Ask about pica behavior or unusual cravings in high-risk groups, such as pregnant women and children and adults with autism.Treatments that combined reinforcement and response reduction showed good efficacy.²³ An example of the combined approach would be to stop the patient from eating nonnutritive items while redirecting him to eat food instead.²²

Supplementation. Iron supplementation has decreased or even reversed pica in patients whose clinical symptoms and behavior were associated with iron deficiency.^35,40

Medications. Successful treatment with selective serotonin reuptake inhibitors (escitalopram), atypical neuroleptics (olanzapine), and attention-deficit/hyperactivity disorder medications (methylphenidate) has been reported in some patients, but case reports are few, and the evidence for the drugs’ efficacy is limited.^41-43

Be alert for pica. Primary care physicians need to be aware of pica and proactively seek information about cravings or behaviors suggesting the condition from patients in high-risk populations—pregnant women, children, immigrants and refugees, people with developmental disabilities—or their caregivers. Once pica is identified, clinicians should undertake appropriate laboratory investigation and behavior modification attempts.

CORRESPONDENCE
Ranit Mishori, MD, MHS, Department of Family Medicine, Georgetown University School of Medicine, Pre-Clinical Building, GB-01D, 3900 Reservoir Road, NW, Washington, DC 20007; [email protected]

CASE › A 6-year-old African girl, developing and growing appropriately for age, was brought to our clinic by her father with the chief complaint of “eating the textbooks at school.” The child had eaten paper for years, the father said; he never thought it unusual until her teacher brought it to his attention. The father reported that his daughter had met all developmental milestones and was up to date with her immunizations. When asked why she ate paper, the child responded, “I don’t know.”

The child was diagnosed with pica and, because we were concerned that she was eating other nonnutritive foods, we ordered hematologic studies. Her lead level (2 mcg/dL) was within the normal range; her hemoglobin/hematocrit was 10.4 g/dL/32.3%. Iron therapy was started. At follow-up 4 weeks later, the child’s paper-eating behavior had resolved.

The word pica comes from the Latin word for magpie, a bird with a reputation for eating practically anything. The Diagnostic and Statistical Manual of Mental Disorders, 5th edition, defines pica as persistent eating of nonnutritive substances for at least 1 month that is inappropriate to developmental level and not part of a culturally supported or socially normative practice.¹

Case reports on paper pica are few, but numerous reports describe other forms of the behavior, including eating ice; dirt, soil, and clay; starch; burnt matches; cardboard; hair; laundry detergent; chalk; soap; firecrackers; and metal artifacts such as coins.^2-16

Pica has been described in the literature as “underreported” and “unrecognized.” Its true prevalence is difficult to assess because most people don’t report it and the methodology of data collection varies among populations, as does the definition of pica. According to some estimates, more than 50% of children ages 18 to 36 months seek and ingest nonfood items. The practice reportedly decreases as a child ages, but an estimated 10% of children older than 12 years may engage in it.¹⁷

Pica has been reported since antiquity. Many medical and anthropological studies refer to the practice of geophagia, or dirt eating, which is prevalent in Africa and among small children and women, particularly women who are native to the southern United States, African-American, or pregnant.^5-10,18,19

Pica often occurs in people with developmental disabilities such as autism and is considered a psychiatric condition in that context.^{3,11,15,20-31} However, because many forms of pica, especially geophagia, aren’t associated with mental health issues, researchers disagree about whether to consider it an abnormal behavior. A 2000 workshop on pica organized by the Agency for Toxic Substances and Disease Registry concluded that geophagia is not an abnormal behavior.¹⁷ One of the most compelling arguments for this view is that dirt eating is far too common around the world to be considered abnormal, and dirt is held in some cultures to have therapeutic powers.^7,13,24

Adverse outcomes linked to pica

Pica is associated with adverse outcomes, however. A study by the Agency for Healthcare Research and Quality found that despite an overall decline in hospitalizations for eating disorders, hospitalizations for pica have risen.²⁵ From 1999 to 2009, pica-related hospitalizations jumped 93%, although the overall number of patients hospitalized for the condition remains small (964 in 1999 to 2000, 1862 in 2008-2009).

Adverse effects of pica include potassium abnormalities and GI conditions ranging from abdominal pain to perforation, blockage, and colon ischemia.Documented adverse effects of pica include potassium abnormalities and gastrointestinal conditions ranging from irritation and abdominal pain to perforation, blockage, and colon ischemia.^3,11,26-29 Reported bidirectional effects (which both result from and contribute to pica) include iron deficiency, parasitic infections, and heavy metal exposure—particularly lead, mercury, and arsenic.^{4,6,9,20,30-38}

Diagnosis: Focus on history and selective testing

Pica is a clinical diagnosis, confirmed by the patient’s history, not any single laboratory test. Providers should ask about pica behavior or unusual cravings in certain high-risk groups: pregnant women, particularly women from the southern United States, immigrants or refugees, and children and adults with autism or other developmental disabilities.^18,22

Testing should be based on the type of pica behavior. Because various forms of pica are commonly associated with iron-deficiency anemia, obtain serum hemoglobin and hematocrit levels along with iron levels if necessary in patients who report cravings for unusual substances. Pica in pregnancy is a sign of iron deficiency, but it also may signal iron deficiency in patients who aren’t pregnant. In one study of 262 nonpregnant adults with iron-deficiency anemia, 45% reported pica behaviors; of these, 87.3% reported eating ice.³⁴

Check serum lead levels in children who engage in geophagia since dirt may contain lead. Because ingestion of soil or clay is associated with soil-borne parasitic infections, also consider testing for ova and parasites if clinically indicated. Patients who eat paper may be exposed to mercury poisoning, so a serum mercury level is advisable.

Management: Prevention and behavior modification are key

Treatment for pica varies by patient and the specific behavior. Management approaches are primarily preventive, educational, and directed toward behavior modification.

Prevention. Residential facilities and primary care offices that care for people with developmental disabilities may screen for pica by means of prevalence surveys, direct observation, stool checks, review of medical history records, and interviews with caregivers.

Residential facilities can create a pica-safe environment by training staff in pica prevention, instituting regular on-site monitoring to ensure that no dangerous objects are available, and developing procedures to guide staff behavior, such as safe disposal of rubber gloves.²² Parents and caregivers of young children or children with developmental disabilities who don’t live in residential facilities should be aware of pica and monitor what their children are ingesting.

Behavior modification. Behavior-based approaches have proved effective for treating pica in developmentally disabled patients. Applied behavioral analysis “was found to have the most robust empirical support to treat this behavior.”³⁹ Patients found to have pica may be referred for further assessment to a behavior specialist or a psychologist with experience in treating the condition.^22,39

A review of 26 studies found that, in 25 studies, behavioral therapy reduced pica behavior by 80% or more.²³ Behavioral treatments included reinforcement procedures alone, response reduction procedures alone, and combined reinforcement and response reduction procedures. Reinforcement shapes behavior by controlling the consequences of the behavior using a combination of rewards and punishments.²³ Response reduction, or blocking, involves obstructing every attempt to eat inedible items.²²

Ask about pica behavior or unusual cravings in high-risk groups, such as pregnant women and children and adults with autism.Treatments that combined reinforcement and response reduction showed good efficacy.²³ An example of the combined approach would be to stop the patient from eating nonnutritive items while redirecting him to eat food instead.²²

Supplementation. Iron supplementation has decreased or even reversed pica in patients whose clinical symptoms and behavior were associated with iron deficiency.^35,40

Medications. Successful treatment with selective serotonin reuptake inhibitors (escitalopram), atypical neuroleptics (olanzapine), and attention-deficit/hyperactivity disorder medications (methylphenidate) has been reported in some patients, but case reports are few, and the evidence for the drugs’ efficacy is limited.^41-43

Be alert for pica. Primary care physicians need to be aware of pica and proactively seek information about cravings or behaviors suggesting the condition from patients in high-risk populations—pregnant women, children, immigrants and refugees, people with developmental disabilities—or their caregivers. Once pica is identified, clinicians should undertake appropriate laboratory investigation and behavior modification attempts.

CORRESPONDENCE
Ranit Mishori, MD, MHS, Department of Family Medicine, Georgetown University School of Medicine, Pre-Clinical Building, GB-01D, 3900 Reservoir Road, NW, Washington, DC 20007; [email protected]

For many years, articles about autism cited prevalence rates of approximately 7 in 10,000.¹ Over the past few years, however, there appears to have been an explosion in the rate at which autism is diagnosed: More recent estimates range from about 30 in 10,000² to one in 68.³ References to an autism epidemic appear to have originated in a 2002 California legislative report suggesting a 273% increase in autism from 1987 to 1998.⁴

Concerns about rising rates of autism, however, are not new. In 1943, Leo Kanner, MD, a psychiatrist and pioneer in the study of autism, published a paper titled, “Autistic disturbances of affective contact.”⁵ The result? “Almost overnight, the country seemed to be populated by a multitude of autistic children,” he later observed.⁶

To what should we attribute the current rise in reported autism rates? Even a casual review of the autism literature suggests a number of potential causes that may account for at least a portion of the recent increase.

We know more about the disorder

In his paper, Kanner described the “peculiarities” of 11 children whom he had cared for.⁵ While several had been diagnosed with mental retardation, childhood schizophrenia, or both, what stood out to Kanner was an “autistic aloneness” evident from the beginning of life. This was in contrast to childhood schizophrenia, in which a child experienced a departure from previously normal interrelations. His insight contributed to our understanding of what is now recognized as autism spectrum disorder (ASD).

Also in 1943, Hans Asperger, MD, was studying families with children exhibiting behaviors similar to those described by Kanner. The following year, Asperger published an article (in German) describing these children. Unfortunately, this paper—titled “Autistic psychopathy in childhood”⁷—was not translated into English until the early 1990s.⁸

Since Kanner and Asperger first called attention to the disorder, there have been numerous changes in societal and medical understanding of autism. Bruno Bettelheim, PhD, an Austrian-born child psychologist with a particular interest in emotionally disturbed children, theorized that poor maternal bonding and lack of maternal affection were responsible for autistic characteristics.⁹ Bernard Rimland, PhD, a psychologist and the father of an autistic child, argued for a biological basis of the disorder.¹⁰ Rimland’s theory of autism as a neurodevelopmental disorder with an unidentified organic etiology is most consistent with current medical opinion.

Definitions and diagnostic criteria have evolved

The first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) was published in 1952.¹¹ Although it was nearly a decade after Dr. Kanner clearly described autism as an entity separate from childhood schizophrenia, the word autism was used in this edition only once—to describe psychotic reactions associated with schizophrenia. Similarly, autism was referred to only in relation to childhood schizophrenia in the DSMII,¹² published in 1968.

DSM-III adds diagnostic criteria. In DSM-III (1980),13 specific diagnostic criteria for infantile autism and pervasive developmental disorder (PDD) appeared for the first time. Both diagnoses were clearly contrasted with a diagnosis of schizophrenia, and were used to identify children who exhibited a pervasive lack of responsiveness to others.

The DSM-III-Revised (R) (1987)14 added a classification scheme more consistent with the current standard. It included: 1) qualitative impairment in reciprocal social interaction; 2) qualitative impairment in communication and imaginative activity; and 3) restricted activity/interests.

PDD Not Otherwise Specified was also included, and served to identify those who had qualitative impairments in social interaction and communication skills but did not meet the full criteria for autism disorder or PDD.

The publication of DSM-IV (1994)¹⁵ brought another change: Specific criteria were outlined for the diagnoses of Asperger syndrome, Rett syndrome, and childhood disintegrative disorders; in the DSM-IV-Text Revision (TR) (2000),16 this structure remained relatively stable.

DSM-5 (2013) took another step, consolidating these various disorders into a single, unifying diagnosis of ASD.¹⁷ Significant controversy surrounded this change, with some viewing it as an oversimplification that does not accurately reflect important distinctions among divergent disorders18 and others arguing that it will result in unrecognized cases and exclusion of affected individuals.¹⁹ Notably, a recent study using concurrent DSM-IV and DSM-5 criteria to diagnose autism and PDD in more than 4000 children documented a high level of agreement between them.²⁰

Diagnostic tools have improved

As the number of children diagnosed with autism has increased, so have efforts to more accurately diagnose autism as a distinct disorder.

A blood test to identify differences in gene expression between children with and without ASD has shown initial promise—particularly in males. In the 1960s, guidelines for diagnosing autism focused primarily on Kanner’s original descriptive criteria. Even in the 1980s, after DSM-III criteria identified autism as a distinct disorder, children were being evaluated with generalized developmental screening tools focused on behaviors characteristic of a severe mental handicap, without differentiating between autistic and nonautistic children.

Early autism-specific observational and structured interview tools (eg, Childhood Autism Rating Scale,²¹ Autism Diagnostic Interview [ADI],²² and Autistic Diagnostic Observation Schedule)²³ emerged from a need for standardized diagnostic instruments that were comparable and reproducible.²⁴ But because these tools were highly specific and initially studied in research settings with high-risk populations, they lacked the sensitivity to identify children at risk in the general population, particularly those with milder symptoms.

As the diagnosis of autism became more standardized following publication of the DSM III-R in 1987, developmental specialists were able to construct increasingly sensitive evaluation tools. The ADI was revised²⁵ to facilitate earlier and more efficient diagnosis, allowing for assessment of children as young as 19 months of age. The Checklist for Autism in Toddlers (CHAT)²⁶ and subsequent modification (M-CHAT)²⁷ were among the earliest and most effective screening tools, appropriate for use in children as young as 16 months old.

Tools that followed the original CHAT (eg, the Autism Spectrum Screening Questionnaire [ASSQ]²⁸) were adapted to better identify high-functioning children with Asperger syndrome, as well as those with autism.

Another revision of the M-CHAT—the M-CHAT-R/F (Revised with Follow-up) was validated earlier this year. In a study involving 16,000 children, 95% of those who had positive tests were found to have some form of developmental delay and almost half (47%) received an ASD diagnosis.²⁹

Other diagnostic aids are being explored as a means of promoting earlier identification of ASD. For example, a blood test to identify differences in gene expression between children with and without ASD³⁰ has shown initial promise, particularly in males. This test is licensed by SynapDx (Lexington, Mass) and a clinical trial to evaluate it has begun.

Recent studies suggest that advanced paternal age can increase the risk of autism.Results of another study demonstrating normalization of brain activity in autistic children after they’ve undergone intensive treatment³¹ raise the possibility of using cortical activation as measured by electroencephalography as an early biomarker for autism.

Treatment options, advocacy affect rate of diagnosis

Improvements in diagnosis and targeted identification of potentially treatable symptoms³² led to the development of new treatment options. And greater use of day care and preschool programs prompted networking among parents, who touted the benefits of early evaluation, diagnosis, and treatment. Earlier screening, not surprisingly, led to an increase in the target population.

Local as well as national advocacy groups, led primarily by parents, have become powerful voices for improvements in services offered to children with autism. And research studies with varying degrees of sophistication continue to be published, further fueling the demand for school systems to provide supportive learning environments as required by the Individuals with Disabilities Education Act, initially enacted in 1975 and amended in 2004.³³

Federal and state funds in the form of Medicaid waivers are available to provide long-term care services in home and community settings, while private insurers typically pay for associated treatment modalities for those with an autism diagnosis, including physical, occupational, and speech therapy, among other services.

Finally, politicians and celebrities with personal connections to autism have joined the effort to increase awareness and improve the quality and availability of services—further assuring that autism is recognized as a legitimate, definable, and treatable disorder.

As an autism diagnosis has become more socially acceptable, it has at times replaced diagnoses of learning disability and mental retardation, a trend known as “diagnostic substitution.”³⁴ Indeed, having a child with an ASD diagnosis often makes it possible for parents to secure services that might otherwise be unavailable to them.

Is the incidence of autism linked to the environment?

Numerous environmental, nutritional, and pharmaceutical changes have been cited as reasons for what is perceived as an increasing incidence of autism in recent years. For example, some contend that greater use of food preservatives and greater exposure of young children to environmental toxins are contributing factors.³⁵

Thimerosal. Perhaps most notable is the assertion—since disproven—that thimerosal, a substance previously used in the manufacture of several childhood vaccines, was a leading cause of autism.^36,37 In fact, one study documented an increase in autism after thimerosal had been discontinued.³⁸ (For more on the thimerosal controversy, see “Autism: 5 misconceptions that can complicate care”.)

Autism comorbidity. As autism is frequently comorbid with other developmental disabilities, advances in medical technology that have led to a decline in neonatal death and overall mortality among the disabled may mean more survivors are subsequently diagnosed with an autism comorbidity.

Biological factors. Recent studies suggest that advanced paternal age can increase the risk of autism.³⁹ Twin studies suggest moderate genetic heritability, along with a substantial environmental contribution to the development of autism.⁴⁰ And new research suggests that maternal stressors during pregnancy—eg, trauma, illness, or substance abuse—may increase a child’s risk of developing autism, among other psychiatric disorders.⁴¹

The important role you play in the diagnosis of autism

It is clear that autism is more common than previously thought³ and that various factors are at work. Ensuring that children are promptly and properly evaluated begins when primary care physicians take parents’ concerns seriously and keep an eye out for common symptoms and characteristic developmental delays that may be evident even in the first year of life.

If symptoms are not severe enough to be detected in the child’s first several years, the next likely presentation will be when a parent gets a call from school suggesting that their child be tested for autism. While this delayed presentation suggests a higher level of functioning, a full evaluation, including the use of questionnaires such as the Social Responsiveness Scale⁴² and ASSQ,²⁸ and appropriate referrals are still vital.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted, and vague behavioral issues or concerns about intellectual impairment may become sufficiently troublesome to prompt a thorough evaluation. While behavioral therapy can be very helpful in such cases, medical therapy directed at co-occurring conditions (ie, a mood or anxiety disorder, including obsessive-compulsive disorder) may also provide significant benefit.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted. Whatever the age of the child, his or her parents should be counseled as to the general nature of autism, reassured of the availability of treatment options and given the appropriate referrals, and encouraged to learn more by availing themselves of resources (TABLE) and support groups. A consumer update from the US Food and Drug Administration titled “Beware of false or misleading claims for treating autism” (http://www.fda.gov/ForConsumers/ConsumerUpdates/ucm394757.htm) will be helpful for many parents, as well.

CORRESPONDENCE
Robert G. Zylstra, EdD, LCSW, Department of Family Medicine, The University of Tennessee College of Medicine, 110 East Third Street, Chattanooga, TN 37403; [email protected]

For many years, articles about autism cited prevalence rates of approximately 7 in 10,000.¹ Over the past few years, however, there appears to have been an explosion in the rate at which autism is diagnosed: More recent estimates range from about 30 in 10,000² to one in 68.³ References to an autism epidemic appear to have originated in a 2002 California legislative report suggesting a 273% increase in autism from 1987 to 1998.⁴

Concerns about rising rates of autism, however, are not new. In 1943, Leo Kanner, MD, a psychiatrist and pioneer in the study of autism, published a paper titled, “Autistic disturbances of affective contact.”⁵ The result? “Almost overnight, the country seemed to be populated by a multitude of autistic children,” he later observed.⁶

To what should we attribute the current rise in reported autism rates? Even a casual review of the autism literature suggests a number of potential causes that may account for at least a portion of the recent increase.

We know more about the disorder

In his paper, Kanner described the “peculiarities” of 11 children whom he had cared for.⁵ While several had been diagnosed with mental retardation, childhood schizophrenia, or both, what stood out to Kanner was an “autistic aloneness” evident from the beginning of life. This was in contrast to childhood schizophrenia, in which a child experienced a departure from previously normal interrelations. His insight contributed to our understanding of what is now recognized as autism spectrum disorder (ASD).

Also in 1943, Hans Asperger, MD, was studying families with children exhibiting behaviors similar to those described by Kanner. The following year, Asperger published an article (in German) describing these children. Unfortunately, this paper—titled “Autistic psychopathy in childhood”⁷—was not translated into English until the early 1990s.⁸

Since Kanner and Asperger first called attention to the disorder, there have been numerous changes in societal and medical understanding of autism. Bruno Bettelheim, PhD, an Austrian-born child psychologist with a particular interest in emotionally disturbed children, theorized that poor maternal bonding and lack of maternal affection were responsible for autistic characteristics.⁹ Bernard Rimland, PhD, a psychologist and the father of an autistic child, argued for a biological basis of the disorder.¹⁰ Rimland’s theory of autism as a neurodevelopmental disorder with an unidentified organic etiology is most consistent with current medical opinion.

Definitions and diagnostic criteria have evolved

The first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) was published in 1952.¹¹ Although it was nearly a decade after Dr. Kanner clearly described autism as an entity separate from childhood schizophrenia, the word autism was used in this edition only once—to describe psychotic reactions associated with schizophrenia. Similarly, autism was referred to only in relation to childhood schizophrenia in the DSMII,¹² published in 1968.

DSM-III adds diagnostic criteria. In DSM-III (1980),13 specific diagnostic criteria for infantile autism and pervasive developmental disorder (PDD) appeared for the first time. Both diagnoses were clearly contrasted with a diagnosis of schizophrenia, and were used to identify children who exhibited a pervasive lack of responsiveness to others.

The DSM-III-Revised (R) (1987)14 added a classification scheme more consistent with the current standard. It included: 1) qualitative impairment in reciprocal social interaction; 2) qualitative impairment in communication and imaginative activity; and 3) restricted activity/interests.

PDD Not Otherwise Specified was also included, and served to identify those who had qualitative impairments in social interaction and communication skills but did not meet the full criteria for autism disorder or PDD.

The publication of DSM-IV (1994)¹⁵ brought another change: Specific criteria were outlined for the diagnoses of Asperger syndrome, Rett syndrome, and childhood disintegrative disorders; in the DSM-IV-Text Revision (TR) (2000),16 this structure remained relatively stable.

DSM-5 (2013) took another step, consolidating these various disorders into a single, unifying diagnosis of ASD.¹⁷ Significant controversy surrounded this change, with some viewing it as an oversimplification that does not accurately reflect important distinctions among divergent disorders18 and others arguing that it will result in unrecognized cases and exclusion of affected individuals.¹⁹ Notably, a recent study using concurrent DSM-IV and DSM-5 criteria to diagnose autism and PDD in more than 4000 children documented a high level of agreement between them.²⁰

Diagnostic tools have improved

As the number of children diagnosed with autism has increased, so have efforts to more accurately diagnose autism as a distinct disorder.

A blood test to identify differences in gene expression between children with and without ASD has shown initial promise—particularly in males. In the 1960s, guidelines for diagnosing autism focused primarily on Kanner’s original descriptive criteria. Even in the 1980s, after DSM-III criteria identified autism as a distinct disorder, children were being evaluated with generalized developmental screening tools focused on behaviors characteristic of a severe mental handicap, without differentiating between autistic and nonautistic children.

Early autism-specific observational and structured interview tools (eg, Childhood Autism Rating Scale,²¹ Autism Diagnostic Interview [ADI],²² and Autistic Diagnostic Observation Schedule)²³ emerged from a need for standardized diagnostic instruments that were comparable and reproducible.²⁴ But because these tools were highly specific and initially studied in research settings with high-risk populations, they lacked the sensitivity to identify children at risk in the general population, particularly those with milder symptoms.

As the diagnosis of autism became more standardized following publication of the DSM III-R in 1987, developmental specialists were able to construct increasingly sensitive evaluation tools. The ADI was revised²⁵ to facilitate earlier and more efficient diagnosis, allowing for assessment of children as young as 19 months of age. The Checklist for Autism in Toddlers (CHAT)²⁶ and subsequent modification (M-CHAT)²⁷ were among the earliest and most effective screening tools, appropriate for use in children as young as 16 months old.

Tools that followed the original CHAT (eg, the Autism Spectrum Screening Questionnaire [ASSQ]²⁸) were adapted to better identify high-functioning children with Asperger syndrome, as well as those with autism.

Another revision of the M-CHAT—the M-CHAT-R/F (Revised with Follow-up) was validated earlier this year. In a study involving 16,000 children, 95% of those who had positive tests were found to have some form of developmental delay and almost half (47%) received an ASD diagnosis.²⁹

Other diagnostic aids are being explored as a means of promoting earlier identification of ASD. For example, a blood test to identify differences in gene expression between children with and without ASD³⁰ has shown initial promise, particularly in males. This test is licensed by SynapDx (Lexington, Mass) and a clinical trial to evaluate it has begun.

Recent studies suggest that advanced paternal age can increase the risk of autism.Results of another study demonstrating normalization of brain activity in autistic children after they’ve undergone intensive treatment³¹ raise the possibility of using cortical activation as measured by electroencephalography as an early biomarker for autism.

Treatment options, advocacy affect rate of diagnosis

Improvements in diagnosis and targeted identification of potentially treatable symptoms³² led to the development of new treatment options. And greater use of day care and preschool programs prompted networking among parents, who touted the benefits of early evaluation, diagnosis, and treatment. Earlier screening, not surprisingly, led to an increase in the target population.

Local as well as national advocacy groups, led primarily by parents, have become powerful voices for improvements in services offered to children with autism. And research studies with varying degrees of sophistication continue to be published, further fueling the demand for school systems to provide supportive learning environments as required by the Individuals with Disabilities Education Act, initially enacted in 1975 and amended in 2004.³³

Federal and state funds in the form of Medicaid waivers are available to provide long-term care services in home and community settings, while private insurers typically pay for associated treatment modalities for those with an autism diagnosis, including physical, occupational, and speech therapy, among other services.

Finally, politicians and celebrities with personal connections to autism have joined the effort to increase awareness and improve the quality and availability of services—further assuring that autism is recognized as a legitimate, definable, and treatable disorder.

As an autism diagnosis has become more socially acceptable, it has at times replaced diagnoses of learning disability and mental retardation, a trend known as “diagnostic substitution.”³⁴ Indeed, having a child with an ASD diagnosis often makes it possible for parents to secure services that might otherwise be unavailable to them.

Is the incidence of autism linked to the environment?

Numerous environmental, nutritional, and pharmaceutical changes have been cited as reasons for what is perceived as an increasing incidence of autism in recent years. For example, some contend that greater use of food preservatives and greater exposure of young children to environmental toxins are contributing factors.³⁵

Thimerosal. Perhaps most notable is the assertion—since disproven—that thimerosal, a substance previously used in the manufacture of several childhood vaccines, was a leading cause of autism.^36,37 In fact, one study documented an increase in autism after thimerosal had been discontinued.³⁸ (For more on the thimerosal controversy, see “Autism: 5 misconceptions that can complicate care”.)

Autism comorbidity. As autism is frequently comorbid with other developmental disabilities, advances in medical technology that have led to a decline in neonatal death and overall mortality among the disabled may mean more survivors are subsequently diagnosed with an autism comorbidity.

Biological factors. Recent studies suggest that advanced paternal age can increase the risk of autism.³⁹ Twin studies suggest moderate genetic heritability, along with a substantial environmental contribution to the development of autism.⁴⁰ And new research suggests that maternal stressors during pregnancy—eg, trauma, illness, or substance abuse—may increase a child’s risk of developing autism, among other psychiatric disorders.⁴¹

The important role you play in the diagnosis of autism

It is clear that autism is more common than previously thought³ and that various factors are at work. Ensuring that children are promptly and properly evaluated begins when primary care physicians take parents’ concerns seriously and keep an eye out for common symptoms and characteristic developmental delays that may be evident even in the first year of life.

If symptoms are not severe enough to be detected in the child’s first several years, the next likely presentation will be when a parent gets a call from school suggesting that their child be tested for autism. While this delayed presentation suggests a higher level of functioning, a full evaluation, including the use of questionnaires such as the Social Responsiveness Scale⁴² and ASSQ,²⁸ and appropriate referrals are still vital.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted, and vague behavioral issues or concerns about intellectual impairment may become sufficiently troublesome to prompt a thorough evaluation. While behavioral therapy can be very helpful in such cases, medical therapy directed at co-occurring conditions (ie, a mood or anxiety disorder, including obsessive-compulsive disorder) may also provide significant benefit.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted. Whatever the age of the child, his or her parents should be counseled as to the general nature of autism, reassured of the availability of treatment options and given the appropriate referrals, and encouraged to learn more by availing themselves of resources (TABLE) and support groups. A consumer update from the US Food and Drug Administration titled “Beware of false or misleading claims for treating autism” (http://www.fda.gov/ForConsumers/ConsumerUpdates/ucm394757.htm) will be helpful for many parents, as well.

CORRESPONDENCE
Robert G. Zylstra, EdD, LCSW, Department of Family Medicine, The University of Tennessee College of Medicine, 110 East Third Street, Chattanooga, TN 37403; [email protected]

For many years, articles about autism cited prevalence rates of approximately 7 in 10,000.¹ Over the past few years, however, there appears to have been an explosion in the rate at which autism is diagnosed: More recent estimates range from about 30 in 10,000² to one in 68.³ References to an autism epidemic appear to have originated in a 2002 California legislative report suggesting a 273% increase in autism from 1987 to 1998.⁴

Concerns about rising rates of autism, however, are not new. In 1943, Leo Kanner, MD, a psychiatrist and pioneer in the study of autism, published a paper titled, “Autistic disturbances of affective contact.”⁵ The result? “Almost overnight, the country seemed to be populated by a multitude of autistic children,” he later observed.⁶

To what should we attribute the current rise in reported autism rates? Even a casual review of the autism literature suggests a number of potential causes that may account for at least a portion of the recent increase.

We know more about the disorder

In his paper, Kanner described the “peculiarities” of 11 children whom he had cared for.⁵ While several had been diagnosed with mental retardation, childhood schizophrenia, or both, what stood out to Kanner was an “autistic aloneness” evident from the beginning of life. This was in contrast to childhood schizophrenia, in which a child experienced a departure from previously normal interrelations. His insight contributed to our understanding of what is now recognized as autism spectrum disorder (ASD).

Also in 1943, Hans Asperger, MD, was studying families with children exhibiting behaviors similar to those described by Kanner. The following year, Asperger published an article (in German) describing these children. Unfortunately, this paper—titled “Autistic psychopathy in childhood”⁷—was not translated into English until the early 1990s.⁸

Since Kanner and Asperger first called attention to the disorder, there have been numerous changes in societal and medical understanding of autism. Bruno Bettelheim, PhD, an Austrian-born child psychologist with a particular interest in emotionally disturbed children, theorized that poor maternal bonding and lack of maternal affection were responsible for autistic characteristics.⁹ Bernard Rimland, PhD, a psychologist and the father of an autistic child, argued for a biological basis of the disorder.¹⁰ Rimland’s theory of autism as a neurodevelopmental disorder with an unidentified organic etiology is most consistent with current medical opinion.

Definitions and diagnostic criteria have evolved

The first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) was published in 1952.¹¹ Although it was nearly a decade after Dr. Kanner clearly described autism as an entity separate from childhood schizophrenia, the word autism was used in this edition only once—to describe psychotic reactions associated with schizophrenia. Similarly, autism was referred to only in relation to childhood schizophrenia in the DSMII,¹² published in 1968.

DSM-III adds diagnostic criteria. In DSM-III (1980),13 specific diagnostic criteria for infantile autism and pervasive developmental disorder (PDD) appeared for the first time. Both diagnoses were clearly contrasted with a diagnosis of schizophrenia, and were used to identify children who exhibited a pervasive lack of responsiveness to others.

The DSM-III-Revised (R) (1987)14 added a classification scheme more consistent with the current standard. It included: 1) qualitative impairment in reciprocal social interaction; 2) qualitative impairment in communication and imaginative activity; and 3) restricted activity/interests.

PDD Not Otherwise Specified was also included, and served to identify those who had qualitative impairments in social interaction and communication skills but did not meet the full criteria for autism disorder or PDD.

The publication of DSM-IV (1994)¹⁵ brought another change: Specific criteria were outlined for the diagnoses of Asperger syndrome, Rett syndrome, and childhood disintegrative disorders; in the DSM-IV-Text Revision (TR) (2000),16 this structure remained relatively stable.

DSM-5 (2013) took another step, consolidating these various disorders into a single, unifying diagnosis of ASD.¹⁷ Significant controversy surrounded this change, with some viewing it as an oversimplification that does not accurately reflect important distinctions among divergent disorders18 and others arguing that it will result in unrecognized cases and exclusion of affected individuals.¹⁹ Notably, a recent study using concurrent DSM-IV and DSM-5 criteria to diagnose autism and PDD in more than 4000 children documented a high level of agreement between them.²⁰

Diagnostic tools have improved

As the number of children diagnosed with autism has increased, so have efforts to more accurately diagnose autism as a distinct disorder.

A blood test to identify differences in gene expression between children with and without ASD has shown initial promise—particularly in males. In the 1960s, guidelines for diagnosing autism focused primarily on Kanner’s original descriptive criteria. Even in the 1980s, after DSM-III criteria identified autism as a distinct disorder, children were being evaluated with generalized developmental screening tools focused on behaviors characteristic of a severe mental handicap, without differentiating between autistic and nonautistic children.

Early autism-specific observational and structured interview tools (eg, Childhood Autism Rating Scale,²¹ Autism Diagnostic Interview [ADI],²² and Autistic Diagnostic Observation Schedule)²³ emerged from a need for standardized diagnostic instruments that were comparable and reproducible.²⁴ But because these tools were highly specific and initially studied in research settings with high-risk populations, they lacked the sensitivity to identify children at risk in the general population, particularly those with milder symptoms.

As the diagnosis of autism became more standardized following publication of the DSM III-R in 1987, developmental specialists were able to construct increasingly sensitive evaluation tools. The ADI was revised²⁵ to facilitate earlier and more efficient diagnosis, allowing for assessment of children as young as 19 months of age. The Checklist for Autism in Toddlers (CHAT)²⁶ and subsequent modification (M-CHAT)²⁷ were among the earliest and most effective screening tools, appropriate for use in children as young as 16 months old.

Tools that followed the original CHAT (eg, the Autism Spectrum Screening Questionnaire [ASSQ]²⁸) were adapted to better identify high-functioning children with Asperger syndrome, as well as those with autism.

Another revision of the M-CHAT—the M-CHAT-R/F (Revised with Follow-up) was validated earlier this year. In a study involving 16,000 children, 95% of those who had positive tests were found to have some form of developmental delay and almost half (47%) received an ASD diagnosis.²⁹

Other diagnostic aids are being explored as a means of promoting earlier identification of ASD. For example, a blood test to identify differences in gene expression between children with and without ASD³⁰ has shown initial promise, particularly in males. This test is licensed by SynapDx (Lexington, Mass) and a clinical trial to evaluate it has begun.

Recent studies suggest that advanced paternal age can increase the risk of autism.Results of another study demonstrating normalization of brain activity in autistic children after they’ve undergone intensive treatment³¹ raise the possibility of using cortical activation as measured by electroencephalography as an early biomarker for autism.

Treatment options, advocacy affect rate of diagnosis

Improvements in diagnosis and targeted identification of potentially treatable symptoms³² led to the development of new treatment options. And greater use of day care and preschool programs prompted networking among parents, who touted the benefits of early evaluation, diagnosis, and treatment. Earlier screening, not surprisingly, led to an increase in the target population.

Local as well as national advocacy groups, led primarily by parents, have become powerful voices for improvements in services offered to children with autism. And research studies with varying degrees of sophistication continue to be published, further fueling the demand for school systems to provide supportive learning environments as required by the Individuals with Disabilities Education Act, initially enacted in 1975 and amended in 2004.³³

Federal and state funds in the form of Medicaid waivers are available to provide long-term care services in home and community settings, while private insurers typically pay for associated treatment modalities for those with an autism diagnosis, including physical, occupational, and speech therapy, among other services.

Finally, politicians and celebrities with personal connections to autism have joined the effort to increase awareness and improve the quality and availability of services—further assuring that autism is recognized as a legitimate, definable, and treatable disorder.

As an autism diagnosis has become more socially acceptable, it has at times replaced diagnoses of learning disability and mental retardation, a trend known as “diagnostic substitution.”³⁴ Indeed, having a child with an ASD diagnosis often makes it possible for parents to secure services that might otherwise be unavailable to them.

Is the incidence of autism linked to the environment?

Numerous environmental, nutritional, and pharmaceutical changes have been cited as reasons for what is perceived as an increasing incidence of autism in recent years. For example, some contend that greater use of food preservatives and greater exposure of young children to environmental toxins are contributing factors.³⁵

Thimerosal. Perhaps most notable is the assertion—since disproven—that thimerosal, a substance previously used in the manufacture of several childhood vaccines, was a leading cause of autism.^36,37 In fact, one study documented an increase in autism after thimerosal had been discontinued.³⁸ (For more on the thimerosal controversy, see “Autism: 5 misconceptions that can complicate care”.)

Autism comorbidity. As autism is frequently comorbid with other developmental disabilities, advances in medical technology that have led to a decline in neonatal death and overall mortality among the disabled may mean more survivors are subsequently diagnosed with an autism comorbidity.

Biological factors. Recent studies suggest that advanced paternal age can increase the risk of autism.³⁹ Twin studies suggest moderate genetic heritability, along with a substantial environmental contribution to the development of autism.⁴⁰ And new research suggests that maternal stressors during pregnancy—eg, trauma, illness, or substance abuse—may increase a child’s risk of developing autism, among other psychiatric disorders.⁴¹

The important role you play in the diagnosis of autism

It is clear that autism is more common than previously thought³ and that various factors are at work. Ensuring that children are promptly and properly evaluated begins when primary care physicians take parents’ concerns seriously and keep an eye out for common symptoms and characteristic developmental delays that may be evident even in the first year of life.

If symptoms are not severe enough to be detected in the child’s first several years, the next likely presentation will be when a parent gets a call from school suggesting that their child be tested for autism. While this delayed presentation suggests a higher level of functioning, a full evaluation, including the use of questionnaires such as the Social Responsiveness Scale⁴² and ASSQ,²⁸ and appropriate referrals are still vital.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted, and vague behavioral issues or concerns about intellectual impairment may become sufficiently troublesome to prompt a thorough evaluation. While behavioral therapy can be very helpful in such cases, medical therapy directed at co-occurring conditions (ie, a mood or anxiety disorder, including obsessive-compulsive disorder) may also provide significant benefit.

Puberty is another time when characteristics suggestive of autism that escaped earlier detection may be noted. Whatever the age of the child, his or her parents should be counseled as to the general nature of autism, reassured of the availability of treatment options and given the appropriate referrals, and encouraged to learn more by availing themselves of resources (TABLE) and support groups. A consumer update from the US Food and Drug Administration titled “Beware of false or misleading claims for treating autism” (http://www.fda.gov/ForConsumers/ConsumerUpdates/ucm394757.htm) will be helpful for many parents, as well.

CORRESPONDENCE
Robert G. Zylstra, EdD, LCSW, Department of Family Medicine, The University of Tennessee College of Medicine, 110 East Third Street, Chattanooga, TN 37403; [email protected]

Autism spectrum disorder (ASD) affects approximately one in 68 children in the United States, according to the Centers for Disease Control and Prevention (CDC).¹ Growing public awareness of autism means that family physicians are increasingly likely to hear from anxious new (and expectant) parents. Unfortunately, misinformation about autism continues to be perpetuated, through word of mouth, the Internet, and misinformed advocacy groups. This article addresses 5 of the most common misconceptions, and can help you set the record straight and respond appropriately to parental concerns.

Misconception 1: Autism is a single condition

While autistic disorder was previously considered one of 5 pervasive developmental disorders, in 2013 the Diagnostic and Statistical Manual of Mental Disorder, 5th edition (DSM-5) redefined it. (To learn more about how shifts in our understanding of autism were reflected in each new edition of the DSM, see “Autism: Why the rise in rates?”)

ASD is now an umbrella term that encompasses autism, Asperger syndrome, pervasive developmental disorder not otherwise specified, childhood disintegrative disorder, and Rett syndrome.² The new term is meant to highlight the continuum of symptoms and frequent variability of presentation among those affected, ranging from mild to more severe impairment. Anyone who was classified under DSM-IV criteria, of course, should continue to have an autism/ASD diagnosis.

As with previous definitions, ASD is characterized by communication deficits (eg, inappropriate responses in conversation, misinterpreted nonverbal interactions, and significant challenges in age-appropriate bonding/friendship development). While previous definitions were focused on identifying school-age deficits, the update requires early childhood symptoms—regardless of the age of formal diagnosis.²

Misconception 2: Only symptomatic children should be screened for ASD

Although the decision to screen all children remains a controversial one, at least one medical society—the American Academy of Pediatrics (AAP)—calls for universal screening.³ The American Academy of Family Physicians does not have or endorse a formal guideline about screening for ASD. The US Preventive Services Task Force has a guideline, but it is in the process of being revised.⁴

Given the advances in early childhood interventions, there is little doubt that early identification of those at risk for developmental delay is beneficial. But opponents of universal screening cite concerns about unnecessary testing, anxiety, and overdiagnosis due to false positives associated with traditional screening methods.

The AAP calls for screening and surveillance. Since 2006, the AAP has recommended surveillance at all well-child visits, combined with screening for developmental delays at 9, 18, and 24 or 30 months of age, using a standardized screening tool.^3,5 A parent-completed tool (eg, the Modified Checklist for Autism in Toddlers [M-CHAT] or, most recently, the M-CHAT Revised with Follow-up⁶; Parents’ Evaluation of Developmental Status; or Ages and Stages Questionnaire, 3rd ed) should be used rather than a directly administered tool.^3,5 An algorithm detailing the AAP’s approach is available at http://www.cdc.gov/ncbddd/actearly/autism/case-modules/pdf/diagnosis/AAP%20Screening%20Guidelines.pdf.

Avoid a "wait and see" approach. Parents of children with ASD often broach the subject by the baby's first birthday.Physicians should also be prepared to evaluate any child whose parents raise concerns about his or her development during a routine visit. A “wait and see” approach is strongly discouraged. Parents of children with ASD often broach the subject by the baby’s first birthday.⁵ Common concerns include the child’s inability to babble, point or gesture meaningfully, or respond to his or her name; poor eye contact; failure to play with toys; and/or loss of (or failure to develop) language or social skills.⁵

An ASD mnemonic. The CDC, in collaboration with the AAP, the American Academy of Neurology, and the Child Neurology Society, has released a simplified guideline with the mnemonic ALARM to summarize recommendations for developmental screening, surveillance, diagnosis, and management of ASD.⁵ ALARM stands for:
• Autism is prevalent
• Listen to parents
• Act early
• Refer
• Monitor.

Misconception 3: Since ASD can't be cured, early intervention offers no benefit

While there is no cure for ASD and it is not considered reversible, there is an array of potential ASD therapies and proven benefits of early intervention. Therapies range from diet to medication and behavioral skills development, but only a few have ample evidence of efficacy (TABLE).^7-17

Patients with ASD are less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.Randomized controlled trials of early developmental and behavioral therapy have shown some promise in decreasing symptoms associated with ASD and improving parent-child communication and social engagement.^18-21 A systematic review found that young children with ASD can improve cognitive performance, language skills, and adaptive behavioral skills through behavioral interventions or more comprehensive approaches using developmental and behavioral frameworks.²² In addition to increasing the likelihood of overall school success, early intervention programs that improve communication and social skills can have a significant impact on the individual’s eventual quality of life (QOL), employability, and independence.^23-25

TABLE
Therapeutic options for ASD: What works?^7-17

^{ASD, autism spectrum disorder; IV, intravenous; SOR, strength of recommendation.
*SOR: A, Good-quality patient-oriented evidence; B, inconsistent or limited-quality patient-oriented evidence; C, consensus, usual practice, opinion, disease-oriented evidence, case series.}

Misconception 4: Individuals with ASD are more intellectually disabled and can't function independently

As many as 96% of children with ASD have a coexisting developmental disability, such as generalized developmental delay (found in 80% of those with ASD), learning disabilities (affecting 60%), or attention-deficit/hyperactivity disorder (in 42%).²⁶ Although many parents—and some physicians—assume that children with ASD are intellectually disabled, in fact, less than one in 5 (19%) has an intellectual disability.²⁶

Individuals with ASD do, however, often have difficulty living independently. In one study of post high school living arrangements, those with ASD were less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.²³

Another analysis found that only about half (53.4%) of young adults with ASD had ever worked for pay outside the home since leaving high school—the lowest rate among disability groups.²⁴ Of those who had worked outside the home, young adults with ASD earned an average of $8.10 per hour, significantly lower than the comparison groups. Not surprisingly, young adults with ASD who had better conversational and/or functional skills had a higher likelihood of ever having had worked for pay outside the home.²⁴

Social participation also is considered an indicator of overall QOL and independent function. Young adults with ASD were found to be significantly more likely than those with other types of developmental delays to be socially isolated—never seeing friends, getting calls from friends, or being invited to activities. Lower communication and functional skills, as well as living with a parent, were predictors of less social participation in young adults with ASD.²⁵

Misconception 5: Thimerosal vaccines cause ASD

The controversy and concern about a correlation between mercury and ASD began in the late 1990s, when a published study appeared to link thimerosal-containing vaccines to the increasing incidence of autism.²⁷ This notion appeared to be further strengthened by a 2002 study,²⁸ done by the same researchers and reaching similar conclusions. Since then, the correlation has been disproven by a number of studies and further review of the initial studies revealed them to be flawed.²⁹

Despite the lack of evidence to support any long-term effects from the minimal exposure to mercury in the preservative, in July 1999, US Public Health Service agencies, the AAP, and vaccine manufacturers agreed that thimerosal should be reduced or eliminated in vaccines as a precautionary measure.³⁰ With the exception of a limited number of multidose influenza vaccines, childhood vaccines are now thimerosal-free.

Fact: Early referral is key

When and where to refer parents for additional evaluation of a child with developmental delays depends on the resources in your community. Typically, a team of providers participates in the evaluation and management of a child with ASD, often including a developmental pediatrician, psychiatrist and other mental health professional, neurologist, speech pathologist, audiologist, physical therapist, and special education teacher.

Although every community may not have easy access to a pediatric subspecialist referral center, all 50 states and US territories are required by law to provide access to early intervention programs under Part C of the 2004 Individuals with Disabilities Education Act.31 (For a list of resources, see “Autism spectrum disorder: Where to learn more”.)

Help families prepare

Despite a lack of evidence connecting thimerosal to autism, the agent has been removed from most childhood vaccines as a precautionary measure.Advise families whom you refer for evaluation that they are unlikely to have a definitive answer after the first visit. Explain that the evaluation is quite thorough and generally takes several visits. Typically, a child suspected of developmental delay will undergo a comprehensive evaluation, including history and physical exam, blood work (including lead testing and, in some locations, genetic testing), hearing and vision screening, speech and language evaluation, and sensorimotor and cognitive evaluation. Additional information may be requested from daycare providers, preschool teachers, or others who spend significant amounts of time with the child.

Once a diagnosis is made, the team will work with the parents to develop an individualized care plan for the child, which often includes a mix of cognitive, physical, and speech development services in addition to nutrition and support services. The primary care physician, of course, will continue to oversee, monitor, and coordinate care.

CORRESPONDENCE
Margot Savoy, MD, MPH, FAAF P, CPE, FA BC, Christiana Care Health System, Department of Family & Community Medicine, 1401 Foulk Road, Suite 100B, Wilmington, DE 19803; [email protected]

Autism spectrum disorder (ASD) affects approximately one in 68 children in the United States, according to the Centers for Disease Control and Prevention (CDC).¹ Growing public awareness of autism means that family physicians are increasingly likely to hear from anxious new (and expectant) parents. Unfortunately, misinformation about autism continues to be perpetuated, through word of mouth, the Internet, and misinformed advocacy groups. This article addresses 5 of the most common misconceptions, and can help you set the record straight and respond appropriately to parental concerns.

Misconception 1: Autism is a single condition

While autistic disorder was previously considered one of 5 pervasive developmental disorders, in 2013 the Diagnostic and Statistical Manual of Mental Disorder, 5th edition (DSM-5) redefined it. (To learn more about how shifts in our understanding of autism were reflected in each new edition of the DSM, see “Autism: Why the rise in rates?”)

ASD is now an umbrella term that encompasses autism, Asperger syndrome, pervasive developmental disorder not otherwise specified, childhood disintegrative disorder, and Rett syndrome.² The new term is meant to highlight the continuum of symptoms and frequent variability of presentation among those affected, ranging from mild to more severe impairment. Anyone who was classified under DSM-IV criteria, of course, should continue to have an autism/ASD diagnosis.

As with previous definitions, ASD is characterized by communication deficits (eg, inappropriate responses in conversation, misinterpreted nonverbal interactions, and significant challenges in age-appropriate bonding/friendship development). While previous definitions were focused on identifying school-age deficits, the update requires early childhood symptoms—regardless of the age of formal diagnosis.²

Misconception 2: Only symptomatic children should be screened for ASD

Although the decision to screen all children remains a controversial one, at least one medical society—the American Academy of Pediatrics (AAP)—calls for universal screening.³ The American Academy of Family Physicians does not have or endorse a formal guideline about screening for ASD. The US Preventive Services Task Force has a guideline, but it is in the process of being revised.⁴

Given the advances in early childhood interventions, there is little doubt that early identification of those at risk for developmental delay is beneficial. But opponents of universal screening cite concerns about unnecessary testing, anxiety, and overdiagnosis due to false positives associated with traditional screening methods.

The AAP calls for screening and surveillance. Since 2006, the AAP has recommended surveillance at all well-child visits, combined with screening for developmental delays at 9, 18, and 24 or 30 months of age, using a standardized screening tool.^3,5 A parent-completed tool (eg, the Modified Checklist for Autism in Toddlers [M-CHAT] or, most recently, the M-CHAT Revised with Follow-up⁶; Parents’ Evaluation of Developmental Status; or Ages and Stages Questionnaire, 3rd ed) should be used rather than a directly administered tool.^3,5 An algorithm detailing the AAP’s approach is available at http://www.cdc.gov/ncbddd/actearly/autism/case-modules/pdf/diagnosis/AAP%20Screening%20Guidelines.pdf.

Avoid a "wait and see" approach. Parents of children with ASD often broach the subject by the baby's first birthday.Physicians should also be prepared to evaluate any child whose parents raise concerns about his or her development during a routine visit. A “wait and see” approach is strongly discouraged. Parents of children with ASD often broach the subject by the baby’s first birthday.⁵ Common concerns include the child’s inability to babble, point or gesture meaningfully, or respond to his or her name; poor eye contact; failure to play with toys; and/or loss of (or failure to develop) language or social skills.⁵

An ASD mnemonic. The CDC, in collaboration with the AAP, the American Academy of Neurology, and the Child Neurology Society, has released a simplified guideline with the mnemonic ALARM to summarize recommendations for developmental screening, surveillance, diagnosis, and management of ASD.⁵ ALARM stands for:
• Autism is prevalent
• Listen to parents
• Act early
• Refer
• Monitor.

Misconception 3: Since ASD can't be cured, early intervention offers no benefit

While there is no cure for ASD and it is not considered reversible, there is an array of potential ASD therapies and proven benefits of early intervention. Therapies range from diet to medication and behavioral skills development, but only a few have ample evidence of efficacy (TABLE).^7-17

Patients with ASD are less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.Randomized controlled trials of early developmental and behavioral therapy have shown some promise in decreasing symptoms associated with ASD and improving parent-child communication and social engagement.^18-21 A systematic review found that young children with ASD can improve cognitive performance, language skills, and adaptive behavioral skills through behavioral interventions or more comprehensive approaches using developmental and behavioral frameworks.²² In addition to increasing the likelihood of overall school success, early intervention programs that improve communication and social skills can have a significant impact on the individual’s eventual quality of life (QOL), employability, and independence.^23-25

TABLE
Therapeutic options for ASD: What works?^7-17

^{ASD, autism spectrum disorder; IV, intravenous; SOR, strength of recommendation.
*SOR: A, Good-quality patient-oriented evidence; B, inconsistent or limited-quality patient-oriented evidence; C, consensus, usual practice, opinion, disease-oriented evidence, case series.}

Misconception 4: Individuals with ASD are more intellectually disabled and can't function independently

As many as 96% of children with ASD have a coexisting developmental disability, such as generalized developmental delay (found in 80% of those with ASD), learning disabilities (affecting 60%), or attention-deficit/hyperactivity disorder (in 42%).²⁶ Although many parents—and some physicians—assume that children with ASD are intellectually disabled, in fact, less than one in 5 (19%) has an intellectual disability.²⁶

Individuals with ASD do, however, often have difficulty living independently. In one study of post high school living arrangements, those with ASD were less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.²³

Another analysis found that only about half (53.4%) of young adults with ASD had ever worked for pay outside the home since leaving high school—the lowest rate among disability groups.²⁴ Of those who had worked outside the home, young adults with ASD earned an average of $8.10 per hour, significantly lower than the comparison groups. Not surprisingly, young adults with ASD who had better conversational and/or functional skills had a higher likelihood of ever having had worked for pay outside the home.²⁴

Social participation also is considered an indicator of overall QOL and independent function. Young adults with ASD were found to be significantly more likely than those with other types of developmental delays to be socially isolated—never seeing friends, getting calls from friends, or being invited to activities. Lower communication and functional skills, as well as living with a parent, were predictors of less social participation in young adults with ASD.²⁵

Misconception 5: Thimerosal vaccines cause ASD

The controversy and concern about a correlation between mercury and ASD began in the late 1990s, when a published study appeared to link thimerosal-containing vaccines to the increasing incidence of autism.²⁷ This notion appeared to be further strengthened by a 2002 study,²⁸ done by the same researchers and reaching similar conclusions. Since then, the correlation has been disproven by a number of studies and further review of the initial studies revealed them to be flawed.²⁹

Despite the lack of evidence to support any long-term effects from the minimal exposure to mercury in the preservative, in July 1999, US Public Health Service agencies, the AAP, and vaccine manufacturers agreed that thimerosal should be reduced or eliminated in vaccines as a precautionary measure.³⁰ With the exception of a limited number of multidose influenza vaccines, childhood vaccines are now thimerosal-free.

Fact: Early referral is key

When and where to refer parents for additional evaluation of a child with developmental delays depends on the resources in your community. Typically, a team of providers participates in the evaluation and management of a child with ASD, often including a developmental pediatrician, psychiatrist and other mental health professional, neurologist, speech pathologist, audiologist, physical therapist, and special education teacher.

Although every community may not have easy access to a pediatric subspecialist referral center, all 50 states and US territories are required by law to provide access to early intervention programs under Part C of the 2004 Individuals with Disabilities Education Act.31 (For a list of resources, see “Autism spectrum disorder: Where to learn more”.)

Help families prepare

Despite a lack of evidence connecting thimerosal to autism, the agent has been removed from most childhood vaccines as a precautionary measure.Advise families whom you refer for evaluation that they are unlikely to have a definitive answer after the first visit. Explain that the evaluation is quite thorough and generally takes several visits. Typically, a child suspected of developmental delay will undergo a comprehensive evaluation, including history and physical exam, blood work (including lead testing and, in some locations, genetic testing), hearing and vision screening, speech and language evaluation, and sensorimotor and cognitive evaluation. Additional information may be requested from daycare providers, preschool teachers, or others who spend significant amounts of time with the child.

Once a diagnosis is made, the team will work with the parents to develop an individualized care plan for the child, which often includes a mix of cognitive, physical, and speech development services in addition to nutrition and support services. The primary care physician, of course, will continue to oversee, monitor, and coordinate care.

CORRESPONDENCE
Margot Savoy, MD, MPH, FAAF P, CPE, FA BC, Christiana Care Health System, Department of Family & Community Medicine, 1401 Foulk Road, Suite 100B, Wilmington, DE 19803; [email protected]

Autism spectrum disorder (ASD) affects approximately one in 68 children in the United States, according to the Centers for Disease Control and Prevention (CDC).¹ Growing public awareness of autism means that family physicians are increasingly likely to hear from anxious new (and expectant) parents. Unfortunately, misinformation about autism continues to be perpetuated, through word of mouth, the Internet, and misinformed advocacy groups. This article addresses 5 of the most common misconceptions, and can help you set the record straight and respond appropriately to parental concerns.

Misconception 1: Autism is a single condition

While autistic disorder was previously considered one of 5 pervasive developmental disorders, in 2013 the Diagnostic and Statistical Manual of Mental Disorder, 5th edition (DSM-5) redefined it. (To learn more about how shifts in our understanding of autism were reflected in each new edition of the DSM, see “Autism: Why the rise in rates?”)

ASD is now an umbrella term that encompasses autism, Asperger syndrome, pervasive developmental disorder not otherwise specified, childhood disintegrative disorder, and Rett syndrome.² The new term is meant to highlight the continuum of symptoms and frequent variability of presentation among those affected, ranging from mild to more severe impairment. Anyone who was classified under DSM-IV criteria, of course, should continue to have an autism/ASD diagnosis.

As with previous definitions, ASD is characterized by communication deficits (eg, inappropriate responses in conversation, misinterpreted nonverbal interactions, and significant challenges in age-appropriate bonding/friendship development). While previous definitions were focused on identifying school-age deficits, the update requires early childhood symptoms—regardless of the age of formal diagnosis.²

Misconception 2: Only symptomatic children should be screened for ASD

Although the decision to screen all children remains a controversial one, at least one medical society—the American Academy of Pediatrics (AAP)—calls for universal screening.³ The American Academy of Family Physicians does not have or endorse a formal guideline about screening for ASD. The US Preventive Services Task Force has a guideline, but it is in the process of being revised.⁴

Given the advances in early childhood interventions, there is little doubt that early identification of those at risk for developmental delay is beneficial. But opponents of universal screening cite concerns about unnecessary testing, anxiety, and overdiagnosis due to false positives associated with traditional screening methods.

The AAP calls for screening and surveillance. Since 2006, the AAP has recommended surveillance at all well-child visits, combined with screening for developmental delays at 9, 18, and 24 or 30 months of age, using a standardized screening tool.^3,5 A parent-completed tool (eg, the Modified Checklist for Autism in Toddlers [M-CHAT] or, most recently, the M-CHAT Revised with Follow-up⁶; Parents’ Evaluation of Developmental Status; or Ages and Stages Questionnaire, 3rd ed) should be used rather than a directly administered tool.^3,5 An algorithm detailing the AAP’s approach is available at http://www.cdc.gov/ncbddd/actearly/autism/case-modules/pdf/diagnosis/AAP%20Screening%20Guidelines.pdf.

Avoid a "wait and see" approach. Parents of children with ASD often broach the subject by the baby's first birthday.Physicians should also be prepared to evaluate any child whose parents raise concerns about his or her development during a routine visit. A “wait and see” approach is strongly discouraged. Parents of children with ASD often broach the subject by the baby’s first birthday.⁵ Common concerns include the child’s inability to babble, point or gesture meaningfully, or respond to his or her name; poor eye contact; failure to play with toys; and/or loss of (or failure to develop) language or social skills.⁵

An ASD mnemonic. The CDC, in collaboration with the AAP, the American Academy of Neurology, and the Child Neurology Society, has released a simplified guideline with the mnemonic ALARM to summarize recommendations for developmental screening, surveillance, diagnosis, and management of ASD.⁵ ALARM stands for:
• Autism is prevalent
• Listen to parents
• Act early
• Refer
• Monitor.

Misconception 3: Since ASD can't be cured, early intervention offers no benefit

While there is no cure for ASD and it is not considered reversible, there is an array of potential ASD therapies and proven benefits of early intervention. Therapies range from diet to medication and behavioral skills development, but only a few have ample evidence of efficacy (TABLE).^7-17

Patients with ASD are less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.Randomized controlled trials of early developmental and behavioral therapy have shown some promise in decreasing symptoms associated with ASD and improving parent-child communication and social engagement.^18-21 A systematic review found that young children with ASD can improve cognitive performance, language skills, and adaptive behavioral skills through behavioral interventions or more comprehensive approaches using developmental and behavioral frameworks.²² In addition to increasing the likelihood of overall school success, early intervention programs that improve communication and social skills can have a significant impact on the individual’s eventual quality of life (QOL), employability, and independence.^23-25

TABLE
Therapeutic options for ASD: What works?^7-17

^{ASD, autism spectrum disorder; IV, intravenous; SOR, strength of recommendation.
*SOR: A, Good-quality patient-oriented evidence; B, inconsistent or limited-quality patient-oriented evidence; C, consensus, usual practice, opinion, disease-oriented evidence, case series.}

Misconception 4: Individuals with ASD are more intellectually disabled and can't function independently

As many as 96% of children with ASD have a coexisting developmental disability, such as generalized developmental delay (found in 80% of those with ASD), learning disabilities (affecting 60%), or attention-deficit/hyperactivity disorder (in 42%).²⁶ Although many parents—and some physicians—assume that children with ASD are intellectually disabled, in fact, less than one in 5 (19%) has an intellectual disability.²⁶

Individuals with ASD do, however, often have difficulty living independently. In one study of post high school living arrangements, those with ASD were less likely to have ever lived on their own than those with learning disabilities, intellectual disabilities, or emotional disturbances.²³

Another analysis found that only about half (53.4%) of young adults with ASD had ever worked for pay outside the home since leaving high school—the lowest rate among disability groups.²⁴ Of those who had worked outside the home, young adults with ASD earned an average of $8.10 per hour, significantly lower than the comparison groups. Not surprisingly, young adults with ASD who had better conversational and/or functional skills had a higher likelihood of ever having had worked for pay outside the home.²⁴

Social participation also is considered an indicator of overall QOL and independent function. Young adults with ASD were found to be significantly more likely than those with other types of developmental delays to be socially isolated—never seeing friends, getting calls from friends, or being invited to activities. Lower communication and functional skills, as well as living with a parent, were predictors of less social participation in young adults with ASD.²⁵

Misconception 5: Thimerosal vaccines cause ASD

The controversy and concern about a correlation between mercury and ASD began in the late 1990s, when a published study appeared to link thimerosal-containing vaccines to the increasing incidence of autism.²⁷ This notion appeared to be further strengthened by a 2002 study,²⁸ done by the same researchers and reaching similar conclusions. Since then, the correlation has been disproven by a number of studies and further review of the initial studies revealed them to be flawed.²⁹

Despite the lack of evidence to support any long-term effects from the minimal exposure to mercury in the preservative, in July 1999, US Public Health Service agencies, the AAP, and vaccine manufacturers agreed that thimerosal should be reduced or eliminated in vaccines as a precautionary measure.³⁰ With the exception of a limited number of multidose influenza vaccines, childhood vaccines are now thimerosal-free.

Fact: Early referral is key

When and where to refer parents for additional evaluation of a child with developmental delays depends on the resources in your community. Typically, a team of providers participates in the evaluation and management of a child with ASD, often including a developmental pediatrician, psychiatrist and other mental health professional, neurologist, speech pathologist, audiologist, physical therapist, and special education teacher.

Although every community may not have easy access to a pediatric subspecialist referral center, all 50 states and US territories are required by law to provide access to early intervention programs under Part C of the 2004 Individuals with Disabilities Education Act.31 (For a list of resources, see “Autism spectrum disorder: Where to learn more”.)

Help families prepare

Despite a lack of evidence connecting thimerosal to autism, the agent has been removed from most childhood vaccines as a precautionary measure.Advise families whom you refer for evaluation that they are unlikely to have a definitive answer after the first visit. Explain that the evaluation is quite thorough and generally takes several visits. Typically, a child suspected of developmental delay will undergo a comprehensive evaluation, including history and physical exam, blood work (including lead testing and, in some locations, genetic testing), hearing and vision screening, speech and language evaluation, and sensorimotor and cognitive evaluation. Additional information may be requested from daycare providers, preschool teachers, or others who spend significant amounts of time with the child.

Once a diagnosis is made, the team will work with the parents to develop an individualized care plan for the child, which often includes a mix of cognitive, physical, and speech development services in addition to nutrition and support services. The primary care physician, of course, will continue to oversee, monitor, and coordinate care.

CORRESPONDENCE
Margot Savoy, MD, MPH, FAAF P, CPE, FA BC, Christiana Care Health System, Department of Family & Community Medicine, 1401 Foulk Road, Suite 100B, Wilmington, DE 19803; [email protected]

Pressure to reform our health care system is at an all-time high, driven by relentlessly rising costs and fragmentation of care. These persistent problems have led to lower quality care and limited access to care for a large proportion of the US population—issues that accountable care organizations (ACOs), as well as other value-based models, are designed to address.

While the terms used to describe the means by which health care systems attempt to do more to meet the needs of those they serve may vary, the importance of reorganizing care delivery to better integrate services is gaining traction nationwide. As we move to new models, primary care takes center stage.

ACOs (or ACO-type arrangements) anchored by primary care networks can help meet the goals of health care reform by responding to changes in reimbursement, reducing fragmented care, and focusing on improving the quality of care for defined patient populations. In addition, these delivery models can take advantage of new health information technology (IT) and the move toward patient-centered medical homes (PCMHs).

In the pages that follow, we examine opportunities for new care delivery models to slow rising costs and improve population health in family medicine. The introduction of these models has important implications for patients, physicians, and provider organizations, and our aim is to ensure that family physicians are prepared to take these vital steps toward achieving health reform goals.

Shifting from volume-based to value-based reimbursement

According to the Congressional Budget Office, ACOs are expected to save $5 billion during their first 8 years of existence.¹ After one year of ACO activity, the Centers for Medicare and Medicaid Services (CMS) reported savings of $30 million.² The expected savings will be driven by the increased provider accountability associated with ACOs.

Various means of increasing provider accountability through changes in reimbursement strategies have been proposed; several are new, while others are improvements on—or variations of—methods that have been tried before. The most common approaches—shared savings, shared savings plus penalty, capitation, episodic payment, prospective payment, pay-for-performance, and hospital-physician bundling—are detailed in TABLE 1.^3-7 Broad implementation of any of these reimbursement mechanisms within a new model of care would represent a shift away from the traditional volume-based provider payment model to a value-based system—a key step in slowing the rise in health care costs.

TABLE 1
Reimbursement strategies designed to promote physician accountability^3-7

^{ACO, accountable care organization; CMS, Centers for Medicare and Medicaid Services; DRG, diagnosis-related group; FFS, fee-for-service; FPs, family physicians; IT, information technology; PCMH, patient-centered medical home.}

Although it is too soon to be certain of the effect such changes will have on the earnings of family physicians, it is reasonable to think that new payment strategies—and a larger role for primary care providers—will improve their financial standing.

Moving toward population health management

New ACO-type models also make it easier to improve health care for specific populations, using strategies designed to organize, provide, and manage care for defined groups. In addition to controlling the cost of caring for specific groups, well-designed and implemented population health management strategies can increase continuity of care by ensuring oversight of patients across the spectrum of health care settings.

Five broad categories of population health management are most prominent: lifestyle management and demand management (both for relatively healthy people), disease management (for those with chronic conditions), catastrophic care management (for patients with rare or catastrophic illness or injury), and disability management (for groups of employees).⁸TABLE 2⁸ describes the population targeted and activities associated with each. It is important to remember, however, that no single strategy is mutually exclusive for a particular group.

Comprehensive Primary Care (CPC) initiative. In a collaboration made possible by the Affordable Care Act, CMS initiated the CPC initiative in 2012.⁹ A 4-year project designed to test and further identify the benefits of population health management and strengthen coordination of care for Medicare patients within primary care settings, the CPC initiative involves 497 sites and 2347 providers caring for approximately 315,000 Medicare beneficiaries.⁹ More information is available at http://innovation.cms.gov/initiatives/comprehensive-primary-care-initiative/.

TABLE 2
Population health management strategies⁸

^{QOL, quality of life.}

Building infrastructure and leveraging IT

ACOs and ACO-type models will take a variety of forms, depending in part on geographic need and local demographics. Yet, all share a common need for a strong infrastructure to support improved transitions, integration, and coordination of care. Incorporation of a strong health IT system is critical so that data regarding the process and cost of care, as well as outcomes, can be collected and put to optimal use.^10-13

Across health care settings, health IT innovations are being successfully implemented in efforts to enhance physician decision support, improve patient safety, increase guideline adherence, and improve chronic disease treatment.^11,13 In primary care, for example, an IT infrastructure can create disease registries so that data on patients with specific conditions can be tracked and acted upon—eg, a diabetes registry could be used to identify and contact patients who have an hemoglobin A1c >9 and have not been seen in 9 months or more.

Similarly, an IT system with the ability to identify patients at high risk for disease decompensation, hospitalization, and/or increased morbidity and mortality linked to progression of a chronic disease is needed. Identifying medical conditions associated with higher costs would make it possible to focus care coordination and chronic care management efforts on this targeted population.

As health IT continues to evolve, additional means of interacting with patients and improving patient care will be developed. Physicians and organizations that are ready to take advantage of these advances in technology will be well positioned to address the goals of health reform. (See “Health care reform: Recommendations for family physicians” below.^3-7,14,15)

How the patient-centered medical home fits into the picture

The implementation of ACOs and other new models of care has promising implications for the establishment of PCMHs. Consistent with the goals of health reform, the PCMH movement focuses on a coordinated teamwork approach, anchored within a general practice or family medicine setting.

An evaluation of the PCMH National Demonstration Project funded by the American Academy of Family Physicians found that the adoption of more components of the PCMH at the practice level was associated with improvement in patient outcomes, as measured by the Ambulatory Care Quality Alliance starter set16 (a compilation of clinical performance measures developed by a broad coalition of providers, payers, consumers, and government agencies).

A recent look at the Southeastern Pennsylvania Chronic Care Initiative¹⁷ found less promising results. “A multipayer medical home pilot, in which participating practices adopted new structural capabilities and received NCQA [National Committee for Quality Assurance] certification, was associated with limited improvements in quality and was not associated with reductions in utilization of hospital, emergency department, or ambulatory care services or total costs over 3 years,” Friedberg et al¹⁷ concluded. The authors did note, however, that NCQA recognition was what the practices involved in this initiative were rewarded for—not PCMH activity.

Coordinated delivery models that integrate primary care and mental health services have been shown to be cost-effective.It is also important to keep in mind that PCMH activity has been shown to improve care and reduce costs—not NCQA recognition in and of itself. In fact, a large body of evidence clearly demonstrates the positive patient care outcomes and reductions in overall cost associated with the PCMH. These findings were compiled by the Patient-Centered Primary Care Collaborative—which issues annual reports on the progress of the PCMH—in a January 2014 update.¹⁸

In a PCMH model, the focus shifts away from the procedure(s) or treatment to the whole person. However, all components of care (eg, primary and specialty care, hospital, ancillary services, laboratory, and radiology) are vital and need to be connected to increase efficiency and reduce cost—creating what is sometimes referred to as a “medical neighborhood.”^19-21

What’s in the neighborhood?

In a medical neighborhood, such as an ACO, each patient is cared for by a team of providers at multiple locations. The PCMH serves as the base, ensuring that all providers work together toward a common goal. In addition to providing primary care, the PCMH coordinates each patient’s specialty and support services and communicates the care plan to all involved.

Successful implementation of a medical neighborhood requires a close working relationship among providers, payers, and community resources. For example, payers can provide real-time information about patients who have been admitted to the hospital or discharged from the emergency department, which enables close follow-up and coordination across multiple systems.

Nontraditional settings. Another facet of the medical neighborhood is the provision of health care services in nontraditional settings. For example, some grocery stores in our area employ nutritionists to whom we refer patients for nutritional counseling regarding their health in general or a disease process in particular.

Changes in reimbursement also will affect how care is delivered within the medical neighborhood. As we move away from fee-for-service (volume-based) to value-based payments, physicians who have made the transition from working individually with a panel of patients to providing team-based care within a PCMH will be better positioned to meet the goals of health care reform. (See “Team-based care is key inside the PCMH, too”²² below.) Nonetheless, the transition is a dynamic process. With changes in reimbursement and delivery models, physicians also will be expected to develop and implement continuous quality improvement measures so patient care can be continually evaluated and improved.

Now comes the hard part

While a PCMH requires primary care physicians to collaborate with other health professionals, it has the potential to lead to conflicts and debates about who is at the head of the health care team. This is particularly true within mental health services because, while primary care visits are frequently related to psychosocial issues, the mental health and general practice sectors have traditionally been distinct. In recent years, however, coordinated delivery models that integrate primary care and mental health services have been shown to increase access and reduce the stigma associated with mental health services—and to be cost-effective.²³

In many ways, moving the primary care culture from the traditional focus on the physician as “captain of the ship” to a physician-led, team-based approach is one of the most difficult tasks for organizations attempting to transform their care delivery models.^3,24 Physicians historically have been autonomous providers of medical care, relying on their own experience, expertise, and beliefs to guide decisions about patient care. Now they’re being asked to give up some of the direct control they may have had over patient care decisions and learn to work more collaboratively with other providers, as well as nonclinicians (eg, health coaches), to achieve desired outcomes.³ A successful transition depends on a reimbursement framework in which patient care goals are properly aligned with incentives for primary care physicians to work in a team-based environment.^3,25-27

Helping patients help themselves. Moving toward more patient-focused care will also require a concerted effort to increase patients’ engagement in their own health and medical care. In practice, very little of an individual’s time is spent in a physician’s office. Thus, optimal outcomes can be achieved only when patients are actively involved. Helping patients become proactive—ie, by arming them with the knowledge, skill, and confidence to do their part in staying healthy²⁸—also represents a major shift in primary care culture, as patients become active participants in medical decision making rather than passive recipients of physicians’ advice.

Alternative approaches. To deliver the continuum of care that is central to new Moving primary care culture to a physician-led, team-based approach is one of the most difficult tasks when transforming an organization's care delivery model.care delivery models and shift the culture of primary care toward a PCMH, physicians can implement a number of clinic-based engagement approaches—interacting with patients via e-visits such as e-mail through a secure portal, telemedicine, and group medical visits, for example. Physicians can encourage patient participation by starting patient interest groups and advisory panels²⁹—recommended by the NCQA and the Agency for Healthcare Research and Quality and used at our institution—and conducting patient needs assessments on a regular basis. Opportunities for primary care practices to engage with the community include partnering with local health departments, churches, nonprofits, and advocacy organizations to conduct health promotion and educational activities.

CORRESPONDENCE
Randy Wexler, MD, MPH, Department of Family Medicine, College of Medicine, The Ohio State University, 2231 North High Street, Northwood and High Building, Columbus, OH 43201; [email protected]

Pressure to reform our health care system is at an all-time high, driven by relentlessly rising costs and fragmentation of care. These persistent problems have led to lower quality care and limited access to care for a large proportion of the US population—issues that accountable care organizations (ACOs), as well as other value-based models, are designed to address.

While the terms used to describe the means by which health care systems attempt to do more to meet the needs of those they serve may vary, the importance of reorganizing care delivery to better integrate services is gaining traction nationwide. As we move to new models, primary care takes center stage.

ACOs (or ACO-type arrangements) anchored by primary care networks can help meet the goals of health care reform by responding to changes in reimbursement, reducing fragmented care, and focusing on improving the quality of care for defined patient populations. In addition, these delivery models can take advantage of new health information technology (IT) and the move toward patient-centered medical homes (PCMHs).

In the pages that follow, we examine opportunities for new care delivery models to slow rising costs and improve population health in family medicine. The introduction of these models has important implications for patients, physicians, and provider organizations, and our aim is to ensure that family physicians are prepared to take these vital steps toward achieving health reform goals.

Shifting from volume-based to value-based reimbursement

According to the Congressional Budget Office, ACOs are expected to save $5 billion during their first 8 years of existence.¹ After one year of ACO activity, the Centers for Medicare and Medicaid Services (CMS) reported savings of $30 million.² The expected savings will be driven by the increased provider accountability associated with ACOs.

Various means of increasing provider accountability through changes in reimbursement strategies have been proposed; several are new, while others are improvements on—or variations of—methods that have been tried before. The most common approaches—shared savings, shared savings plus penalty, capitation, episodic payment, prospective payment, pay-for-performance, and hospital-physician bundling—are detailed in TABLE 1.^3-7 Broad implementation of any of these reimbursement mechanisms within a new model of care would represent a shift away from the traditional volume-based provider payment model to a value-based system—a key step in slowing the rise in health care costs.

TABLE 1
Reimbursement strategies designed to promote physician accountability^3-7

^{ACO, accountable care organization; CMS, Centers for Medicare and Medicaid Services; DRG, diagnosis-related group; FFS, fee-for-service; FPs, family physicians; IT, information technology; PCMH, patient-centered medical home.}

Although it is too soon to be certain of the effect such changes will have on the earnings of family physicians, it is reasonable to think that new payment strategies—and a larger role for primary care providers—will improve their financial standing.

Moving toward population health management

New ACO-type models also make it easier to improve health care for specific populations, using strategies designed to organize, provide, and manage care for defined groups. In addition to controlling the cost of caring for specific groups, well-designed and implemented population health management strategies can increase continuity of care by ensuring oversight of patients across the spectrum of health care settings.

Five broad categories of population health management are most prominent: lifestyle management and demand management (both for relatively healthy people), disease management (for those with chronic conditions), catastrophic care management (for patients with rare or catastrophic illness or injury), and disability management (for groups of employees).⁸TABLE 2⁸ describes the population targeted and activities associated with each. It is important to remember, however, that no single strategy is mutually exclusive for a particular group.

Comprehensive Primary Care (CPC) initiative. In a collaboration made possible by the Affordable Care Act, CMS initiated the CPC initiative in 2012.⁹ A 4-year project designed to test and further identify the benefits of population health management and strengthen coordination of care for Medicare patients within primary care settings, the CPC initiative involves 497 sites and 2347 providers caring for approximately 315,000 Medicare beneficiaries.⁹ More information is available at http://innovation.cms.gov/initiatives/comprehensive-primary-care-initiative/.

TABLE 2
Population health management strategies⁸

^{QOL, quality of life.}

Building infrastructure and leveraging IT

ACOs and ACO-type models will take a variety of forms, depending in part on geographic need and local demographics. Yet, all share a common need for a strong infrastructure to support improved transitions, integration, and coordination of care. Incorporation of a strong health IT system is critical so that data regarding the process and cost of care, as well as outcomes, can be collected and put to optimal use.^10-13

Across health care settings, health IT innovations are being successfully implemented in efforts to enhance physician decision support, improve patient safety, increase guideline adherence, and improve chronic disease treatment.^11,13 In primary care, for example, an IT infrastructure can create disease registries so that data on patients with specific conditions can be tracked and acted upon—eg, a diabetes registry could be used to identify and contact patients who have an hemoglobin A1c >9 and have not been seen in 9 months or more.

Similarly, an IT system with the ability to identify patients at high risk for disease decompensation, hospitalization, and/or increased morbidity and mortality linked to progression of a chronic disease is needed. Identifying medical conditions associated with higher costs would make it possible to focus care coordination and chronic care management efforts on this targeted population.

As health IT continues to evolve, additional means of interacting with patients and improving patient care will be developed. Physicians and organizations that are ready to take advantage of these advances in technology will be well positioned to address the goals of health reform. (See “Health care reform: Recommendations for family physicians” below.^3-7,14,15)

How the patient-centered medical home fits into the picture

The implementation of ACOs and other new models of care has promising implications for the establishment of PCMHs. Consistent with the goals of health reform, the PCMH movement focuses on a coordinated teamwork approach, anchored within a general practice or family medicine setting.

An evaluation of the PCMH National Demonstration Project funded by the American Academy of Family Physicians found that the adoption of more components of the PCMH at the practice level was associated with improvement in patient outcomes, as measured by the Ambulatory Care Quality Alliance starter set16 (a compilation of clinical performance measures developed by a broad coalition of providers, payers, consumers, and government agencies).

A recent look at the Southeastern Pennsylvania Chronic Care Initiative¹⁷ found less promising results. “A multipayer medical home pilot, in which participating practices adopted new structural capabilities and received NCQA [National Committee for Quality Assurance] certification, was associated with limited improvements in quality and was not associated with reductions in utilization of hospital, emergency department, or ambulatory care services or total costs over 3 years,” Friedberg et al¹⁷ concluded. The authors did note, however, that NCQA recognition was what the practices involved in this initiative were rewarded for—not PCMH activity.

Coordinated delivery models that integrate primary care and mental health services have been shown to be cost-effective.It is also important to keep in mind that PCMH activity has been shown to improve care and reduce costs—not NCQA recognition in and of itself. In fact, a large body of evidence clearly demonstrates the positive patient care outcomes and reductions in overall cost associated with the PCMH. These findings were compiled by the Patient-Centered Primary Care Collaborative—which issues annual reports on the progress of the PCMH—in a January 2014 update.¹⁸

In a PCMH model, the focus shifts away from the procedure(s) or treatment to the whole person. However, all components of care (eg, primary and specialty care, hospital, ancillary services, laboratory, and radiology) are vital and need to be connected to increase efficiency and reduce cost—creating what is sometimes referred to as a “medical neighborhood.”^19-21

What’s in the neighborhood?

In a medical neighborhood, such as an ACO, each patient is cared for by a team of providers at multiple locations. The PCMH serves as the base, ensuring that all providers work together toward a common goal. In addition to providing primary care, the PCMH coordinates each patient’s specialty and support services and communicates the care plan to all involved.

Successful implementation of a medical neighborhood requires a close working relationship among providers, payers, and community resources. For example, payers can provide real-time information about patients who have been admitted to the hospital or discharged from the emergency department, which enables close follow-up and coordination across multiple systems.

Nontraditional settings. Another facet of the medical neighborhood is the provision of health care services in nontraditional settings. For example, some grocery stores in our area employ nutritionists to whom we refer patients for nutritional counseling regarding their health in general or a disease process in particular.

Changes in reimbursement also will affect how care is delivered within the medical neighborhood. As we move away from fee-for-service (volume-based) to value-based payments, physicians who have made the transition from working individually with a panel of patients to providing team-based care within a PCMH will be better positioned to meet the goals of health care reform. (See “Team-based care is key inside the PCMH, too”²² below.) Nonetheless, the transition is a dynamic process. With changes in reimbursement and delivery models, physicians also will be expected to develop and implement continuous quality improvement measures so patient care can be continually evaluated and improved.

Now comes the hard part

While a PCMH requires primary care physicians to collaborate with other health professionals, it has the potential to lead to conflicts and debates about who is at the head of the health care team. This is particularly true within mental health services because, while primary care visits are frequently related to psychosocial issues, the mental health and general practice sectors have traditionally been distinct. In recent years, however, coordinated delivery models that integrate primary care and mental health services have been shown to increase access and reduce the stigma associated with mental health services—and to be cost-effective.²³

In many ways, moving the primary care culture from the traditional focus on the physician as “captain of the ship” to a physician-led, team-based approach is one of the most difficult tasks for organizations attempting to transform their care delivery models.^3,24 Physicians historically have been autonomous providers of medical care, relying on their own experience, expertise, and beliefs to guide decisions about patient care. Now they’re being asked to give up some of the direct control they may have had over patient care decisions and learn to work more collaboratively with other providers, as well as nonclinicians (eg, health coaches), to achieve desired outcomes.³ A successful transition depends on a reimbursement framework in which patient care goals are properly aligned with incentives for primary care physicians to work in a team-based environment.^3,25-27