Sarah-Anne Schumann, MD

ILLUSTRATIVE CASE

A 71-year-old woman with diabetes and coronary artery disease has just been admitted to the ICU, where she’ll receive treatment for sepsis, multilobar pneumonia, and respiratory failure requiring mechanical ventilation. Her blood sugar is 253 mg/dL. In writing her admission orders, you contemplate targets for glycemic control. How low should you go?

Hyperglycemia is common in patients admitted to intensive care, whether or not they have diabetes. Elevated blood sugar is associated with stress and trauma and affects both postoperative and critically ill medical patients. A wealth of evidence has demonstrated that hyperglycemia is associated with poorer outcomes and increased mortality in this patient population, including those with myocardial infarction, stroke, trauma, and other medical conditions.^2-5 Thus, intensive glucose control is the standard of care in the ICU, based on consensus guidelines from such groups as the American Diabetes Association (ADA) and the Surviving Sepsis Campaign—an initiative developed by 3 critical care organizations and endorsed by 16 specialty groups.^6-8

Is intense therapy better? Study results differ
The association between hyperglycemia and an increased risk of death led investigators to study the effectiveness of aggressive treatment with insulin in decreasing morbidity and mortality. A 2004 meta-analysis of 35 trials comparing insulin vs no insulin in critically ill hospitalized patients demonstrated a 15% reduction in short-term mortality among patients treated with insulin.⁹ A 2008 meta-analysis of 29 randomized trials, including data from 8432 adult ICU patients, compared intensive insulin therapy with conventional therapy—and found that intensive therapy did not lower hospital mortality rates compared with conventional therapy. In addition, this meta-analysis revealed a marked increase in severe hypoglycemia (blood sugar ≤40 mg/dL) in the intensive therapy group.¹⁰ (The intensive therapy group included studies with glucose goals of ≤110 mg/dL and <150 mg/dL in about equal numbers; conventional therapy goals were generally between 180 and 200 mg/dL.)

The studies included in both the meta-analyses, however, were mostly small, single-center trials, and of low-to-medium quality. In addition, methods for achieving glycemic control varied. Nonetheless, current consensus guidelines set a goal for glucose levels of 80 to 110 mg/dL for all critically ill hospitalized patients.^6-8 But because of the lack of sufficient high-quality evidence from a single large RCT, Finfer et al conducted the large study described here to clearly establish that intensive glycemic control decreases all-cause mortality. Given their hypothesis, the results were surprising.

STUDY SUMMARY: Intensive therapy does more harm than help

NICE-SUGAR (Normoglycaemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation) was a large-scale, multicenter, multinational trial comparing aggressive blood sugar control (goal 81-108 mg/dL) with conventional therapy (goal ≤180 mg/dL) in 6104 critically ill hospitalized patients with hyperglycemia. Patients were followed for 90 days. The primary end point was death from any cause 90 days after randomization. Secondary outcomes included survival time during the first 90 days, specific cause of death, duration of mechanical ventilation, renal replacement therapy, and length of stays in the ICU and in the hospital. Other outcomes included death from any cause within 28 days, place of death, new organ failure, positive blood culture, blood transfusion, and units of blood transfused.

The study was conducted in 42 hospitals in Canada, Australia, and New Zealand. Patients had to have an anticipated ICU admission of 3 days or more and randomization had to occur within 24 hours of admission. The study protocol was discontinued when patients began eating or were discharged from the ICU; if they were readmitted to the ICU within 90 days of randomization, the study protocol was resumed.

Treatment assignment was revealed to clinical staff after randomization, and was determined by a specific algorithm ( https://studies.thegeorgeinstitute.org/nice/ ). Blood sugar levels were managed with insulin infusions.

In the conventional group, insulin was started at 1 unit/h for glucose levels >180 mg/dL, and decreased or stopped when levels were <144 mg/dL, depending on previous glucose value and current rate of drip. In the intensive therapy group, insulin was initiated for lower levels (blood sugar >109 mg/dL) and at a higher rate (2 units/h). The insulin rate was decreased or maintained for glucose levels from 64 to 80 mg/dL, depending on previous glucose value and current rate of drip. Insulin was withheld for blood sugar levels of <64 mg/dL.

Contrary to the hypothesis, intensive therapy spelled trouble. Patients with intensive glycemic control had an all-cause mortality rate of 27.5%, compared with a rate of 24.9% for patients in the conventional therapy group (P=.04, number needed to harm [NNH]=38). Severe hypoglycemia (glucose ≤40 mg/dL) occurred in 6.8% of those in the intensive therapy group, compared with 0.5% in the conventional therapy group (P=.03, NNH=16).

Most of the deaths in both groups occurred in the ICU or in the hospital. Deaths from cardiovascular causes were more common among those in the intensive therapy group. There were no significant differences in any other outcomes. The mean glucose level in the intensive therapy group was 118, vs 145 mg/dL in the conventional therapy group.

For multivariate and subgroup analyses, the patients were assigned strata (Canada or Australia/New Zealand; operative vs nonoperative admission) or classified into groups (traumatic vs atraumatic; diabetes vs no diabetes; corticosteroids in previous 72 hours or not; high vs low critical illness symptom severity) based on predefined characteristics. No subgroups had significantly improved outcomes with intensive therapy.¹

WHAT’S NEW: Now we know: Don’t go too low

This study, in contrast to a number of smaller studies of lower quality, demonstrates a higher all-cause mortality rate at 90 days for critically ill patients receiving intensive glucose therapy. It is now clear that, among critically ill hospitalized patients, aiming for intensive glucose control (81-108 mg/dL) is associated with an increased rate of severe hypoglycemic events and all-cause mortality at 90 days. The previously used goal of conventional therapy (≤180 mg/dL) is safer.

CAVEATS: Study population may not reflect primary care

There are 2 caveats to this study. The first is that because of the nature of the research, it was impossible to maintain blinding of the clinical staff to patient assignments. The second important caveat pertains to the severity of illness among participants in this multicenter study: Most of these patients were in ICUs at tertiary care medical centers and had an expected ICU length of stay of 3 or more days. Although many family physicians manage patients in ICUs, the patients randomized in this study may represent a sicker than average patient population for some hospitals.

CHALLENGES TO IMPLEMENTATION: Some may doubt validity of this outcome

Less aggressive glycemic control for critically ill patients should be easier to achieve, not more difficult. However, a change in glucose targets may require new admission order sets and, notably, reeducation of physicians and nurses who have been convinced by earlier studies that more intensive glucose control is superior.

Acknowledgments

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 71-year-old woman with diabetes and coronary artery disease has just been admitted to the ICU, where she’ll receive treatment for sepsis, multilobar pneumonia, and respiratory failure requiring mechanical ventilation. Her blood sugar is 253 mg/dL. In writing her admission orders, you contemplate targets for glycemic control. How low should you go?

Hyperglycemia is common in patients admitted to intensive care, whether or not they have diabetes. Elevated blood sugar is associated with stress and trauma and affects both postoperative and critically ill medical patients. A wealth of evidence has demonstrated that hyperglycemia is associated with poorer outcomes and increased mortality in this patient population, including those with myocardial infarction, stroke, trauma, and other medical conditions.^2-5 Thus, intensive glucose control is the standard of care in the ICU, based on consensus guidelines from such groups as the American Diabetes Association (ADA) and the Surviving Sepsis Campaign—an initiative developed by 3 critical care organizations and endorsed by 16 specialty groups.^6-8

Is intense therapy better? Study results differ
The association between hyperglycemia and an increased risk of death led investigators to study the effectiveness of aggressive treatment with insulin in decreasing morbidity and mortality. A 2004 meta-analysis of 35 trials comparing insulin vs no insulin in critically ill hospitalized patients demonstrated a 15% reduction in short-term mortality among patients treated with insulin.⁹ A 2008 meta-analysis of 29 randomized trials, including data from 8432 adult ICU patients, compared intensive insulin therapy with conventional therapy—and found that intensive therapy did not lower hospital mortality rates compared with conventional therapy. In addition, this meta-analysis revealed a marked increase in severe hypoglycemia (blood sugar ≤40 mg/dL) in the intensive therapy group.¹⁰ (The intensive therapy group included studies with glucose goals of ≤110 mg/dL and <150 mg/dL in about equal numbers; conventional therapy goals were generally between 180 and 200 mg/dL.)

The studies included in both the meta-analyses, however, were mostly small, single-center trials, and of low-to-medium quality. In addition, methods for achieving glycemic control varied. Nonetheless, current consensus guidelines set a goal for glucose levels of 80 to 110 mg/dL for all critically ill hospitalized patients.^6-8 But because of the lack of sufficient high-quality evidence from a single large RCT, Finfer et al conducted the large study described here to clearly establish that intensive glycemic control decreases all-cause mortality. Given their hypothesis, the results were surprising.

STUDY SUMMARY: Intensive therapy does more harm than help

NICE-SUGAR (Normoglycaemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation) was a large-scale, multicenter, multinational trial comparing aggressive blood sugar control (goal 81-108 mg/dL) with conventional therapy (goal ≤180 mg/dL) in 6104 critically ill hospitalized patients with hyperglycemia. Patients were followed for 90 days. The primary end point was death from any cause 90 days after randomization. Secondary outcomes included survival time during the first 90 days, specific cause of death, duration of mechanical ventilation, renal replacement therapy, and length of stays in the ICU and in the hospital. Other outcomes included death from any cause within 28 days, place of death, new organ failure, positive blood culture, blood transfusion, and units of blood transfused.

The study was conducted in 42 hospitals in Canada, Australia, and New Zealand. Patients had to have an anticipated ICU admission of 3 days or more and randomization had to occur within 24 hours of admission. The study protocol was discontinued when patients began eating or were discharged from the ICU; if they were readmitted to the ICU within 90 days of randomization, the study protocol was resumed.

Treatment assignment was revealed to clinical staff after randomization, and was determined by a specific algorithm ( https://studies.thegeorgeinstitute.org/nice/ ). Blood sugar levels were managed with insulin infusions.

In the conventional group, insulin was started at 1 unit/h for glucose levels >180 mg/dL, and decreased or stopped when levels were <144 mg/dL, depending on previous glucose value and current rate of drip. In the intensive therapy group, insulin was initiated for lower levels (blood sugar >109 mg/dL) and at a higher rate (2 units/h). The insulin rate was decreased or maintained for glucose levels from 64 to 80 mg/dL, depending on previous glucose value and current rate of drip. Insulin was withheld for blood sugar levels of <64 mg/dL.

Contrary to the hypothesis, intensive therapy spelled trouble. Patients with intensive glycemic control had an all-cause mortality rate of 27.5%, compared with a rate of 24.9% for patients in the conventional therapy group (P=.04, number needed to harm [NNH]=38). Severe hypoglycemia (glucose ≤40 mg/dL) occurred in 6.8% of those in the intensive therapy group, compared with 0.5% in the conventional therapy group (P=.03, NNH=16).

Most of the deaths in both groups occurred in the ICU or in the hospital. Deaths from cardiovascular causes were more common among those in the intensive therapy group. There were no significant differences in any other outcomes. The mean glucose level in the intensive therapy group was 118, vs 145 mg/dL in the conventional therapy group.

For multivariate and subgroup analyses, the patients were assigned strata (Canada or Australia/New Zealand; operative vs nonoperative admission) or classified into groups (traumatic vs atraumatic; diabetes vs no diabetes; corticosteroids in previous 72 hours or not; high vs low critical illness symptom severity) based on predefined characteristics. No subgroups had significantly improved outcomes with intensive therapy.¹

WHAT’S NEW: Now we know: Don’t go too low

This study, in contrast to a number of smaller studies of lower quality, demonstrates a higher all-cause mortality rate at 90 days for critically ill patients receiving intensive glucose therapy. It is now clear that, among critically ill hospitalized patients, aiming for intensive glucose control (81-108 mg/dL) is associated with an increased rate of severe hypoglycemic events and all-cause mortality at 90 days. The previously used goal of conventional therapy (≤180 mg/dL) is safer.

CAVEATS: Study population may not reflect primary care

There are 2 caveats to this study. The first is that because of the nature of the research, it was impossible to maintain blinding of the clinical staff to patient assignments. The second important caveat pertains to the severity of illness among participants in this multicenter study: Most of these patients were in ICUs at tertiary care medical centers and had an expected ICU length of stay of 3 or more days. Although many family physicians manage patients in ICUs, the patients randomized in this study may represent a sicker than average patient population for some hospitals.

CHALLENGES TO IMPLEMENTATION: Some may doubt validity of this outcome

Less aggressive glycemic control for critically ill patients should be easier to achieve, not more difficult. However, a change in glucose targets may require new admission order sets and, notably, reeducation of physicians and nurses who have been convinced by earlier studies that more intensive glucose control is superior.

Acknowledgments

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 71-year-old woman with diabetes and coronary artery disease has just been admitted to the ICU, where she’ll receive treatment for sepsis, multilobar pneumonia, and respiratory failure requiring mechanical ventilation. Her blood sugar is 253 mg/dL. In writing her admission orders, you contemplate targets for glycemic control. How low should you go?

Hyperglycemia is common in patients admitted to intensive care, whether or not they have diabetes. Elevated blood sugar is associated with stress and trauma and affects both postoperative and critically ill medical patients. A wealth of evidence has demonstrated that hyperglycemia is associated with poorer outcomes and increased mortality in this patient population, including those with myocardial infarction, stroke, trauma, and other medical conditions.^2-5 Thus, intensive glucose control is the standard of care in the ICU, based on consensus guidelines from such groups as the American Diabetes Association (ADA) and the Surviving Sepsis Campaign—an initiative developed by 3 critical care organizations and endorsed by 16 specialty groups.^6-8

Is intense therapy better? Study results differ
The association between hyperglycemia and an increased risk of death led investigators to study the effectiveness of aggressive treatment with insulin in decreasing morbidity and mortality. A 2004 meta-analysis of 35 trials comparing insulin vs no insulin in critically ill hospitalized patients demonstrated a 15% reduction in short-term mortality among patients treated with insulin.⁹ A 2008 meta-analysis of 29 randomized trials, including data from 8432 adult ICU patients, compared intensive insulin therapy with conventional therapy—and found that intensive therapy did not lower hospital mortality rates compared with conventional therapy. In addition, this meta-analysis revealed a marked increase in severe hypoglycemia (blood sugar ≤40 mg/dL) in the intensive therapy group.¹⁰ (The intensive therapy group included studies with glucose goals of ≤110 mg/dL and <150 mg/dL in about equal numbers; conventional therapy goals were generally between 180 and 200 mg/dL.)

The studies included in both the meta-analyses, however, were mostly small, single-center trials, and of low-to-medium quality. In addition, methods for achieving glycemic control varied. Nonetheless, current consensus guidelines set a goal for glucose levels of 80 to 110 mg/dL for all critically ill hospitalized patients.^6-8 But because of the lack of sufficient high-quality evidence from a single large RCT, Finfer et al conducted the large study described here to clearly establish that intensive glycemic control decreases all-cause mortality. Given their hypothesis, the results were surprising.

STUDY SUMMARY: Intensive therapy does more harm than help

NICE-SUGAR (Normoglycaemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation) was a large-scale, multicenter, multinational trial comparing aggressive blood sugar control (goal 81-108 mg/dL) with conventional therapy (goal ≤180 mg/dL) in 6104 critically ill hospitalized patients with hyperglycemia. Patients were followed for 90 days. The primary end point was death from any cause 90 days after randomization. Secondary outcomes included survival time during the first 90 days, specific cause of death, duration of mechanical ventilation, renal replacement therapy, and length of stays in the ICU and in the hospital. Other outcomes included death from any cause within 28 days, place of death, new organ failure, positive blood culture, blood transfusion, and units of blood transfused.

The study was conducted in 42 hospitals in Canada, Australia, and New Zealand. Patients had to have an anticipated ICU admission of 3 days or more and randomization had to occur within 24 hours of admission. The study protocol was discontinued when patients began eating or were discharged from the ICU; if they were readmitted to the ICU within 90 days of randomization, the study protocol was resumed.

Treatment assignment was revealed to clinical staff after randomization, and was determined by a specific algorithm ( https://studies.thegeorgeinstitute.org/nice/ ). Blood sugar levels were managed with insulin infusions.

In the conventional group, insulin was started at 1 unit/h for glucose levels >180 mg/dL, and decreased or stopped when levels were <144 mg/dL, depending on previous glucose value and current rate of drip. In the intensive therapy group, insulin was initiated for lower levels (blood sugar >109 mg/dL) and at a higher rate (2 units/h). The insulin rate was decreased or maintained for glucose levels from 64 to 80 mg/dL, depending on previous glucose value and current rate of drip. Insulin was withheld for blood sugar levels of <64 mg/dL.

Contrary to the hypothesis, intensive therapy spelled trouble. Patients with intensive glycemic control had an all-cause mortality rate of 27.5%, compared with a rate of 24.9% for patients in the conventional therapy group (P=.04, number needed to harm [NNH]=38). Severe hypoglycemia (glucose ≤40 mg/dL) occurred in 6.8% of those in the intensive therapy group, compared with 0.5% in the conventional therapy group (P=.03, NNH=16).

Most of the deaths in both groups occurred in the ICU or in the hospital. Deaths from cardiovascular causes were more common among those in the intensive therapy group. There were no significant differences in any other outcomes. The mean glucose level in the intensive therapy group was 118, vs 145 mg/dL in the conventional therapy group.

For multivariate and subgroup analyses, the patients were assigned strata (Canada or Australia/New Zealand; operative vs nonoperative admission) or classified into groups (traumatic vs atraumatic; diabetes vs no diabetes; corticosteroids in previous 72 hours or not; high vs low critical illness symptom severity) based on predefined characteristics. No subgroups had significantly improved outcomes with intensive therapy.¹

WHAT’S NEW: Now we know: Don’t go too low

This study, in contrast to a number of smaller studies of lower quality, demonstrates a higher all-cause mortality rate at 90 days for critically ill patients receiving intensive glucose therapy. It is now clear that, among critically ill hospitalized patients, aiming for intensive glucose control (81-108 mg/dL) is associated with an increased rate of severe hypoglycemic events and all-cause mortality at 90 days. The previously used goal of conventional therapy (≤180 mg/dL) is safer.

CAVEATS: Study population may not reflect primary care

There are 2 caveats to this study. The first is that because of the nature of the research, it was impossible to maintain blinding of the clinical staff to patient assignments. The second important caveat pertains to the severity of illness among participants in this multicenter study: Most of these patients were in ICUs at tertiary care medical centers and had an expected ICU length of stay of 3 or more days. Although many family physicians manage patients in ICUs, the patients randomized in this study may represent a sicker than average patient population for some hospitals.

CHALLENGES TO IMPLEMENTATION: Some may doubt validity of this outcome

Less aggressive glycemic control for critically ill patients should be easier to achieve, not more difficult. However, a change in glucose targets may require new admission order sets and, notably, reeducation of physicians and nurses who have been convinced by earlier studies that more intensive glucose control is superior.

Acknowledgments

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old patient with a body mass index (BMI) of 18 tells you she’s fat and she’s afraid of gaining weight. Further questioning reveals that your patient binges on cookies and potato chips about once a week, then compensates for overeating by taking laxatives or exercising excessively—a practice she’s been following since she started college several years ago. The eating disorder she describes does not meet the criteria in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) for bulimia or anorexia nervosa, although it has elements of both. Rather, it fits the diagnostic criteria for eating disorder NOS. You’re aware that CBT is the first-line behavioral treatment for bulimia, and wonder whether it would be helpful for your patient.

Eating disorders often go undetected and untreated in primary care practices,³ as many patients don’t volunteer information about their weight or behaviors related to food, and physicians often fail to ask. Overall, as few as 10% of those with eating disorders receive any form of treatment.¹

Would you recognize this loosely defined disorder?

In the United States, the lifetime prevalence of eating disorders is 0.6% for anorexia nervosa (0.3% for men and 0.9% for women), 1.0% for bulimia (0.5% for men and 1.5% for women), and 2.8% for binge-eating disorder (2.0% for men, 3.5% for women).⁴ Eating disorder NOS, which encompasses subthreshold cases of anorexia or bulimia, patients with elements of both anorexia and bulimia, and patients with binge-eating disorder, accounts for 50% to 80% of eating disorder diagnoses in outpatient settings. Yet there have been few studies of the treatment of these patients.^2,5,6

A review of DSM-IV criteria

The diagnostic criteria for anorexia nervosa include a refusal to maintain a weight of at least 85% of normal body weight (or having a BMI ≤17.5), intense fear of gaining weight, disturbance in the way one’s body shape is experienced, and amenorrhea in females who are post-menarche.

Criteria for bulimia include recurrent episodes of binge eating (consuming a large amount of food with a sense of lack of control over eating) and recurrent inappropriate compensatory behaviors to prevent weight gain (self-induced vomiting, excessive exercise, fasting, laxatives, diuretics, or enemas) at least twice weekly for 3 months; and self-evaluation that is unduly influenced by body shape and weight.⁷ Most patients with eating disorder NOS have clinical features of both anorexia and bulimia.⁶

APA guidelines are silent on NOS

CBT has consistently proven to be the most useful behavioral treatment for patients with bulimia.¹ Selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine—the only medication with Food and Drug Administration approval for the treatment of an eating disorder⁸ —are about as effective as CBT, and the combination of CBT and an SSRI is superior to either treatment alone.⁹ CBT has also been found to be somewhat effective in treating binge-eating disorder.¹⁰

Anorexia nervosa, the most deadly eating disorder (the mortality rate is 6.6%¹¹ ) and the most difficult to treat, is the exception. Several studies have assessed CBT for treating anorexia, but it has not been found to be very effective.^10,12,13

The 2006 American Psychiatric Association practice guidelines for the treatment of patients with eating disorders feature recommendations for anorexia, bulimia, and binge-eating disorder, but do not address eating disorder NOS.¹⁰ The National Institute for Clinical Excellence (NICE) in the United Kingdom issued guidelines for the treatment of eating disorders in 2004. In response to the lack of evidence for treating eating disorder NOS, NICE recommended basing treatment on the form of eating disorder that most closely resembles the patient’s presentation.¹⁴ Fairburn et al addressed the lack of evidence for treatment of eating disorder NOS with the study summarized here.

STUDY SUMMARY: Both broad and focused CBT delivered results

Conducted at 2 eating disorder centers in the United Kingdom, this RCT included 154 patients, 18 to 65 years of age, who met DSM-IV criteria for either bulimia or eating disorder NOS. Exclusion criteria included prior treatment with CBT or other evidence-based treatment for the same eating disorder, and a BMI ≤17.5.

Most of the patients were female (95.5%) and white (90.3%), with a median age of 26 years and a median duration of eating disorder of 8.6 years. Sixty-two percent of the patients had a diagnosis of eating disorder NOS, and 38% were diagnosed with bulimia. Half of the patients had another current psychiatric diagnosis—a major depressive disorder, an anxiety disorder, or substance abuse.

The patients were randomized into 4 groups: Two received immediate treatment, and the other 2, referred to as waiting list controls, waited 8 weeks before beginning treatment. Treatment consisted of 1 of 2 forms of CBT-E, an enhanced form of CBT used to treat adult outpatients with eating disorders. Patients either received CBT-Ef, a focused form of CBT that exclusively targets eating disorder psychopathology, or CBT-Eb, a broader form of therapy that also addresses other problems that are common in patients with eating disorders, such as perfectionism and low self-esteem.

Both types of CBT-E featured a 90-minute preparatory session, 20 50-minute sessions, and 1 review session 20 weeks after completion of treatment. In the first 4 sessions, CBT-Ef and CBT-Eb were identical—addressing the eating disorder exclusively. CBT-Ef continued to focus on the eating disorder for the rest of the sessions, while subsequent CBT-Eb sessions also dealt with mood intolerance, interpersonal difficulties, and related issues. Five therapists—4 psychologists and 1 nurse-therapist—conducted the treatments.

Patients were weaned from ongoing psychiatric therapy during the study, but those who were on antidepressant therapy (n=76) were able to continue it. Patients were assessed before treatment, at the end of the waiting period for those in the control groups, after 8 weeks of treatment, at the end of treatment, and 20, 40, and 60 weeks after completion of treatment. (Twenty-two percent of the enrollees did not complete treatment.)

Primary outcomes were based on the Eating Disorder Examination (EDE), administered by assessors who were not involved in the treatment and were blinded to the patients’ group assignment. Change in severity of eating disorder features was measured by the global EDE score (0-6) and attaining a global EDE score <1.74 (<1 standard deviation above the community mean).

No treatment vs CBT. The waiting period left little doubt of the short-term efficacy of CBT: After 8 weeks, there was significant improvement in eating disorder behaviors and overall severity in both the CBT-Ef and CBT-Eb groups (EDE scores fell from 4.15 at baseline to 3.26 and from 4.04 to 2.89, respectively). In the same time period, scores for the waiting list control groups remained flat (from 4.08 at baseline to 3.99).

At the end of treatment and at the 60-week follow-up, patients in both forms of CBT-E showed significant improvement across all measures, with no significant difference between treatments. By the end of treatment, 66.4% of those who completed all of the CBT sessions had global EDE scores <1.74 (considered clinically significant).

Subgroup analysis offers opportunity for fine-tuning

When analyzed separately, the patients with bulimia and those with eating disorder NOS did equally well at the end of treatment: 52.7% of those with bulimia and 53.3% of those with eating disorder NOS had global EDE scores <1.74. At the 60-week follow-up, the patients with bulimia maintained their improvement slightly more: 61.4% had global EDE scores <1.74, compared with 45.7% of the patients with eating disorder NOS.

The researchers also compared the outcomes of patients with the most complex additional psychopathology with those of patients with less complex problems. Greater complexity was defined as moderate ratings in at least 2 of the following domains: mood intolerance, clinical perfectionism, low self-esteem, and interpersonal difficulties.

Broad focus more effective for high complexity. Overall, those in the more complex subgroup did not respond as well; 48% had global EDE scores <1.74, vs 60% of those in the less complex group. However, those in the more complex subgroup did better with the broad form of CBT (at 60-week follow up, 60% had scores <1.74 with CBT-Eb, compared with 40% in the CBT-Ef treatment arm), while the less complex subgroup did better with the more tightly focused CBT-Ef. ²

WHAT’S NEW: Evidence supports CBT for NOS diagnosis

The most recent (2004) Cochrane review of “psychotherapy for bulimia nervosa and binging” included 40 RCTs of patients with bulimia, binge-eating disorder, and eating disorder NOS with recurrent binge-eating episodes (included in 7 studies). While the review confirmed that CBT is effective for bulimia and “similar syndromes,” it identified a need for larger and higher quality trials of CBT, particularly in patients with eating disorder NOS.¹ The study reviewed in this PURL—the first large, high-quality trial to include a number of patients with eating disorder NOS—provides strong evidence that CBT is effective for this group of patients.²

CAVEATS: Limited wait time leaves unanswered questions

One limitation of this study is the lack of a control group beyond the 8-week waiting period. Prior studies of CBT for bulimia that delayed therapy for those in the control groups for a longer duration have consistently shown that patients receiving CBT did significantly better than those in the control group.⁹ While a “no treatment” group would have made the results more robust in this case, it would not have been ethical to withhold treatment for the entire length of the study.

It is noteworthy, too, that this study only included patients with a BMI >17.5. Patients with a diagnosis of anorexia nervosa, who by definition have a lower BMI, will need other treatments, including hospitalization in some cases.

CHALLENGES TO IMPLEMENTATION: Identifying patients and therapists

The primary challenge is to determine which of your patients have eating disorders. When discussing diet, adding a simple question such as, “Are you happy with your current weight?” can help you identify those who meet the criteria for an eating disorder or are at high risk.³

Identifying local mental health providers who are trained to provide CBT for patients with eating disorders is another concern. Insurance coverage for this intensive treatment may also be a limiting factor in some cases.

Many studies support the use of fluoxetine for patients with bulimia, and combined treatment with SSRIs and CBT has been shown to be superior to either treatment alone.^8,10,14 Consider starting the patient on an antidepressant while she (or he) awaits the start of CBT.

Acknowledgements

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURLs) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old patient with a body mass index (BMI) of 18 tells you she’s fat and she’s afraid of gaining weight. Further questioning reveals that your patient binges on cookies and potato chips about once a week, then compensates for overeating by taking laxatives or exercising excessively—a practice she’s been following since she started college several years ago. The eating disorder she describes does not meet the criteria in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) for bulimia or anorexia nervosa, although it has elements of both. Rather, it fits the diagnostic criteria for eating disorder NOS. You’re aware that CBT is the first-line behavioral treatment for bulimia, and wonder whether it would be helpful for your patient.

Eating disorders often go undetected and untreated in primary care practices,³ as many patients don’t volunteer information about their weight or behaviors related to food, and physicians often fail to ask. Overall, as few as 10% of those with eating disorders receive any form of treatment.¹

Would you recognize this loosely defined disorder?

In the United States, the lifetime prevalence of eating disorders is 0.6% for anorexia nervosa (0.3% for men and 0.9% for women), 1.0% for bulimia (0.5% for men and 1.5% for women), and 2.8% for binge-eating disorder (2.0% for men, 3.5% for women).⁴ Eating disorder NOS, which encompasses subthreshold cases of anorexia or bulimia, patients with elements of both anorexia and bulimia, and patients with binge-eating disorder, accounts for 50% to 80% of eating disorder diagnoses in outpatient settings. Yet there have been few studies of the treatment of these patients.^2,5,6

A review of DSM-IV criteria

The diagnostic criteria for anorexia nervosa include a refusal to maintain a weight of at least 85% of normal body weight (or having a BMI ≤17.5), intense fear of gaining weight, disturbance in the way one’s body shape is experienced, and amenorrhea in females who are post-menarche.

Criteria for bulimia include recurrent episodes of binge eating (consuming a large amount of food with a sense of lack of control over eating) and recurrent inappropriate compensatory behaviors to prevent weight gain (self-induced vomiting, excessive exercise, fasting, laxatives, diuretics, or enemas) at least twice weekly for 3 months; and self-evaluation that is unduly influenced by body shape and weight.⁷ Most patients with eating disorder NOS have clinical features of both anorexia and bulimia.⁶

APA guidelines are silent on NOS

CBT has consistently proven to be the most useful behavioral treatment for patients with bulimia.¹ Selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine—the only medication with Food and Drug Administration approval for the treatment of an eating disorder⁸ —are about as effective as CBT, and the combination of CBT and an SSRI is superior to either treatment alone.⁹ CBT has also been found to be somewhat effective in treating binge-eating disorder.¹⁰

Anorexia nervosa, the most deadly eating disorder (the mortality rate is 6.6%¹¹ ) and the most difficult to treat, is the exception. Several studies have assessed CBT for treating anorexia, but it has not been found to be very effective.^10,12,13

The 2006 American Psychiatric Association practice guidelines for the treatment of patients with eating disorders feature recommendations for anorexia, bulimia, and binge-eating disorder, but do not address eating disorder NOS.¹⁰ The National Institute for Clinical Excellence (NICE) in the United Kingdom issued guidelines for the treatment of eating disorders in 2004. In response to the lack of evidence for treating eating disorder NOS, NICE recommended basing treatment on the form of eating disorder that most closely resembles the patient’s presentation.¹⁴ Fairburn et al addressed the lack of evidence for treatment of eating disorder NOS with the study summarized here.

STUDY SUMMARY: Both broad and focused CBT delivered results

Conducted at 2 eating disorder centers in the United Kingdom, this RCT included 154 patients, 18 to 65 years of age, who met DSM-IV criteria for either bulimia or eating disorder NOS. Exclusion criteria included prior treatment with CBT or other evidence-based treatment for the same eating disorder, and a BMI ≤17.5.

Most of the patients were female (95.5%) and white (90.3%), with a median age of 26 years and a median duration of eating disorder of 8.6 years. Sixty-two percent of the patients had a diagnosis of eating disorder NOS, and 38% were diagnosed with bulimia. Half of the patients had another current psychiatric diagnosis—a major depressive disorder, an anxiety disorder, or substance abuse.

The patients were randomized into 4 groups: Two received immediate treatment, and the other 2, referred to as waiting list controls, waited 8 weeks before beginning treatment. Treatment consisted of 1 of 2 forms of CBT-E, an enhanced form of CBT used to treat adult outpatients with eating disorders. Patients either received CBT-Ef, a focused form of CBT that exclusively targets eating disorder psychopathology, or CBT-Eb, a broader form of therapy that also addresses other problems that are common in patients with eating disorders, such as perfectionism and low self-esteem.

Both types of CBT-E featured a 90-minute preparatory session, 20 50-minute sessions, and 1 review session 20 weeks after completion of treatment. In the first 4 sessions, CBT-Ef and CBT-Eb were identical—addressing the eating disorder exclusively. CBT-Ef continued to focus on the eating disorder for the rest of the sessions, while subsequent CBT-Eb sessions also dealt with mood intolerance, interpersonal difficulties, and related issues. Five therapists—4 psychologists and 1 nurse-therapist—conducted the treatments.

Patients were weaned from ongoing psychiatric therapy during the study, but those who were on antidepressant therapy (n=76) were able to continue it. Patients were assessed before treatment, at the end of the waiting period for those in the control groups, after 8 weeks of treatment, at the end of treatment, and 20, 40, and 60 weeks after completion of treatment. (Twenty-two percent of the enrollees did not complete treatment.)

Primary outcomes were based on the Eating Disorder Examination (EDE), administered by assessors who were not involved in the treatment and were blinded to the patients’ group assignment. Change in severity of eating disorder features was measured by the global EDE score (0-6) and attaining a global EDE score <1.74 (<1 standard deviation above the community mean).

No treatment vs CBT. The waiting period left little doubt of the short-term efficacy of CBT: After 8 weeks, there was significant improvement in eating disorder behaviors and overall severity in both the CBT-Ef and CBT-Eb groups (EDE scores fell from 4.15 at baseline to 3.26 and from 4.04 to 2.89, respectively). In the same time period, scores for the waiting list control groups remained flat (from 4.08 at baseline to 3.99).

At the end of treatment and at the 60-week follow-up, patients in both forms of CBT-E showed significant improvement across all measures, with no significant difference between treatments. By the end of treatment, 66.4% of those who completed all of the CBT sessions had global EDE scores <1.74 (considered clinically significant).

Subgroup analysis offers opportunity for fine-tuning

When analyzed separately, the patients with bulimia and those with eating disorder NOS did equally well at the end of treatment: 52.7% of those with bulimia and 53.3% of those with eating disorder NOS had global EDE scores <1.74. At the 60-week follow-up, the patients with bulimia maintained their improvement slightly more: 61.4% had global EDE scores <1.74, compared with 45.7% of the patients with eating disorder NOS.

The researchers also compared the outcomes of patients with the most complex additional psychopathology with those of patients with less complex problems. Greater complexity was defined as moderate ratings in at least 2 of the following domains: mood intolerance, clinical perfectionism, low self-esteem, and interpersonal difficulties.

Broad focus more effective for high complexity. Overall, those in the more complex subgroup did not respond as well; 48% had global EDE scores <1.74, vs 60% of those in the less complex group. However, those in the more complex subgroup did better with the broad form of CBT (at 60-week follow up, 60% had scores <1.74 with CBT-Eb, compared with 40% in the CBT-Ef treatment arm), while the less complex subgroup did better with the more tightly focused CBT-Ef. ²

WHAT’S NEW: Evidence supports CBT for NOS diagnosis

The most recent (2004) Cochrane review of “psychotherapy for bulimia nervosa and binging” included 40 RCTs of patients with bulimia, binge-eating disorder, and eating disorder NOS with recurrent binge-eating episodes (included in 7 studies). While the review confirmed that CBT is effective for bulimia and “similar syndromes,” it identified a need for larger and higher quality trials of CBT, particularly in patients with eating disorder NOS.¹ The study reviewed in this PURL—the first large, high-quality trial to include a number of patients with eating disorder NOS—provides strong evidence that CBT is effective for this group of patients.²

CAVEATS: Limited wait time leaves unanswered questions

One limitation of this study is the lack of a control group beyond the 8-week waiting period. Prior studies of CBT for bulimia that delayed therapy for those in the control groups for a longer duration have consistently shown that patients receiving CBT did significantly better than those in the control group.⁹ While a “no treatment” group would have made the results more robust in this case, it would not have been ethical to withhold treatment for the entire length of the study.

It is noteworthy, too, that this study only included patients with a BMI >17.5. Patients with a diagnosis of anorexia nervosa, who by definition have a lower BMI, will need other treatments, including hospitalization in some cases.

CHALLENGES TO IMPLEMENTATION: Identifying patients and therapists

The primary challenge is to determine which of your patients have eating disorders. When discussing diet, adding a simple question such as, “Are you happy with your current weight?” can help you identify those who meet the criteria for an eating disorder or are at high risk.³

Identifying local mental health providers who are trained to provide CBT for patients with eating disorders is another concern. Insurance coverage for this intensive treatment may also be a limiting factor in some cases.

Many studies support the use of fluoxetine for patients with bulimia, and combined treatment with SSRIs and CBT has been shown to be superior to either treatment alone.^8,10,14 Consider starting the patient on an antidepressant while she (or he) awaits the start of CBT.

Acknowledgements

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURLs) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old patient with a body mass index (BMI) of 18 tells you she’s fat and she’s afraid of gaining weight. Further questioning reveals that your patient binges on cookies and potato chips about once a week, then compensates for overeating by taking laxatives or exercising excessively—a practice she’s been following since she started college several years ago. The eating disorder she describes does not meet the criteria in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) for bulimia or anorexia nervosa, although it has elements of both. Rather, it fits the diagnostic criteria for eating disorder NOS. You’re aware that CBT is the first-line behavioral treatment for bulimia, and wonder whether it would be helpful for your patient.

Eating disorders often go undetected and untreated in primary care practices,³ as many patients don’t volunteer information about their weight or behaviors related to food, and physicians often fail to ask. Overall, as few as 10% of those with eating disorders receive any form of treatment.¹

Would you recognize this loosely defined disorder?

In the United States, the lifetime prevalence of eating disorders is 0.6% for anorexia nervosa (0.3% for men and 0.9% for women), 1.0% for bulimia (0.5% for men and 1.5% for women), and 2.8% for binge-eating disorder (2.0% for men, 3.5% for women).⁴ Eating disorder NOS, which encompasses subthreshold cases of anorexia or bulimia, patients with elements of both anorexia and bulimia, and patients with binge-eating disorder, accounts for 50% to 80% of eating disorder diagnoses in outpatient settings. Yet there have been few studies of the treatment of these patients.^2,5,6

A review of DSM-IV criteria

The diagnostic criteria for anorexia nervosa include a refusal to maintain a weight of at least 85% of normal body weight (or having a BMI ≤17.5), intense fear of gaining weight, disturbance in the way one’s body shape is experienced, and amenorrhea in females who are post-menarche.

Criteria for bulimia include recurrent episodes of binge eating (consuming a large amount of food with a sense of lack of control over eating) and recurrent inappropriate compensatory behaviors to prevent weight gain (self-induced vomiting, excessive exercise, fasting, laxatives, diuretics, or enemas) at least twice weekly for 3 months; and self-evaluation that is unduly influenced by body shape and weight.⁷ Most patients with eating disorder NOS have clinical features of both anorexia and bulimia.⁶

APA guidelines are silent on NOS

CBT has consistently proven to be the most useful behavioral treatment for patients with bulimia.¹ Selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine—the only medication with Food and Drug Administration approval for the treatment of an eating disorder⁸ —are about as effective as CBT, and the combination of CBT and an SSRI is superior to either treatment alone.⁹ CBT has also been found to be somewhat effective in treating binge-eating disorder.¹⁰

Anorexia nervosa, the most deadly eating disorder (the mortality rate is 6.6%¹¹ ) and the most difficult to treat, is the exception. Several studies have assessed CBT for treating anorexia, but it has not been found to be very effective.^10,12,13

The 2006 American Psychiatric Association practice guidelines for the treatment of patients with eating disorders feature recommendations for anorexia, bulimia, and binge-eating disorder, but do not address eating disorder NOS.¹⁰ The National Institute for Clinical Excellence (NICE) in the United Kingdom issued guidelines for the treatment of eating disorders in 2004. In response to the lack of evidence for treating eating disorder NOS, NICE recommended basing treatment on the form of eating disorder that most closely resembles the patient’s presentation.¹⁴ Fairburn et al addressed the lack of evidence for treatment of eating disorder NOS with the study summarized here.

STUDY SUMMARY: Both broad and focused CBT delivered results

Conducted at 2 eating disorder centers in the United Kingdom, this RCT included 154 patients, 18 to 65 years of age, who met DSM-IV criteria for either bulimia or eating disorder NOS. Exclusion criteria included prior treatment with CBT or other evidence-based treatment for the same eating disorder, and a BMI ≤17.5.

Most of the patients were female (95.5%) and white (90.3%), with a median age of 26 years and a median duration of eating disorder of 8.6 years. Sixty-two percent of the patients had a diagnosis of eating disorder NOS, and 38% were diagnosed with bulimia. Half of the patients had another current psychiatric diagnosis—a major depressive disorder, an anxiety disorder, or substance abuse.

The patients were randomized into 4 groups: Two received immediate treatment, and the other 2, referred to as waiting list controls, waited 8 weeks before beginning treatment. Treatment consisted of 1 of 2 forms of CBT-E, an enhanced form of CBT used to treat adult outpatients with eating disorders. Patients either received CBT-Ef, a focused form of CBT that exclusively targets eating disorder psychopathology, or CBT-Eb, a broader form of therapy that also addresses other problems that are common in patients with eating disorders, such as perfectionism and low self-esteem.

Both types of CBT-E featured a 90-minute preparatory session, 20 50-minute sessions, and 1 review session 20 weeks after completion of treatment. In the first 4 sessions, CBT-Ef and CBT-Eb were identical—addressing the eating disorder exclusively. CBT-Ef continued to focus on the eating disorder for the rest of the sessions, while subsequent CBT-Eb sessions also dealt with mood intolerance, interpersonal difficulties, and related issues. Five therapists—4 psychologists and 1 nurse-therapist—conducted the treatments.

Patients were weaned from ongoing psychiatric therapy during the study, but those who were on antidepressant therapy (n=76) were able to continue it. Patients were assessed before treatment, at the end of the waiting period for those in the control groups, after 8 weeks of treatment, at the end of treatment, and 20, 40, and 60 weeks after completion of treatment. (Twenty-two percent of the enrollees did not complete treatment.)

Primary outcomes were based on the Eating Disorder Examination (EDE), administered by assessors who were not involved in the treatment and were blinded to the patients’ group assignment. Change in severity of eating disorder features was measured by the global EDE score (0-6) and attaining a global EDE score <1.74 (<1 standard deviation above the community mean).

No treatment vs CBT. The waiting period left little doubt of the short-term efficacy of CBT: After 8 weeks, there was significant improvement in eating disorder behaviors and overall severity in both the CBT-Ef and CBT-Eb groups (EDE scores fell from 4.15 at baseline to 3.26 and from 4.04 to 2.89, respectively). In the same time period, scores for the waiting list control groups remained flat (from 4.08 at baseline to 3.99).

At the end of treatment and at the 60-week follow-up, patients in both forms of CBT-E showed significant improvement across all measures, with no significant difference between treatments. By the end of treatment, 66.4% of those who completed all of the CBT sessions had global EDE scores <1.74 (considered clinically significant).

Subgroup analysis offers opportunity for fine-tuning

When analyzed separately, the patients with bulimia and those with eating disorder NOS did equally well at the end of treatment: 52.7% of those with bulimia and 53.3% of those with eating disorder NOS had global EDE scores <1.74. At the 60-week follow-up, the patients with bulimia maintained their improvement slightly more: 61.4% had global EDE scores <1.74, compared with 45.7% of the patients with eating disorder NOS.

The researchers also compared the outcomes of patients with the most complex additional psychopathology with those of patients with less complex problems. Greater complexity was defined as moderate ratings in at least 2 of the following domains: mood intolerance, clinical perfectionism, low self-esteem, and interpersonal difficulties.

Broad focus more effective for high complexity. Overall, those in the more complex subgroup did not respond as well; 48% had global EDE scores <1.74, vs 60% of those in the less complex group. However, those in the more complex subgroup did better with the broad form of CBT (at 60-week follow up, 60% had scores <1.74 with CBT-Eb, compared with 40% in the CBT-Ef treatment arm), while the less complex subgroup did better with the more tightly focused CBT-Ef. ²

WHAT’S NEW: Evidence supports CBT for NOS diagnosis

The most recent (2004) Cochrane review of “psychotherapy for bulimia nervosa and binging” included 40 RCTs of patients with bulimia, binge-eating disorder, and eating disorder NOS with recurrent binge-eating episodes (included in 7 studies). While the review confirmed that CBT is effective for bulimia and “similar syndromes,” it identified a need for larger and higher quality trials of CBT, particularly in patients with eating disorder NOS.¹ The study reviewed in this PURL—the first large, high-quality trial to include a number of patients with eating disorder NOS—provides strong evidence that CBT is effective for this group of patients.²

CAVEATS: Limited wait time leaves unanswered questions

One limitation of this study is the lack of a control group beyond the 8-week waiting period. Prior studies of CBT for bulimia that delayed therapy for those in the control groups for a longer duration have consistently shown that patients receiving CBT did significantly better than those in the control group.⁹ While a “no treatment” group would have made the results more robust in this case, it would not have been ethical to withhold treatment for the entire length of the study.

It is noteworthy, too, that this study only included patients with a BMI >17.5. Patients with a diagnosis of anorexia nervosa, who by definition have a lower BMI, will need other treatments, including hospitalization in some cases.

CHALLENGES TO IMPLEMENTATION: Identifying patients and therapists

The primary challenge is to determine which of your patients have eating disorders. When discussing diet, adding a simple question such as, “Are you happy with your current weight?” can help you identify those who meet the criteria for an eating disorder or are at high risk.³

Identifying local mental health providers who are trained to provide CBT for patients with eating disorders is another concern. Insurance coverage for this intensive treatment may also be a limiting factor in some cases.

Many studies support the use of fluoxetine for patients with bulimia, and combined treatment with SSRIs and CBT has been shown to be superior to either treatment alone.^8,10,14 Consider starting the patient on an antidepressant while she (or he) awaits the start of CBT.

Acknowledgements

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURLs) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Lack of evidence for weight loss drugs

The most recent guideline on this topic does not recommend any medication, citing a lack of evidence. In its 2003 consensus statement, a panel representing the American Diabetes Association, American Psychiatric Association, American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity³ recommends:

• That patients taking second-generation antipsychotics have the following assessments at baseline and regular intervals: weight, height, waist circumference, blood pressure, fasting plasma glucose, and fasting lipids.
• Providing nutrition and exercise counseling to all patients who are over-weight or obese at baseline.
• Initiating treatment with one of the second-generation antipsychotics with a lower risk of weight gain for patients at high risk of diabetes (ie, family history) and for patients who gain 5% or more of their initial weight or develop worsening hyperglycemia or dyslipidemia during treatment.

This guideline does not recommend metformin to reduce weight gain.

A 2007 Cochrane review of interventions to reduce weight gain in patients with schizophrenia included 23 randomized controlled trials of a variety of weight loss interventions, including cognitive/behavioral interventions and a variety of medications, including sibutramine, orlistat, fluoxetine, topiramate, and metformin. The authors highlighted the limited number of studies of short duration and with small sample sizes and concluded that the evidence was insufficient for the use of pharmacologic interventions to prevent or treat weight gain.⁵

STUDY SUMMARY: Lifestyle changes and metformin compared

This randomized controlled trial was conducted in China and included 128 adults aged 18 to 45 with a first psychotic episode of schizophrenia. All patients had to have gained more than 10% of their pretreatment body weight during the first year of treatment with an antipsychotic medication (clozapine, olanzapine, risperidone, or sulpiride [not approved for use in the United States]). All study participants had to be under the care of an adult caregiver who monitored and recorded food intake, exercise, and medication intake. Patients with diabetes, cardiovascular disease, liver or renal dysfunction, substance abuse, or psychiatric diagnoses other than schizophrenia were excluded.

Patients were randomized to 1 of 4 groups for the 12 weeks of the study:

• Metformin alone, 250 mg 3 times daily
• Placebo alone
• Lifestyle intervention plus metformin
• Lifestyle intervention plus placebo

The lifestyle intervention included 3 components: (1) education: monthly programs on nutrition and physical activity; (2) diet: the American Heart Association step 2 diet (<30% calories from fat, 55% carbohydrates, >15% protein, with at least 15 g fiber per 1000 kcal); and (3) exercise: 1 week of sessions with an exercise physiologist followed by an individualized home-based exercise program.

Primary outcomes included changes in weight, body mass index (BMI), waist circumference, and fasting glucose ( TABLE 2 ). Ten of the 128 randomized patients either discontinued the study or were lost to follow up, but all 128 patients were included in the analysis.

TABLE 2
Mean difference between baseline and endpoint (week 12) of treatment outcomes (95% confidence intervals)¹

Best result: Lifestyle changes plus metformin

Compared with baseline, weight decreased by 7.3% in the lifestyle plus metformin group, by 4.9% in the metformin-only group, and by 2.2% in the lifestyle-only group; in the placebo group, weight increased by 4.8%.

Participants in all 3 intervention groups also showed significant decreases in the mean fasting glucose, insulin levels, and insulin resistance index (IRI). The insulin levels and the IRI increased in the placebo group.

No significant differences in adverse effects were noted among the 4 treatment groups.¹

WHAT’S NEW: Convincing evidence

This is the first randomized controlled trial to show convincingly that metformin alone or in combination with lifestyle changes is superior to lifestyle changes alone or placebo for reducing weight gain and other adverse metabolic outcomes induced by second-generation antipsychotics.

Intensive lifestyle interventions

Prior studies found that intensive lifestyle interventions can help reduce antipsychotic-related weight gain. A 3-month randomized controlled trial compared an early behavioral intervention (dietary counseling, an exercise program, and behavior therapy) with routine care in 61 patients with first-episode psychosis who were taking risperidone, olanzapine, or haloperidol;⁶ significantly fewer patients assigned to behavioral intervention had an increased initial body weight of more than 7%: 39% in the behavioral intervention group vs 79% in the routine care group (P<.002).

Small samples, small effect sizes

Past studies of metformin for antipsychotic-associated weight gain have generally shown a small benefit, though small sample sizes and small effect sizes prohibited definitive conclusions. Unlike the study by Wu and colleagues,¹ none of these past studies were designed to compare the combination of metformin and lifestyle intervention with metformin alone, lifestyle intervention alone, or placebo alone.

Klein et al conducted a randomized placebo-controlled trial of metformin in 39 children ages 10 to 17 whose weight had increased more than 10% on atypical antipsychotic therapy.⁷ The children treated with placebo gained a mean of 4 kg and increased their mean BMI by 1.12 kg/m² during 16 weeks of treatment, while those in the metformin group did not gain weight and decreased their mean BMI by 0.43 kg/m².

Baptista et al randomized 40 in-patients with schizophrenia, who were being switched from conventional antipsychotics to olanzapine, to either metformin (850-1750 mg/d) or placebo. Both groups gained a similar amount of weight after the 14-week study (5.5 vs 6.3 kg, metformin vs placebo). Three patients who started with high fasting glucose had decreases while taking metformin, and 3 patients given placebo developed elevated fasting glucose during the study.⁸

In another randomized controlled trial of metformin vs placebo in 80 patients who had been taking olanzapine for at least 4 months, Baptista et al found only a small, insignificant difference in weight loss after 12 weeks of treatment (metformin group lost 1.4 kg, placebo group lost 0.18 kg, P=.09). They reported that both groups were highly motivated to lose weight and were compliant with the healthy lifestyle recommendations.⁹

An adequately powered study

The trial¹ highlighted in this PURL had an adequate sample size to compare metformin plus a lifestyle intervention with either treatment alone or placebo. It showed a clinically important effect of metformin both by itself and in conjunction with the lifestyle intervention.

CAVEATS: Consider switching drugs

Before adding metformin to help with weight loss, primary care clinicians should contact the patient’s psychiatrist to discuss the option of switching antipsychotic medications. Switching from a medication with a higher risk for weight gain, such as olanzapine, to one with a lower risk, such as aripiprazole or ziprasidone, can lead to significant weight loss.¹⁰

Not an option for some

However, some patients, especially those taking clozapine, may have already tried multiple antipsychotic agents without success, and switching is not an option for them.

Prescribing metformin

CHALLENGES TO IMPLEMENTATION: Adherence

These study participants were under the care of an adult caregiver who monitored and recorded food and medication intake and exercise level. The lifestyle intervention was thorough and structured and this kind of program is often not available to us for our patients. As a consequence, we may not obtain the same results as in this study. However, even the metformin-alone group showed improvements, and if our patients can reliably take their second-generation antipsychotic, they should also be able to take metformin reliably.

Patient resistance

Some patients may resist taking an additional medication to treat the side effects of their antipsychotic medication. Taking the time to educate them about the increased risk of diabetes and cardiovascular disease related to weight gain may help convince them to do so. Warn them about possible gastrointestinal adverse effects of metformin, which tend to lessen or disappear with time.

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Lack of evidence for weight loss drugs

The most recent guideline on this topic does not recommend any medication, citing a lack of evidence. In its 2003 consensus statement, a panel representing the American Diabetes Association, American Psychiatric Association, American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity³ recommends:

• That patients taking second-generation antipsychotics have the following assessments at baseline and regular intervals: weight, height, waist circumference, blood pressure, fasting plasma glucose, and fasting lipids.
• Providing nutrition and exercise counseling to all patients who are over-weight or obese at baseline.
• Initiating treatment with one of the second-generation antipsychotics with a lower risk of weight gain for patients at high risk of diabetes (ie, family history) and for patients who gain 5% or more of their initial weight or develop worsening hyperglycemia or dyslipidemia during treatment.

This guideline does not recommend metformin to reduce weight gain.

A 2007 Cochrane review of interventions to reduce weight gain in patients with schizophrenia included 23 randomized controlled trials of a variety of weight loss interventions, including cognitive/behavioral interventions and a variety of medications, including sibutramine, orlistat, fluoxetine, topiramate, and metformin. The authors highlighted the limited number of studies of short duration and with small sample sizes and concluded that the evidence was insufficient for the use of pharmacologic interventions to prevent or treat weight gain.⁵

STUDY SUMMARY: Lifestyle changes and metformin compared

This randomized controlled trial was conducted in China and included 128 adults aged 18 to 45 with a first psychotic episode of schizophrenia. All patients had to have gained more than 10% of their pretreatment body weight during the first year of treatment with an antipsychotic medication (clozapine, olanzapine, risperidone, or sulpiride [not approved for use in the United States]). All study participants had to be under the care of an adult caregiver who monitored and recorded food intake, exercise, and medication intake. Patients with diabetes, cardiovascular disease, liver or renal dysfunction, substance abuse, or psychiatric diagnoses other than schizophrenia were excluded.

Patients were randomized to 1 of 4 groups for the 12 weeks of the study:

• Metformin alone, 250 mg 3 times daily
• Placebo alone
• Lifestyle intervention plus metformin
• Lifestyle intervention plus placebo

The lifestyle intervention included 3 components: (1) education: monthly programs on nutrition and physical activity; (2) diet: the American Heart Association step 2 diet (<30% calories from fat, 55% carbohydrates, >15% protein, with at least 15 g fiber per 1000 kcal); and (3) exercise: 1 week of sessions with an exercise physiologist followed by an individualized home-based exercise program.

Primary outcomes included changes in weight, body mass index (BMI), waist circumference, and fasting glucose ( TABLE 2 ). Ten of the 128 randomized patients either discontinued the study or were lost to follow up, but all 128 patients were included in the analysis.

TABLE 2
Mean difference between baseline and endpoint (week 12) of treatment outcomes (95% confidence intervals)¹

Best result: Lifestyle changes plus metformin

Compared with baseline, weight decreased by 7.3% in the lifestyle plus metformin group, by 4.9% in the metformin-only group, and by 2.2% in the lifestyle-only group; in the placebo group, weight increased by 4.8%.

Participants in all 3 intervention groups also showed significant decreases in the mean fasting glucose, insulin levels, and insulin resistance index (IRI). The insulin levels and the IRI increased in the placebo group.

No significant differences in adverse effects were noted among the 4 treatment groups.¹

WHAT’S NEW: Convincing evidence

This is the first randomized controlled trial to show convincingly that metformin alone or in combination with lifestyle changes is superior to lifestyle changes alone or placebo for reducing weight gain and other adverse metabolic outcomes induced by second-generation antipsychotics.

Intensive lifestyle interventions

Prior studies found that intensive lifestyle interventions can help reduce antipsychotic-related weight gain. A 3-month randomized controlled trial compared an early behavioral intervention (dietary counseling, an exercise program, and behavior therapy) with routine care in 61 patients with first-episode psychosis who were taking risperidone, olanzapine, or haloperidol;⁶ significantly fewer patients assigned to behavioral intervention had an increased initial body weight of more than 7%: 39% in the behavioral intervention group vs 79% in the routine care group (P<.002).

Small samples, small effect sizes

Past studies of metformin for antipsychotic-associated weight gain have generally shown a small benefit, though small sample sizes and small effect sizes prohibited definitive conclusions. Unlike the study by Wu and colleagues,¹ none of these past studies were designed to compare the combination of metformin and lifestyle intervention with metformin alone, lifestyle intervention alone, or placebo alone.

Klein et al conducted a randomized placebo-controlled trial of metformin in 39 children ages 10 to 17 whose weight had increased more than 10% on atypical antipsychotic therapy.⁷ The children treated with placebo gained a mean of 4 kg and increased their mean BMI by 1.12 kg/m² during 16 weeks of treatment, while those in the metformin group did not gain weight and decreased their mean BMI by 0.43 kg/m².

Baptista et al randomized 40 in-patients with schizophrenia, who were being switched from conventional antipsychotics to olanzapine, to either metformin (850-1750 mg/d) or placebo. Both groups gained a similar amount of weight after the 14-week study (5.5 vs 6.3 kg, metformin vs placebo). Three patients who started with high fasting glucose had decreases while taking metformin, and 3 patients given placebo developed elevated fasting glucose during the study.⁸

In another randomized controlled trial of metformin vs placebo in 80 patients who had been taking olanzapine for at least 4 months, Baptista et al found only a small, insignificant difference in weight loss after 12 weeks of treatment (metformin group lost 1.4 kg, placebo group lost 0.18 kg, P=.09). They reported that both groups were highly motivated to lose weight and were compliant with the healthy lifestyle recommendations.⁹

An adequately powered study

The trial¹ highlighted in this PURL had an adequate sample size to compare metformin plus a lifestyle intervention with either treatment alone or placebo. It showed a clinically important effect of metformin both by itself and in conjunction with the lifestyle intervention.

CAVEATS: Consider switching drugs

Before adding metformin to help with weight loss, primary care clinicians should contact the patient’s psychiatrist to discuss the option of switching antipsychotic medications. Switching from a medication with a higher risk for weight gain, such as olanzapine, to one with a lower risk, such as aripiprazole or ziprasidone, can lead to significant weight loss.¹⁰

Not an option for some

However, some patients, especially those taking clozapine, may have already tried multiple antipsychotic agents without success, and switching is not an option for them.

Prescribing metformin

CHALLENGES TO IMPLEMENTATION: Adherence

These study participants were under the care of an adult caregiver who monitored and recorded food and medication intake and exercise level. The lifestyle intervention was thorough and structured and this kind of program is often not available to us for our patients. As a consequence, we may not obtain the same results as in this study. However, even the metformin-alone group showed improvements, and if our patients can reliably take their second-generation antipsychotic, they should also be able to take metformin reliably.

Patient resistance

Some patients may resist taking an additional medication to treat the side effects of their antipsychotic medication. Taking the time to educate them about the increased risk of diabetes and cardiovascular disease related to weight gain may help convince them to do so. Warn them about possible gastrointestinal adverse effects of metformin, which tend to lessen or disappear with time.

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Lack of evidence for weight loss drugs

The most recent guideline on this topic does not recommend any medication, citing a lack of evidence. In its 2003 consensus statement, a panel representing the American Diabetes Association, American Psychiatric Association, American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity³ recommends:

• That patients taking second-generation antipsychotics have the following assessments at baseline and regular intervals: weight, height, waist circumference, blood pressure, fasting plasma glucose, and fasting lipids.
• Providing nutrition and exercise counseling to all patients who are over-weight or obese at baseline.
• Initiating treatment with one of the second-generation antipsychotics with a lower risk of weight gain for patients at high risk of diabetes (ie, family history) and for patients who gain 5% or more of their initial weight or develop worsening hyperglycemia or dyslipidemia during treatment.

This guideline does not recommend metformin to reduce weight gain.

A 2007 Cochrane review of interventions to reduce weight gain in patients with schizophrenia included 23 randomized controlled trials of a variety of weight loss interventions, including cognitive/behavioral interventions and a variety of medications, including sibutramine, orlistat, fluoxetine, topiramate, and metformin. The authors highlighted the limited number of studies of short duration and with small sample sizes and concluded that the evidence was insufficient for the use of pharmacologic interventions to prevent or treat weight gain.⁵

STUDY SUMMARY: Lifestyle changes and metformin compared

This randomized controlled trial was conducted in China and included 128 adults aged 18 to 45 with a first psychotic episode of schizophrenia. All patients had to have gained more than 10% of their pretreatment body weight during the first year of treatment with an antipsychotic medication (clozapine, olanzapine, risperidone, or sulpiride [not approved for use in the United States]). All study participants had to be under the care of an adult caregiver who monitored and recorded food intake, exercise, and medication intake. Patients with diabetes, cardiovascular disease, liver or renal dysfunction, substance abuse, or psychiatric diagnoses other than schizophrenia were excluded.

Patients were randomized to 1 of 4 groups for the 12 weeks of the study:

• Metformin alone, 250 mg 3 times daily
• Placebo alone
• Lifestyle intervention plus metformin
• Lifestyle intervention plus placebo

The lifestyle intervention included 3 components: (1) education: monthly programs on nutrition and physical activity; (2) diet: the American Heart Association step 2 diet (<30% calories from fat, 55% carbohydrates, >15% protein, with at least 15 g fiber per 1000 kcal); and (3) exercise: 1 week of sessions with an exercise physiologist followed by an individualized home-based exercise program.

Primary outcomes included changes in weight, body mass index (BMI), waist circumference, and fasting glucose ( TABLE 2 ). Ten of the 128 randomized patients either discontinued the study or were lost to follow up, but all 128 patients were included in the analysis.

TABLE 2
Mean difference between baseline and endpoint (week 12) of treatment outcomes (95% confidence intervals)¹

Best result: Lifestyle changes plus metformin

Compared with baseline, weight decreased by 7.3% in the lifestyle plus metformin group, by 4.9% in the metformin-only group, and by 2.2% in the lifestyle-only group; in the placebo group, weight increased by 4.8%.

Participants in all 3 intervention groups also showed significant decreases in the mean fasting glucose, insulin levels, and insulin resistance index (IRI). The insulin levels and the IRI increased in the placebo group.

No significant differences in adverse effects were noted among the 4 treatment groups.¹

WHAT’S NEW: Convincing evidence

This is the first randomized controlled trial to show convincingly that metformin alone or in combination with lifestyle changes is superior to lifestyle changes alone or placebo for reducing weight gain and other adverse metabolic outcomes induced by second-generation antipsychotics.

Intensive lifestyle interventions

Prior studies found that intensive lifestyle interventions can help reduce antipsychotic-related weight gain. A 3-month randomized controlled trial compared an early behavioral intervention (dietary counseling, an exercise program, and behavior therapy) with routine care in 61 patients with first-episode psychosis who were taking risperidone, olanzapine, or haloperidol;⁶ significantly fewer patients assigned to behavioral intervention had an increased initial body weight of more than 7%: 39% in the behavioral intervention group vs 79% in the routine care group (P<.002).

Small samples, small effect sizes

Past studies of metformin for antipsychotic-associated weight gain have generally shown a small benefit, though small sample sizes and small effect sizes prohibited definitive conclusions. Unlike the study by Wu and colleagues,¹ none of these past studies were designed to compare the combination of metformin and lifestyle intervention with metformin alone, lifestyle intervention alone, or placebo alone.

Klein et al conducted a randomized placebo-controlled trial of metformin in 39 children ages 10 to 17 whose weight had increased more than 10% on atypical antipsychotic therapy.⁷ The children treated with placebo gained a mean of 4 kg and increased their mean BMI by 1.12 kg/m² during 16 weeks of treatment, while those in the metformin group did not gain weight and decreased their mean BMI by 0.43 kg/m².

Baptista et al randomized 40 in-patients with schizophrenia, who were being switched from conventional antipsychotics to olanzapine, to either metformin (850-1750 mg/d) or placebo. Both groups gained a similar amount of weight after the 14-week study (5.5 vs 6.3 kg, metformin vs placebo). Three patients who started with high fasting glucose had decreases while taking metformin, and 3 patients given placebo developed elevated fasting glucose during the study.⁸

In another randomized controlled trial of metformin vs placebo in 80 patients who had been taking olanzapine for at least 4 months, Baptista et al found only a small, insignificant difference in weight loss after 12 weeks of treatment (metformin group lost 1.4 kg, placebo group lost 0.18 kg, P=.09). They reported that both groups were highly motivated to lose weight and were compliant with the healthy lifestyle recommendations.⁹

An adequately powered study

The trial¹ highlighted in this PURL had an adequate sample size to compare metformin plus a lifestyle intervention with either treatment alone or placebo. It showed a clinically important effect of metformin both by itself and in conjunction with the lifestyle intervention.

CAVEATS: Consider switching drugs

Before adding metformin to help with weight loss, primary care clinicians should contact the patient’s psychiatrist to discuss the option of switching antipsychotic medications. Switching from a medication with a higher risk for weight gain, such as olanzapine, to one with a lower risk, such as aripiprazole or ziprasidone, can lead to significant weight loss.¹⁰

Not an option for some

However, some patients, especially those taking clozapine, may have already tried multiple antipsychotic agents without success, and switching is not an option for them.

Prescribing metformin

CHALLENGES TO IMPLEMENTATION: Adherence

These study participants were under the care of an adult caregiver who monitored and recorded food and medication intake and exercise level. The lifestyle intervention was thorough and structured and this kind of program is often not available to us for our patients. As a consequence, we may not obtain the same results as in this study. However, even the metformin-alone group showed improvements, and if our patients can reliably take their second-generation antipsychotic, they should also be able to take metformin reliably.

Patient resistance

Some patients may resist taking an additional medication to treat the side effects of their antipsychotic medication. Taking the time to educate them about the increased risk of diabetes and cardiovascular disease related to weight gain may help convince them to do so. Warn them about possible gastrointestinal adverse effects of metformin, which tend to lessen or disappear with time.

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

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old woman presents to your office with a 1-week history of cough, purulent nasal discharge, and unilateral facial pain. You diagnose acute sinusitis.

Should you prescribe an antibiotic?

No. Yet it’s no wonder that most adults treated for acute sinusitis leave the doctor’s office with a prescription for antibiotics. Until the publication of the meta-analysis by Young and colleagues¹ featured in this PURL, we have lacked A-level evidence from studies conducted in realistic settings—like your practice and ours.

Review of serial data from the National Ambulatory Medical Care Surveys (NAMCS) from 1999 through 2005 does show a slight downward trend in antibiotic prescribing for acute sinusitis: 1999-2002 data showed that 83% of cases of acute sinusitis were treated with an antibiotic.² Data from the 2004 and 2005 NAMCS reveal that family physicians prescribed antibiotics for 80% of patients with acute sinusitis in 2004 and 76% of patients in 2005 (S. Medvedev, unpublished data, NAMCS database, March 2008).

Is this continued high rate of antibiotic prescribing justified?

Do antibiotics improve symptoms and shorten the duration of illness or not?

These questions are important, obviously, not only because of the high rate of prescribing but also because sinusitis is one of the most common diagnoses: approximately 20 million cases annually in the United States, or about 21% of all outpatient antibiotic prescriptions for adults.²

Which patients might benefit from antibiotics?

Common clinical signs and symptoms cannot identify patients with rhinosinusitis for whom treatment is clearly justified, given the cost, adverse events, and bacterial resistance associated with antibiotic use

• Severity of symptoms is important only in that signs suggestive of a serious complication are the sole reason for immediate antibiotic treatment
• Purulent discharge noted in the pharynx on exam was associated with a higher likelihood of benefit from antibiotics, but NNT was 8
• Antibiotics are not justified even if a patient reports having symptoms for longer than 7-10 days

Source: Young et al.¹

A diagnostic dilemma

Before we discuss the evidence that is summarized in the excellent meta-analysis by Young and colleagues,¹ let’s acknowledge that acute sinusitis is undeniably a diagnostic dilemma. Distinguishing bacterial from viral infection is nearly impossible on clinical grounds because the symptoms are so similar. A litany of identical upper respiratory symptoms accompanies both viral and bacterial sinus infections. Purulent nasal secretions, maxillary facial pain (especially with unilateral predominance), maxillary tooth pain (which is uncommon with sinus infection), altered sense of smell, and worsening illness after improvement constitute the short list of signs and symptoms that has some predictive value, but even the presence of all of these is not a terrific predictor of bacterial sinus infection. Plain x-rays have low accuracy in distinguishing viral from bacterial infection. Computed tomography (CT) sinus scans are better but far from perfect, are not readily available in practice, and are expensive.

Sinusitis in the real world

How effective are antibiotics for patients diagnosed not by sinus x-rays or CTs, but by signs and symptoms—as we typically do in daily practice?

A meta-analysis³ of 13 randomized controlled trials (RCTs) found that sinusitis improved without antibiotics, but it included trials in which patients were recruited based on results of imaging studies and cultures, which are not normally used in primary care clinical practice. That study compared antibiotic treatment to placebo for acute uncomplicated sinusitis; 35% of placebo-treated patients were clinically cured by 7 to 12 days and 73% were improved after 7 days. Antibiotic therapy increased cure rates by 15% and improvement rates by 14%, yielding a number needed to treat of 7 to achieve 1 additional positive outcome at 7 days.

STUDY SUMMARY: Meta-analysis of primary care trials

Young and colleagues¹ aggregated and analyzed individual patient-level data from all known placebo-controlled, randomized, antibiotic treatment trials of adults with clinical symptoms of acute sinusitis that were conducted in primary care settings. They excluded trials that used imaging or bacterial culture as part of patient recruitment.

Studies were included that allowed the use of concomitant medication such as nonsteroidal anti-inflammatory drugs, decongestants, or nasal steroids, as long as patients in both groups had access to the same medications. All trials excluded patients with severe symptoms such as high fever, periorbital swelling or erythema, or intense facial pain, important exclusions that we will discuss below.

They identified 10 such studies and completed an intent-to-treat analysis of the 9 double-blind trials for which patient level data were available. Using individual data from 2547 patients, the odds ratio for an overall antibiotic treatment effect was 1.37 (95% confidence interval, 1.13-1.66), with a number needed to treat (NNT) of 15.

This finding means that 15 patients needed to be given an antibiotic for 1 additional patient to be cured at 8 to 15 days after treatment commenced. Using statistical modeling, they determined that 64% of patients treated with placebo were cured at 14 days compared with 70% given an antibiotic. One patient out of 1381 treated with placebo experienced a serious complication, a brain abscess.

Do antibiotics benefit any subgroups?

The investigators also analyzed the prognostic value of specific signs and symptoms to answer the question: Is there any subgroup of patients who might benefit more from antibiotic treatment?

Duration. Patients with a longer duration of symptoms, more severe symptoms, or increased age took longer to cure, but were no more likely to benefit from antibiotic treatment than other patients.

Symptoms, such as a previous common cold, pain on bending, unilateral facial pain, tooth pain, and purulent nasal discharge did not have any prognostic value.

Only one sign—purulent discharge noted in the pharynx on examination—was associated with a higher likelihood of benefit from treatment with antibiotics, but the NNT was still 8 in this group. Patients with symptoms for 7 days or longer were no more likely to respond to antibiotics than those with symptoms for fewer than 7 days.¹

WHAT’S NEW: Realistic evidence from realistic settings

We believe this meta-analysis provides a high level of evidence against routine treatment of sinusitis with antibiotics in primary care practice. Treating 15 patients with an antibiotic to possibly benefit 1 patient 2 weeks after treatment commences does not seem like a good idea when one considers the cost and complications of antibiotic use. Diarrhea and other adverse outcomes are 80% more common among patients with sinusitis who are treated with an antibiotic compared with placebo.³ As noted above, prior meta-analyses of antibiotic treatment for acute sinusitis have been more encouraging than this meta-analysis, with a number needed to treat of 7, but those meta-analyses are clearly overly optimistic for the results one will achieve in primary care practice using clinical signs and symptoms to diagnose acute sinusitis.^3,4 Unlike the Young study, they included trials in specialty clinics with CT scans and sinus puncture and culture used for the diagnostic standard.

Symptoms >1 week are not a reason to prescribe

One very important new finding in this meta-analysis that should change practice is that the duration of illness did not predict a positive response to antibiotics.

Current national recommendations are to use an antibiotic for patients with a duration of illness longer than 1 week, as these patients are presumably more likely to have a bacterial infection.^5-7 However, that recommendation had been based on expert opinion, not on data from clinical trials. A longer duration of symptoms should not be a reason to prescribe an antibiotic for sinusitis symptoms.

How can you help your patient?

What to do, then, for patients with acute sinusitis? Treat the symptoms, which means recommending pain medication for facial pain or headache and saline nasal spray for the nasal discharge, not antibiotics or nasal corticosteroids. Side effects will be fewer and costs will be lower.

• Saline irrigation. A 2007 Cochrane review of 8 chronic and recurrent sinusitis trials showed that nasal saline irrigation is effective for reducing symptoms of chronic and recurrent sinusitis.⁸ Although we do not have high-quality RCT data on saline nasal irrigation for treatment of acute sinusitis, nasal saline irrigation is harmless and inexpensive.
• What about nasal steroids? The evidence is equivocal, and the most recent high-quality RCT of nasal steroids showed no effect.⁹

CAVEATS: Refer seriously ill patients and complicated cases

A very important caveat to our recommendation is that seriously ill patients must be managed differently. Very infrequently a patient develops a serious complication of acute sinusitis such as brain abscess, periorbital cellulitis, or meningitis. Therefore, seriously ill patients with signs and symptoms of acute bacterial sinusitis, such as high fever, periorbital erythema or edema, severe headache, or intense facial pain must be carefully evaluated and treated with great caution and close follow-up. These patients should be referred immediately for consultation with an otolaryngologist.

Of course, mildly ill patients today may become quite ill tomorrow, so always provide advice to patients to return if they are getting worse, a good clinical practice for any condition that is usually benign but occasionally serious.

Patients who have prolonged or recurrent sinusitis symptoms need further evaluation for other diagnoses such as allergies, cystic fibrosis, fungal sinus infection, and other illnesses associated with immune compromise. These complicated patients benefit from consultation with an otolaryngologist who has a specific interest in chronic and recurrent sinusitis, and perhaps consultation from an infectious disease specialist as well.

CHALLENGES TO IMPLEMENTATION: The patient who wants a pill

Some patients may be accustomed to receiving an antibiotic prescription for their “sinus infections” and may resist conservative management. It may be difficult to convince them that antibiotics won’t make a difference when they attribute past resolution of symptoms to antibiotics.

Take enough time to educate your patients on the natural course of illness, the positive benefits of nasal saline, and the reasons not to use unnecessary antibiotics (eg, they are not effective, have potential adverse effects, and can contribute to future antibiotic resistance); this effort will save you time in future visits.¹⁰ A “just in case you don’t get better” prescription to be filled only if the patient is not improving in the next few days is about 50% effective in reducing antibiotic usage for upper respiratory infections.¹¹

Acknowledgement

We acknowledge Sofia Medvedev of the University HealthSystem Consortium (UHC) in Oak Brook, IL for analysis of the National Ambulatory Medical Care Survey data.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old woman presents to your office with a 1-week history of cough, purulent nasal discharge, and unilateral facial pain. You diagnose acute sinusitis.

Should you prescribe an antibiotic?

No. Yet it’s no wonder that most adults treated for acute sinusitis leave the doctor’s office with a prescription for antibiotics. Until the publication of the meta-analysis by Young and colleagues¹ featured in this PURL, we have lacked A-level evidence from studies conducted in realistic settings—like your practice and ours.

Review of serial data from the National Ambulatory Medical Care Surveys (NAMCS) from 1999 through 2005 does show a slight downward trend in antibiotic prescribing for acute sinusitis: 1999-2002 data showed that 83% of cases of acute sinusitis were treated with an antibiotic.² Data from the 2004 and 2005 NAMCS reveal that family physicians prescribed antibiotics for 80% of patients with acute sinusitis in 2004 and 76% of patients in 2005 (S. Medvedev, unpublished data, NAMCS database, March 2008).

Is this continued high rate of antibiotic prescribing justified?

Do antibiotics improve symptoms and shorten the duration of illness or not?

These questions are important, obviously, not only because of the high rate of prescribing but also because sinusitis is one of the most common diagnoses: approximately 20 million cases annually in the United States, or about 21% of all outpatient antibiotic prescriptions for adults.²

Which patients might benefit from antibiotics?

Common clinical signs and symptoms cannot identify patients with rhinosinusitis for whom treatment is clearly justified, given the cost, adverse events, and bacterial resistance associated with antibiotic use

• Severity of symptoms is important only in that signs suggestive of a serious complication are the sole reason for immediate antibiotic treatment
• Purulent discharge noted in the pharynx on exam was associated with a higher likelihood of benefit from antibiotics, but NNT was 8
• Antibiotics are not justified even if a patient reports having symptoms for longer than 7-10 days

Source: Young et al.¹

A diagnostic dilemma

Before we discuss the evidence that is summarized in the excellent meta-analysis by Young and colleagues,¹ let’s acknowledge that acute sinusitis is undeniably a diagnostic dilemma. Distinguishing bacterial from viral infection is nearly impossible on clinical grounds because the symptoms are so similar. A litany of identical upper respiratory symptoms accompanies both viral and bacterial sinus infections. Purulent nasal secretions, maxillary facial pain (especially with unilateral predominance), maxillary tooth pain (which is uncommon with sinus infection), altered sense of smell, and worsening illness after improvement constitute the short list of signs and symptoms that has some predictive value, but even the presence of all of these is not a terrific predictor of bacterial sinus infection. Plain x-rays have low accuracy in distinguishing viral from bacterial infection. Computed tomography (CT) sinus scans are better but far from perfect, are not readily available in practice, and are expensive.

Sinusitis in the real world

How effective are antibiotics for patients diagnosed not by sinus x-rays or CTs, but by signs and symptoms—as we typically do in daily practice?

A meta-analysis³ of 13 randomized controlled trials (RCTs) found that sinusitis improved without antibiotics, but it included trials in which patients were recruited based on results of imaging studies and cultures, which are not normally used in primary care clinical practice. That study compared antibiotic treatment to placebo for acute uncomplicated sinusitis; 35% of placebo-treated patients were clinically cured by 7 to 12 days and 73% were improved after 7 days. Antibiotic therapy increased cure rates by 15% and improvement rates by 14%, yielding a number needed to treat of 7 to achieve 1 additional positive outcome at 7 days.

STUDY SUMMARY: Meta-analysis of primary care trials

Young and colleagues¹ aggregated and analyzed individual patient-level data from all known placebo-controlled, randomized, antibiotic treatment trials of adults with clinical symptoms of acute sinusitis that were conducted in primary care settings. They excluded trials that used imaging or bacterial culture as part of patient recruitment.

Studies were included that allowed the use of concomitant medication such as nonsteroidal anti-inflammatory drugs, decongestants, or nasal steroids, as long as patients in both groups had access to the same medications. All trials excluded patients with severe symptoms such as high fever, periorbital swelling or erythema, or intense facial pain, important exclusions that we will discuss below.

They identified 10 such studies and completed an intent-to-treat analysis of the 9 double-blind trials for which patient level data were available. Using individual data from 2547 patients, the odds ratio for an overall antibiotic treatment effect was 1.37 (95% confidence interval, 1.13-1.66), with a number needed to treat (NNT) of 15.

This finding means that 15 patients needed to be given an antibiotic for 1 additional patient to be cured at 8 to 15 days after treatment commenced. Using statistical modeling, they determined that 64% of patients treated with placebo were cured at 14 days compared with 70% given an antibiotic. One patient out of 1381 treated with placebo experienced a serious complication, a brain abscess.

Do antibiotics benefit any subgroups?

The investigators also analyzed the prognostic value of specific signs and symptoms to answer the question: Is there any subgroup of patients who might benefit more from antibiotic treatment?

Duration. Patients with a longer duration of symptoms, more severe symptoms, or increased age took longer to cure, but were no more likely to benefit from antibiotic treatment than other patients.

Symptoms, such as a previous common cold, pain on bending, unilateral facial pain, tooth pain, and purulent nasal discharge did not have any prognostic value.

Only one sign—purulent discharge noted in the pharynx on examination—was associated with a higher likelihood of benefit from treatment with antibiotics, but the NNT was still 8 in this group. Patients with symptoms for 7 days or longer were no more likely to respond to antibiotics than those with symptoms for fewer than 7 days.¹

WHAT’S NEW: Realistic evidence from realistic settings

We believe this meta-analysis provides a high level of evidence against routine treatment of sinusitis with antibiotics in primary care practice. Treating 15 patients with an antibiotic to possibly benefit 1 patient 2 weeks after treatment commences does not seem like a good idea when one considers the cost and complications of antibiotic use. Diarrhea and other adverse outcomes are 80% more common among patients with sinusitis who are treated with an antibiotic compared with placebo.³ As noted above, prior meta-analyses of antibiotic treatment for acute sinusitis have been more encouraging than this meta-analysis, with a number needed to treat of 7, but those meta-analyses are clearly overly optimistic for the results one will achieve in primary care practice using clinical signs and symptoms to diagnose acute sinusitis.^3,4 Unlike the Young study, they included trials in specialty clinics with CT scans and sinus puncture and culture used for the diagnostic standard.

Symptoms >1 week are not a reason to prescribe

One very important new finding in this meta-analysis that should change practice is that the duration of illness did not predict a positive response to antibiotics.

Current national recommendations are to use an antibiotic for patients with a duration of illness longer than 1 week, as these patients are presumably more likely to have a bacterial infection.^5-7 However, that recommendation had been based on expert opinion, not on data from clinical trials. A longer duration of symptoms should not be a reason to prescribe an antibiotic for sinusitis symptoms.

How can you help your patient?

What to do, then, for patients with acute sinusitis? Treat the symptoms, which means recommending pain medication for facial pain or headache and saline nasal spray for the nasal discharge, not antibiotics or nasal corticosteroids. Side effects will be fewer and costs will be lower.

• Saline irrigation. A 2007 Cochrane review of 8 chronic and recurrent sinusitis trials showed that nasal saline irrigation is effective for reducing symptoms of chronic and recurrent sinusitis.⁸ Although we do not have high-quality RCT data on saline nasal irrigation for treatment of acute sinusitis, nasal saline irrigation is harmless and inexpensive.
• What about nasal steroids? The evidence is equivocal, and the most recent high-quality RCT of nasal steroids showed no effect.⁹

CAVEATS: Refer seriously ill patients and complicated cases

A very important caveat to our recommendation is that seriously ill patients must be managed differently. Very infrequently a patient develops a serious complication of acute sinusitis such as brain abscess, periorbital cellulitis, or meningitis. Therefore, seriously ill patients with signs and symptoms of acute bacterial sinusitis, such as high fever, periorbital erythema or edema, severe headache, or intense facial pain must be carefully evaluated and treated with great caution and close follow-up. These patients should be referred immediately for consultation with an otolaryngologist.

Of course, mildly ill patients today may become quite ill tomorrow, so always provide advice to patients to return if they are getting worse, a good clinical practice for any condition that is usually benign but occasionally serious.

Patients who have prolonged or recurrent sinusitis symptoms need further evaluation for other diagnoses such as allergies, cystic fibrosis, fungal sinus infection, and other illnesses associated with immune compromise. These complicated patients benefit from consultation with an otolaryngologist who has a specific interest in chronic and recurrent sinusitis, and perhaps consultation from an infectious disease specialist as well.

CHALLENGES TO IMPLEMENTATION: The patient who wants a pill

Some patients may be accustomed to receiving an antibiotic prescription for their “sinus infections” and may resist conservative management. It may be difficult to convince them that antibiotics won’t make a difference when they attribute past resolution of symptoms to antibiotics.

Take enough time to educate your patients on the natural course of illness, the positive benefits of nasal saline, and the reasons not to use unnecessary antibiotics (eg, they are not effective, have potential adverse effects, and can contribute to future antibiotic resistance); this effort will save you time in future visits.¹⁰ A “just in case you don’t get better” prescription to be filled only if the patient is not improving in the next few days is about 50% effective in reducing antibiotic usage for upper respiratory infections.¹¹

Acknowledgement

We acknowledge Sofia Medvedev of the University HealthSystem Consortium (UHC) in Oak Brook, IL for analysis of the National Ambulatory Medical Care Survey data.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

ILLUSTRATIVE CASE

A 23-year-old woman presents to your office with a 1-week history of cough, purulent nasal discharge, and unilateral facial pain. You diagnose acute sinusitis.

Should you prescribe an antibiotic?

No. Yet it’s no wonder that most adults treated for acute sinusitis leave the doctor’s office with a prescription for antibiotics. Until the publication of the meta-analysis by Young and colleagues¹ featured in this PURL, we have lacked A-level evidence from studies conducted in realistic settings—like your practice and ours.

Review of serial data from the National Ambulatory Medical Care Surveys (NAMCS) from 1999 through 2005 does show a slight downward trend in antibiotic prescribing for acute sinusitis: 1999-2002 data showed that 83% of cases of acute sinusitis were treated with an antibiotic.² Data from the 2004 and 2005 NAMCS reveal that family physicians prescribed antibiotics for 80% of patients with acute sinusitis in 2004 and 76% of patients in 2005 (S. Medvedev, unpublished data, NAMCS database, March 2008).

Is this continued high rate of antibiotic prescribing justified?

Do antibiotics improve symptoms and shorten the duration of illness or not?

These questions are important, obviously, not only because of the high rate of prescribing but also because sinusitis is one of the most common diagnoses: approximately 20 million cases annually in the United States, or about 21% of all outpatient antibiotic prescriptions for adults.²

Which patients might benefit from antibiotics?

Common clinical signs and symptoms cannot identify patients with rhinosinusitis for whom treatment is clearly justified, given the cost, adverse events, and bacterial resistance associated with antibiotic use

• Severity of symptoms is important only in that signs suggestive of a serious complication are the sole reason for immediate antibiotic treatment
• Purulent discharge noted in the pharynx on exam was associated with a higher likelihood of benefit from antibiotics, but NNT was 8
• Antibiotics are not justified even if a patient reports having symptoms for longer than 7-10 days

Source: Young et al.¹

A diagnostic dilemma

Before we discuss the evidence that is summarized in the excellent meta-analysis by Young and colleagues,¹ let’s acknowledge that acute sinusitis is undeniably a diagnostic dilemma. Distinguishing bacterial from viral infection is nearly impossible on clinical grounds because the symptoms are so similar. A litany of identical upper respiratory symptoms accompanies both viral and bacterial sinus infections. Purulent nasal secretions, maxillary facial pain (especially with unilateral predominance), maxillary tooth pain (which is uncommon with sinus infection), altered sense of smell, and worsening illness after improvement constitute the short list of signs and symptoms that has some predictive value, but even the presence of all of these is not a terrific predictor of bacterial sinus infection. Plain x-rays have low accuracy in distinguishing viral from bacterial infection. Computed tomography (CT) sinus scans are better but far from perfect, are not readily available in practice, and are expensive.

Sinusitis in the real world

How effective are antibiotics for patients diagnosed not by sinus x-rays or CTs, but by signs and symptoms—as we typically do in daily practice?

A meta-analysis³ of 13 randomized controlled trials (RCTs) found that sinusitis improved without antibiotics, but it included trials in which patients were recruited based on results of imaging studies and cultures, which are not normally used in primary care clinical practice. That study compared antibiotic treatment to placebo for acute uncomplicated sinusitis; 35% of placebo-treated patients were clinically cured by 7 to 12 days and 73% were improved after 7 days. Antibiotic therapy increased cure rates by 15% and improvement rates by 14%, yielding a number needed to treat of 7 to achieve 1 additional positive outcome at 7 days.

STUDY SUMMARY: Meta-analysis of primary care trials

Young and colleagues¹ aggregated and analyzed individual patient-level data from all known placebo-controlled, randomized, antibiotic treatment trials of adults with clinical symptoms of acute sinusitis that were conducted in primary care settings. They excluded trials that used imaging or bacterial culture as part of patient recruitment.

Studies were included that allowed the use of concomitant medication such as nonsteroidal anti-inflammatory drugs, decongestants, or nasal steroids, as long as patients in both groups had access to the same medications. All trials excluded patients with severe symptoms such as high fever, periorbital swelling or erythema, or intense facial pain, important exclusions that we will discuss below.

They identified 10 such studies and completed an intent-to-treat analysis of the 9 double-blind trials for which patient level data were available. Using individual data from 2547 patients, the odds ratio for an overall antibiotic treatment effect was 1.37 (95% confidence interval, 1.13-1.66), with a number needed to treat (NNT) of 15.

This finding means that 15 patients needed to be given an antibiotic for 1 additional patient to be cured at 8 to 15 days after treatment commenced. Using statistical modeling, they determined that 64% of patients treated with placebo were cured at 14 days compared with 70% given an antibiotic. One patient out of 1381 treated with placebo experienced a serious complication, a brain abscess.

Do antibiotics benefit any subgroups?

The investigators also analyzed the prognostic value of specific signs and symptoms to answer the question: Is there any subgroup of patients who might benefit more from antibiotic treatment?

Duration. Patients with a longer duration of symptoms, more severe symptoms, or increased age took longer to cure, but were no more likely to benefit from antibiotic treatment than other patients.

Symptoms, such as a previous common cold, pain on bending, unilateral facial pain, tooth pain, and purulent nasal discharge did not have any prognostic value.

Only one sign—purulent discharge noted in the pharynx on examination—was associated with a higher likelihood of benefit from treatment with antibiotics, but the NNT was still 8 in this group. Patients with symptoms for 7 days or longer were no more likely to respond to antibiotics than those with symptoms for fewer than 7 days.¹

WHAT’S NEW: Realistic evidence from realistic settings

We believe this meta-analysis provides a high level of evidence against routine treatment of sinusitis with antibiotics in primary care practice. Treating 15 patients with an antibiotic to possibly benefit 1 patient 2 weeks after treatment commences does not seem like a good idea when one considers the cost and complications of antibiotic use. Diarrhea and other adverse outcomes are 80% more common among patients with sinusitis who are treated with an antibiotic compared with placebo.³ As noted above, prior meta-analyses of antibiotic treatment for acute sinusitis have been more encouraging than this meta-analysis, with a number needed to treat of 7, but those meta-analyses are clearly overly optimistic for the results one will achieve in primary care practice using clinical signs and symptoms to diagnose acute sinusitis.^3,4 Unlike the Young study, they included trials in specialty clinics with CT scans and sinus puncture and culture used for the diagnostic standard.

Symptoms >1 week are not a reason to prescribe

One very important new finding in this meta-analysis that should change practice is that the duration of illness did not predict a positive response to antibiotics.

Current national recommendations are to use an antibiotic for patients with a duration of illness longer than 1 week, as these patients are presumably more likely to have a bacterial infection.^5-7 However, that recommendation had been based on expert opinion, not on data from clinical trials. A longer duration of symptoms should not be a reason to prescribe an antibiotic for sinusitis symptoms.

How can you help your patient?

What to do, then, for patients with acute sinusitis? Treat the symptoms, which means recommending pain medication for facial pain or headache and saline nasal spray for the nasal discharge, not antibiotics or nasal corticosteroids. Side effects will be fewer and costs will be lower.

• Saline irrigation. A 2007 Cochrane review of 8 chronic and recurrent sinusitis trials showed that nasal saline irrigation is effective for reducing symptoms of chronic and recurrent sinusitis.⁸ Although we do not have high-quality RCT data on saline nasal irrigation for treatment of acute sinusitis, nasal saline irrigation is harmless and inexpensive.
• What about nasal steroids? The evidence is equivocal, and the most recent high-quality RCT of nasal steroids showed no effect.⁹

CAVEATS: Refer seriously ill patients and complicated cases

A very important caveat to our recommendation is that seriously ill patients must be managed differently. Very infrequently a patient develops a serious complication of acute sinusitis such as brain abscess, periorbital cellulitis, or meningitis. Therefore, seriously ill patients with signs and symptoms of acute bacterial sinusitis, such as high fever, periorbital erythema or edema, severe headache, or intense facial pain must be carefully evaluated and treated with great caution and close follow-up. These patients should be referred immediately for consultation with an otolaryngologist.

Of course, mildly ill patients today may become quite ill tomorrow, so always provide advice to patients to return if they are getting worse, a good clinical practice for any condition that is usually benign but occasionally serious.

Patients who have prolonged or recurrent sinusitis symptoms need further evaluation for other diagnoses such as allergies, cystic fibrosis, fungal sinus infection, and other illnesses associated with immune compromise. These complicated patients benefit from consultation with an otolaryngologist who has a specific interest in chronic and recurrent sinusitis, and perhaps consultation from an infectious disease specialist as well.

CHALLENGES TO IMPLEMENTATION: The patient who wants a pill

Some patients may be accustomed to receiving an antibiotic prescription for their “sinus infections” and may resist conservative management. It may be difficult to convince them that antibiotics won’t make a difference when they attribute past resolution of symptoms to antibiotics.

Take enough time to educate your patients on the natural course of illness, the positive benefits of nasal saline, and the reasons not to use unnecessary antibiotics (eg, they are not effective, have potential adverse effects, and can contribute to future antibiotic resistance); this effort will save you time in future visits.¹⁰ A “just in case you don’t get better” prescription to be filled only if the patient is not improving in the next few days is about 50% effective in reducing antibiotic usage for upper respiratory infections.¹¹

Acknowledgement

We acknowledge Sofia Medvedev of the University HealthSystem Consortium (UHC) in Oak Brook, IL for analysis of the National Ambulatory Medical Care Survey data.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Illustrative case

Your first 4 patients this morning were a 50-year-old woman with metabolic syndrome, a 62-year-old obese man with high blood pressure, a 44-year-old woman with depression, and a 75-year-old man with a recent admission for myocardial infarction. In addition to managing their medications and reviewing lab results, you have already spent a lot of time discussing the benefits of exercise with each of these patients.

As you prepare to talk with your next patient—a 28-year-old woman with a BMI of 29 whose chief complaint is “wants to lose weight”—you wonder if there are any simple, brief, effective interventions to help your patients increase their physical activity.

BACKGROUND: A long way to go

Although there is no evidence that simply advising patients to walk has any effect, primary care physicians frequently recommend walking as a form of exercise—it is free, requires no special equipment, and is readily accessible to most motivated patients.

The Centers for Disease Control and Prevention recommends that adults engage in moderate physical activity for at least 30 minutes a day, at least 5 days per week.² Yet 40% of adults do not engage in any leisure-time physical activity. This percentage is higher in women (43%), African-Americans (52%), and Hispanics (54%).³

The health benefits of exercise are clear. Regular physical activity has been shown to decrease overweight and obesity.⁴ It has also been shown to improve control of type 2 diabetes⁵ and hypertension.⁶ Frequent exercise is associated with a decreased mortality rate.⁷ Walking has been shown to decrease the risk of cardiovascular events in women, regardless of BMI.⁸

Walking has similarly been shown to decrease overall mortality among men.⁹ Cardiovascular fitness has also been shown to decrease mortality in adults over 60, even in the absence of weight loss.¹⁰

CLINICAL CONTEXT: USPSTF: Advice alone won’t kick-start exercise

We realize, of course, that most of our adult patients could benefit from regular exercise. Exercise is included in the treatment guidelines for overweight/obesity, hypertension, type 2 diabetes, metabolic syndrome, cardiovascular disease, chronic pain, peripheral vascular disease, and depression.¹¹

Eureka! a simple, practical intervention

PURLs EDITOR
Bernard Ewigman. MD, MSPH
Department of Family Medicine
The University of Chicago
be.editor@gmail.com

At last, the humble pedometer gives us a brief intervention for physical exercise that works. yes, we need more research for lots of reasons (always), but this Purl gives us a practical tool that can be recommended in a few minutes, consistent with the realities of daily practice.

The outcomes from this intervention are not dramatic. No lives were saved, no catastrophic diseases averted. yet regular exercise is so fundamentally important to just feeling good and having energy for daily life, not to mention lowering blood pressure and weight.

My guess is that this could become a handy recommendation used daily in family medicine and other primary care practices.

I am interested to know whether you already recommend pedometers to your patients. If not, does this seem like a worthwhile change in your practice?

On a personal note, I made a New year’s resolution to increase my physical activity. as soon as I finish this commentary, I am ordering a pedometer.

However, few office-based interventions have been shown to lead to increased physical activity. Patients sometimes resist making lifestyle changes, and providers are uncertain how to effectively promote physical activity. Furthermore, counseling patients to exercise without a specific intervention has not been shown to lead to long-term increases in physical activity. The US Preventive Services Task Force (USPSTF) finds there is insufficient evidence to recommend behavioral counseling alone for exercise, citing the lack of evidence for long-term efficacy.^12,13

STUDY SUMMARY: Pedometer users walked 2491 additional steps

This meta-analysis included 26 RCTs and observational studies of pedometer use in adult outpatients that reported a change in the number of steps walked per day. The 2767 participants in these studies were 85% women, with a mean age of 49. In the 7 studies that reported race, 93% of patients were white. At baseline, most participants were overweight, with normal blood pressure (mean 129/79 mm Hg) and relatively well-controlled lipid levels (mean total cholesterol 198 mg/dL, HDL 52 mg/dL, LDL 113 mg/dL). The mean baseline activity level was 7473 steps per day (range 2140–12,371). Duration of interventions ranged from 3 to 104 weeks, with a mean of 18 weeks. Sixteen of the studies used the Yamax pedometer, which has been validated for accuracy and reliability.

Participants in the RCTs who used pedometers increased their physical activity by 2491 steps per day more than controls. After excluding 1 study with a much higher increase in physical activity than the others, the increase was 2004 steps per day (95% confidence interval [CI], 878–3129; P<.001). In the observational studies, participants walked 2183 steps per day more than they had at baseline (95% CI, 1571–2796; P<.001). Overall, pedometer users increased their number of steps by 27% over baseline.

Step goal and step diary

Only studies that included a step goal and required participants to keep a step diary showed a significant increase in physical activity with pedometer use. There were no differences in outcomes based on duration of the intervention, inclusion of physical activity counseling, or the brand of pedometer used.

BMI and BP improved; lipids, glucose did not

Intervention participants had a statistically significant decrease in BMI of 0.38, which was associated with older age (P=.001), having a step goal (P=.04), and longer duration of the intervention (P=.07, trend). Intervention participants also had a significant decrease in systolic blood pressure of 3.8 mm Hg and diastolic blood pressure of 0.3 mm Hg (TABLE 1), which was associated with greater systolic blood pressure at baseline (P=.009).

There were no significant differences in serum lipids or fasting serum glucose in the studies that reported these variables.¹

TABLE 1
Pre- and post-intervention body mass index and blood pressure

WHAT’S NEW?: Weight loss without dieting

This study is the first large meta-analysis to show that pedometer use is an effective intervention for promoting physical activity. Another recent meta-analysis shows that pedometer use is also effective for short-term weight loss, even in the absence of dietary changes.¹⁴

Pedometers and goal-setting are simple, relatively inexpensive ways to help patients become physically active. According to systematic reviews,^15,16 telephone-based programs, encouraging stair use, and creating exercise space are other effective interventions to promote physical activity. Some of these interventions are at least as effective as pedometers; however, only encouraging stair walking and pedometer use are practical office-based interventions.

CAVEATS: Price and quality

A 2004 Consumer Reports article ranked pedometers by accuracy, ease of use, and features.¹⁷ Accurate step counts allow patients and physicians to assess whether step goals are being met. Pedometers are more accurate when recording fast walking (2.5–3.0 mph), compared with slow walking. Pedometers may therefore be less accurate in the elderly, very obese, or those who walk slowly.¹⁸

TABLE 2
Consumer Reports top-rated pedometers¹⁷

Negotiate the goal, patient keeps diary

Remember that patients must be counseled to set a step goal and keep a step diary. Most patients will have an initial step goal between 6000 and 10,000 steps per day. The step goal should be individualized to each patient’s current level of activity and gradually increased as activity level increases.

Schedule monthly or semi-monthly follow-up visits to evaluate progress towards activity or weight loss goals and to re-evaluate the step goal. Before beginning an exercise regimen, including walking, patients must be healthy enough for physical activity. In some cases, patients will need stress testing or other evaluation before using a pedometer to increase activity.

CHALLENGES TO IMPLEMENTATION: Time-wise

Counseling patients on the use of pedometers, and coaching them to set an appropriate step goal and keep a step diary, will take up time during the office visit, but it should be a brief intervention and therefore feasible.²¹

Organizing your office staff to assist you, and using a patient handout containing the basic information on pedometers, could reduce the demands on your time. Including information from the 2004 Consumer Reports article and Web sites with pedometers prices (such as www.pedometersusa.com and newlifestyles.com) should provide a good start for those patients who want more information.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Illustrative case

Your first 4 patients this morning were a 50-year-old woman with metabolic syndrome, a 62-year-old obese man with high blood pressure, a 44-year-old woman with depression, and a 75-year-old man with a recent admission for myocardial infarction. In addition to managing their medications and reviewing lab results, you have already spent a lot of time discussing the benefits of exercise with each of these patients.

As you prepare to talk with your next patient—a 28-year-old woman with a BMI of 29 whose chief complaint is “wants to lose weight”—you wonder if there are any simple, brief, effective interventions to help your patients increase their physical activity.

BACKGROUND: A long way to go

Although there is no evidence that simply advising patients to walk has any effect, primary care physicians frequently recommend walking as a form of exercise—it is free, requires no special equipment, and is readily accessible to most motivated patients.

The Centers for Disease Control and Prevention recommends that adults engage in moderate physical activity for at least 30 minutes a day, at least 5 days per week.² Yet 40% of adults do not engage in any leisure-time physical activity. This percentage is higher in women (43%), African-Americans (52%), and Hispanics (54%).³

The health benefits of exercise are clear. Regular physical activity has been shown to decrease overweight and obesity.⁴ It has also been shown to improve control of type 2 diabetes⁵ and hypertension.⁶ Frequent exercise is associated with a decreased mortality rate.⁷ Walking has been shown to decrease the risk of cardiovascular events in women, regardless of BMI.⁸

Walking has similarly been shown to decrease overall mortality among men.⁹ Cardiovascular fitness has also been shown to decrease mortality in adults over 60, even in the absence of weight loss.¹⁰

CLINICAL CONTEXT: USPSTF: Advice alone won’t kick-start exercise

We realize, of course, that most of our adult patients could benefit from regular exercise. Exercise is included in the treatment guidelines for overweight/obesity, hypertension, type 2 diabetes, metabolic syndrome, cardiovascular disease, chronic pain, peripheral vascular disease, and depression.¹¹

Eureka! a simple, practical intervention

PURLs EDITOR
Bernard Ewigman. MD, MSPH
Department of Family Medicine
The University of Chicago
be.editor@gmail.com

At last, the humble pedometer gives us a brief intervention for physical exercise that works. yes, we need more research for lots of reasons (always), but this Purl gives us a practical tool that can be recommended in a few minutes, consistent with the realities of daily practice.

The outcomes from this intervention are not dramatic. No lives were saved, no catastrophic diseases averted. yet regular exercise is so fundamentally important to just feeling good and having energy for daily life, not to mention lowering blood pressure and weight.

My guess is that this could become a handy recommendation used daily in family medicine and other primary care practices.

I am interested to know whether you already recommend pedometers to your patients. If not, does this seem like a worthwhile change in your practice?

On a personal note, I made a New year’s resolution to increase my physical activity. as soon as I finish this commentary, I am ordering a pedometer.

However, few office-based interventions have been shown to lead to increased physical activity. Patients sometimes resist making lifestyle changes, and providers are uncertain how to effectively promote physical activity. Furthermore, counseling patients to exercise without a specific intervention has not been shown to lead to long-term increases in physical activity. The US Preventive Services Task Force (USPSTF) finds there is insufficient evidence to recommend behavioral counseling alone for exercise, citing the lack of evidence for long-term efficacy.^12,13

STUDY SUMMARY: Pedometer users walked 2491 additional steps

This meta-analysis included 26 RCTs and observational studies of pedometer use in adult outpatients that reported a change in the number of steps walked per day. The 2767 participants in these studies were 85% women, with a mean age of 49. In the 7 studies that reported race, 93% of patients were white. At baseline, most participants were overweight, with normal blood pressure (mean 129/79 mm Hg) and relatively well-controlled lipid levels (mean total cholesterol 198 mg/dL, HDL 52 mg/dL, LDL 113 mg/dL). The mean baseline activity level was 7473 steps per day (range 2140–12,371). Duration of interventions ranged from 3 to 104 weeks, with a mean of 18 weeks. Sixteen of the studies used the Yamax pedometer, which has been validated for accuracy and reliability.

Participants in the RCTs who used pedometers increased their physical activity by 2491 steps per day more than controls. After excluding 1 study with a much higher increase in physical activity than the others, the increase was 2004 steps per day (95% confidence interval [CI], 878–3129; P<.001). In the observational studies, participants walked 2183 steps per day more than they had at baseline (95% CI, 1571–2796; P<.001). Overall, pedometer users increased their number of steps by 27% over baseline.

Step goal and step diary

Only studies that included a step goal and required participants to keep a step diary showed a significant increase in physical activity with pedometer use. There were no differences in outcomes based on duration of the intervention, inclusion of physical activity counseling, or the brand of pedometer used.

BMI and BP improved; lipids, glucose did not

Intervention participants had a statistically significant decrease in BMI of 0.38, which was associated with older age (P=.001), having a step goal (P=.04), and longer duration of the intervention (P=.07, trend). Intervention participants also had a significant decrease in systolic blood pressure of 3.8 mm Hg and diastolic blood pressure of 0.3 mm Hg (TABLE 1), which was associated with greater systolic blood pressure at baseline (P=.009).

There were no significant differences in serum lipids or fasting serum glucose in the studies that reported these variables.¹

TABLE 1
Pre- and post-intervention body mass index and blood pressure

WHAT’S NEW?: Weight loss without dieting

This study is the first large meta-analysis to show that pedometer use is an effective intervention for promoting physical activity. Another recent meta-analysis shows that pedometer use is also effective for short-term weight loss, even in the absence of dietary changes.¹⁴

Pedometers and goal-setting are simple, relatively inexpensive ways to help patients become physically active. According to systematic reviews,^15,16 telephone-based programs, encouraging stair use, and creating exercise space are other effective interventions to promote physical activity. Some of these interventions are at least as effective as pedometers; however, only encouraging stair walking and pedometer use are practical office-based interventions.

CAVEATS: Price and quality

A 2004 Consumer Reports article ranked pedometers by accuracy, ease of use, and features.¹⁷ Accurate step counts allow patients and physicians to assess whether step goals are being met. Pedometers are more accurate when recording fast walking (2.5–3.0 mph), compared with slow walking. Pedometers may therefore be less accurate in the elderly, very obese, or those who walk slowly.¹⁸

TABLE 2
Consumer Reports top-rated pedometers¹⁷

Negotiate the goal, patient keeps diary

Remember that patients must be counseled to set a step goal and keep a step diary. Most patients will have an initial step goal between 6000 and 10,000 steps per day. The step goal should be individualized to each patient’s current level of activity and gradually increased as activity level increases.

Schedule monthly or semi-monthly follow-up visits to evaluate progress towards activity or weight loss goals and to re-evaluate the step goal. Before beginning an exercise regimen, including walking, patients must be healthy enough for physical activity. In some cases, patients will need stress testing or other evaluation before using a pedometer to increase activity.

CHALLENGES TO IMPLEMENTATION: Time-wise

Counseling patients on the use of pedometers, and coaching them to set an appropriate step goal and keep a step diary, will take up time during the office visit, but it should be a brief intervention and therefore feasible.²¹

Organizing your office staff to assist you, and using a patient handout containing the basic information on pedometers, could reduce the demands on your time. Including information from the 2004 Consumer Reports article and Web sites with pedometers prices (such as www.pedometersusa.com and newlifestyles.com) should provide a good start for those patients who want more information.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Illustrative case

Your first 4 patients this morning were a 50-year-old woman with metabolic syndrome, a 62-year-old obese man with high blood pressure, a 44-year-old woman with depression, and a 75-year-old man with a recent admission for myocardial infarction. In addition to managing their medications and reviewing lab results, you have already spent a lot of time discussing the benefits of exercise with each of these patients.

As you prepare to talk with your next patient—a 28-year-old woman with a BMI of 29 whose chief complaint is “wants to lose weight”—you wonder if there are any simple, brief, effective interventions to help your patients increase their physical activity.

BACKGROUND: A long way to go

Although there is no evidence that simply advising patients to walk has any effect, primary care physicians frequently recommend walking as a form of exercise—it is free, requires no special equipment, and is readily accessible to most motivated patients.

The Centers for Disease Control and Prevention recommends that adults engage in moderate physical activity for at least 30 minutes a day, at least 5 days per week.² Yet 40% of adults do not engage in any leisure-time physical activity. This percentage is higher in women (43%), African-Americans (52%), and Hispanics (54%).³

The health benefits of exercise are clear. Regular physical activity has been shown to decrease overweight and obesity.⁴ It has also been shown to improve control of type 2 diabetes⁵ and hypertension.⁶ Frequent exercise is associated with a decreased mortality rate.⁷ Walking has been shown to decrease the risk of cardiovascular events in women, regardless of BMI.⁸

Walking has similarly been shown to decrease overall mortality among men.⁹ Cardiovascular fitness has also been shown to decrease mortality in adults over 60, even in the absence of weight loss.¹⁰

CLINICAL CONTEXT: USPSTF: Advice alone won’t kick-start exercise

We realize, of course, that most of our adult patients could benefit from regular exercise. Exercise is included in the treatment guidelines for overweight/obesity, hypertension, type 2 diabetes, metabolic syndrome, cardiovascular disease, chronic pain, peripheral vascular disease, and depression.¹¹

Eureka! a simple, practical intervention

PURLs EDITOR
Bernard Ewigman. MD, MSPH
Department of Family Medicine
The University of Chicago
be.editor@gmail.com

At last, the humble pedometer gives us a brief intervention for physical exercise that works. yes, we need more research for lots of reasons (always), but this Purl gives us a practical tool that can be recommended in a few minutes, consistent with the realities of daily practice.

The outcomes from this intervention are not dramatic. No lives were saved, no catastrophic diseases averted. yet regular exercise is so fundamentally important to just feeling good and having energy for daily life, not to mention lowering blood pressure and weight.

My guess is that this could become a handy recommendation used daily in family medicine and other primary care practices.

I am interested to know whether you already recommend pedometers to your patients. If not, does this seem like a worthwhile change in your practice?

On a personal note, I made a New year’s resolution to increase my physical activity. as soon as I finish this commentary, I am ordering a pedometer.

However, few office-based interventions have been shown to lead to increased physical activity. Patients sometimes resist making lifestyle changes, and providers are uncertain how to effectively promote physical activity. Furthermore, counseling patients to exercise without a specific intervention has not been shown to lead to long-term increases in physical activity. The US Preventive Services Task Force (USPSTF) finds there is insufficient evidence to recommend behavioral counseling alone for exercise, citing the lack of evidence for long-term efficacy.^12,13

STUDY SUMMARY: Pedometer users walked 2491 additional steps

This meta-analysis included 26 RCTs and observational studies of pedometer use in adult outpatients that reported a change in the number of steps walked per day. The 2767 participants in these studies were 85% women, with a mean age of 49. In the 7 studies that reported race, 93% of patients were white. At baseline, most participants were overweight, with normal blood pressure (mean 129/79 mm Hg) and relatively well-controlled lipid levels (mean total cholesterol 198 mg/dL, HDL 52 mg/dL, LDL 113 mg/dL). The mean baseline activity level was 7473 steps per day (range 2140–12,371). Duration of interventions ranged from 3 to 104 weeks, with a mean of 18 weeks. Sixteen of the studies used the Yamax pedometer, which has been validated for accuracy and reliability.

Participants in the RCTs who used pedometers increased their physical activity by 2491 steps per day more than controls. After excluding 1 study with a much higher increase in physical activity than the others, the increase was 2004 steps per day (95% confidence interval [CI], 878–3129; P<.001). In the observational studies, participants walked 2183 steps per day more than they had at baseline (95% CI, 1571–2796; P<.001). Overall, pedometer users increased their number of steps by 27% over baseline.

Step goal and step diary

Only studies that included a step goal and required participants to keep a step diary showed a significant increase in physical activity with pedometer use. There were no differences in outcomes based on duration of the intervention, inclusion of physical activity counseling, or the brand of pedometer used.

BMI and BP improved; lipids, glucose did not

Intervention participants had a statistically significant decrease in BMI of 0.38, which was associated with older age (P=.001), having a step goal (P=.04), and longer duration of the intervention (P=.07, trend). Intervention participants also had a significant decrease in systolic blood pressure of 3.8 mm Hg and diastolic blood pressure of 0.3 mm Hg (TABLE 1), which was associated with greater systolic blood pressure at baseline (P=.009).

There were no significant differences in serum lipids or fasting serum glucose in the studies that reported these variables.¹

TABLE 1
Pre- and post-intervention body mass index and blood pressure

WHAT’S NEW?: Weight loss without dieting

This study is the first large meta-analysis to show that pedometer use is an effective intervention for promoting physical activity. Another recent meta-analysis shows that pedometer use is also effective for short-term weight loss, even in the absence of dietary changes.¹⁴

Pedometers and goal-setting are simple, relatively inexpensive ways to help patients become physically active. According to systematic reviews,^15,16 telephone-based programs, encouraging stair use, and creating exercise space are other effective interventions to promote physical activity. Some of these interventions are at least as effective as pedometers; however, only encouraging stair walking and pedometer use are practical office-based interventions.

CAVEATS: Price and quality

A 2004 Consumer Reports article ranked pedometers by accuracy, ease of use, and features.¹⁷ Accurate step counts allow patients and physicians to assess whether step goals are being met. Pedometers are more accurate when recording fast walking (2.5–3.0 mph), compared with slow walking. Pedometers may therefore be less accurate in the elderly, very obese, or those who walk slowly.¹⁸

TABLE 2
Consumer Reports top-rated pedometers¹⁷

Negotiate the goal, patient keeps diary

Remember that patients must be counseled to set a step goal and keep a step diary. Most patients will have an initial step goal between 6000 and 10,000 steps per day. The step goal should be individualized to each patient’s current level of activity and gradually increased as activity level increases.

Schedule monthly or semi-monthly follow-up visits to evaluate progress towards activity or weight loss goals and to re-evaluate the step goal. Before beginning an exercise regimen, including walking, patients must be healthy enough for physical activity. In some cases, patients will need stress testing or other evaluation before using a pedometer to increase activity.

CHALLENGES TO IMPLEMENTATION: Time-wise

Counseling patients on the use of pedometers, and coaching them to set an appropriate step goal and keep a step diary, will take up time during the office visit, but it should be a brief intervention and therefore feasible.²¹

Organizing your office staff to assist you, and using a patient handout containing the basic information on pedometers, could reduce the demands on your time. Including information from the 2004 Consumer Reports article and Web sites with pedometers prices (such as www.pedometersusa.com and newlifestyles.com) should provide a good start for those patients who want more information.

PURLs methodology

This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) Surveillance System methodology. The criteria and findings leading to the selection of this study as a PURL can be accessed at www.jfponline.com/purls.

Illustrative case

A 45-year-old man presents to your outpatient clinic within 24 hours after onset of left-sided facial nerve weakness and inability to smile on one side of his face. He asks for a therapy to help improve his symptoms quickly, as his daughter is getting married in a few months, and he will be in the wedding pictures.

Is there a treatment that will hasten his complete recovery?

Background: Insufficient statistical power until now

Many of us treat patients with Bell’s palsy with both corticosteroids and antiviral medications, such as acyclovir or valacyclovir, largely on the basis of pathophysiologic reasoning, because we’ve had no clear guidance from outcome studies. Until now, outcome studies have had mixed findings, and have been inconclusive.² Most outcome studies have lacked the statistical power to either detect or to rule out potential benefits convincingly. The study by Sullivan and colleagues is the first to have a sufficiently large study sample from which to draw more definitive conclusions based on patient-oriented outcomes.

Steroid plus antiviral makes sense, pathophysiologically

Corticosteroids are thought to decrease inflammation of the facial nerve during an episode of facial paralysis. Some have postulated that herpes simplex virus type I may be a cause of facial nerve paralysis, hence the treatment with antivirals.³

Most of our PURL surveillance system clinician reviewers said that they prescribe both corticosteroids and antivirals.

Guidelines: “probably, possibly”

For example, a report of the Quality Standards Subcommittee of the American Academy of Neurology concluded that benefit from both steroids and antivirals has not been well established in patients with Bell’s palsy. However, the report states that evidence suggests that steroids are safe and probably effective, while antivirals are also safe and possibly effective.⁴

In contrast, UpToDate suggests treating all patients seen within a week of symptom onset with corticosteroids (prednisone 60–80 mg daily) plus valacyclovir (1 g 3 times daily) for 1 week.⁵

CLINICAL CONTEXT: Quality of life, risk of permanent harm

Bell’s palsy, defined as an acute peripheral facial weakness of unknown cause, has an annual incidence of 20 to 32 per 100,000. Most patients recover completely, with or without treatment, but 20% to 30% can have permanent facial weakness or paralysis.

The time to resolution is a quality of life issue for those in whom disease does not resolve spontaneously.⁶

We think that this study provides convincing evidence that acyclovir is not indicated for Bell’s palsy and that corticosteroids are.

STUDY SUMMARY: 10-day treatment, starting promptly

This double-blind, placebo-controlled, randomized, multifactorial trial compared recovery of facial nerve function for patients randomized to receive 10 days of treatment with prednisolone (25 mg twice daily), acyclovir (400 mg 5 times daily), both agents, or placebo (lactose).

Inclusion criteria

Patients had to be at least 16 years of age (average age=44), with unilateral facial nerve weakness of no identifiable cause (eg, a diagnosis of Bell’s palsy). They were recruited mostly through their family doctors (75%) but also through emergency rooms and dental offices, and were referred to otolaryngologists at 17 Scottish hospitals within 72 hours.

The degree of initial facial paralysis was moderate to severe, based on the House-Brackmann scale, a widely used system for grading recovery from facial nerve paralysis. After the onset of symptoms, most patients (53.8%) initiated treatment within 24 hours, 32.1% within 48 hours, and 14.1% within 72 hours. Patients were assessed at baseline, 3 months, and 9 months.

Exclusion criteria

Exclusion criteria included pregnancy, breastfeeding, uncontrolled diabetes (Hb A_1c >8.0%), peptic ulcer disease, suppurative otitis media, herpes zoster, multiple sclerosis, systemic infection, sarcoid or other rare disorder, and inability to give informed consent.

FIGURE
Facial weakness or paralysis may be permanent

Primary outcome: Complete recovery

The study was designed to test the effectiveness of prednisolone and acyclovir’s effects on facial nerve recovery. The House-Brackmann scale was used to score recovery. The scale divides patients into 1 of 6 categories depending on the severity of facial nerve dysfunction, with grade 1 describing normal function and grade 6 indicating total paralysis.

The scale was applied to photographs of patients taken while smiling, raising eyebrows, at rest, and closing eyes. The photographs were assessed and graded independently by 3 experts: an otolaryngologist, a neurologist, and a plastic surgeon. They were unaware of the study group assignment or stage of assessment.

Of 496 patients who completed the study, 357 recovered fully at 3 months, with no further treatment needed. Of the remaining patients, 80 had fully recovered at 9 months and 59 still had some facial-nerve deficit. At 3 months, there was a significant difference in recovery rates in prednisolone comparison groups: 83% with prednisolone vs 63.6% without prednisolone, a difference of 19.4 percentage points (95% confidence interval [CI], 11.7 to 27.1; P<.001, number needed to treat [NNT]=5). There was no significant difference in recovery rates in acyclovir comparison groups: 71.2% with acyclovir vs 75% not treated with acyclovir, a difference of 4.5 % percentage points (95% CI, –12.4 to 3.3; unadjusted P=.30; adjusted P=.50). At 9 months, the rates of complete recovery were 94.4% in prednisolone treated groups vs 81.6% in no prednisolone treatment groups (NNT=8) (TABLE).

TABLE
Complete recovery was significantly higher in the group that received prednisolone without an antiviral

Adverse events

Adverse events included an expected range of minor symptoms associated with use of prednisolone and acyclovir, such as dizziness and vomiting. During the study, 3 patients died under circumstances unrelated to treatment: 2 were receiving double placebo and 1 received only acyclovir.¹

WHAT’S NEW: A treatment based on patient-oriented evidence

Neither corticosteroids nor antivirals are new treatments for Bell’s palsy. What is new is that we know what works (corticosteroids) and what does not work (antivirals). This randomized controlled trial finally gives us the evidence on patient-oriented outcomes that we need to make confident recommendations, primarily because it enrolled twice as many patients as all trials compiled for the Cochrane systematic reviews on this topic.^7,8

As an interesting side note, this is a good case study of how pathophysiologic reasoning sometimes leads us to good medical practice (corticosteroids in this case) and sometimes does not (antivirals in this case).

Isn’t it good to know that we can actually help patients with Bell’s palsy with corticosteroids and that antivirals are not necessary?

CAVEATS: Valacyclovir

Hato et al,⁹ in a Japanese study, showed that valacyclovir reaches a level of bioavailability that is 3 to 5 times more than acyclovir and may add some benefit to recovery when used in conjunction with prednisolone, particularly in more severe cases of Bell’s palsy.

The Hato study was a prospective, multicenter, randomized, placebo-controlled study that investigated the effects of valacyclovir (1000 mg/d for 5 days) and prednisolone in comparison with the effects of placebo and prednisolone for the treatment of Bell’s palsy.

The study outcomes included complete recovery from palsy; patients were followed until recovery occurred or more than 6 months in cases with severe prognosis. The patients in the Hato study had an average Yanagihara score of 15 when rating their facial palsy (which falls between House–Brackmann grades 4 and 5).

The overall rate of recovery of those treated with valacyclovir and prednisolone (96.5%) was significantly better (P<.05) than the rate among those treated with placebo and prednisolone (89.7%). In cases of complete or severe palsy, the rates of patients treated with both agents vs prednisolone alone who recovered were 95.7% (n=92) and 86.6% (n=82) (P<.05; NNT=11).

One big difference between the Sullivan and Hato studies is that the patients recruited for the Hato study had much more severe facial palsy (rated between 4 and 5) than in the Sullivan study (average=3.6), which suggests that there may be a use for valacyclovir in treating patients with complete facial palsy.⁶ Patients were all recruited from tertiary care centers as opposed to mainly from primary care settings as in the Sullivan study, consistent with the greater severity of cases in the Hato study.

Outcome assessors were not blinded to treatment assignments or stage of assessment in the Hato study, raising major concerns about the validity of the findings given the nature of facial paralysis as an outcome measure. We find the Sullivan study a more rigorous and convincing study. Nonetheless, future research may verify their findings and support the use of valacyclovir in the most severe cases of Bell’s palsy. For now, we are not convinced.

CHALLENGES TO IMPLEMENTATION: Easy to put into practice

Thankfully, some changes in practice are easy to implement. This is one of them. For those who prefer to prescribe prednisone, the dose of prednisolone used in the study, 25 mg bid, is equivalent to 60 mg of prednisone.

PURL surveillance system methodology
This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) surveillance system methodology.

Illustrative case

A 45-year-old man presents to your outpatient clinic within 24 hours after onset of left-sided facial nerve weakness and inability to smile on one side of his face. He asks for a therapy to help improve his symptoms quickly, as his daughter is getting married in a few months, and he will be in the wedding pictures.

Is there a treatment that will hasten his complete recovery?

Background: Insufficient statistical power until now

Many of us treat patients with Bell’s palsy with both corticosteroids and antiviral medications, such as acyclovir or valacyclovir, largely on the basis of pathophysiologic reasoning, because we’ve had no clear guidance from outcome studies. Until now, outcome studies have had mixed findings, and have been inconclusive.² Most outcome studies have lacked the statistical power to either detect or to rule out potential benefits convincingly. The study by Sullivan and colleagues is the first to have a sufficiently large study sample from which to draw more definitive conclusions based on patient-oriented outcomes.

Steroid plus antiviral makes sense, pathophysiologically

Corticosteroids are thought to decrease inflammation of the facial nerve during an episode of facial paralysis. Some have postulated that herpes simplex virus type I may be a cause of facial nerve paralysis, hence the treatment with antivirals.³

Most of our PURL surveillance system clinician reviewers said that they prescribe both corticosteroids and antivirals.

Guidelines: “probably, possibly”

For example, a report of the Quality Standards Subcommittee of the American Academy of Neurology concluded that benefit from both steroids and antivirals has not been well established in patients with Bell’s palsy. However, the report states that evidence suggests that steroids are safe and probably effective, while antivirals are also safe and possibly effective.⁴

In contrast, UpToDate suggests treating all patients seen within a week of symptom onset with corticosteroids (prednisone 60–80 mg daily) plus valacyclovir (1 g 3 times daily) for 1 week.⁵

CLINICAL CONTEXT: Quality of life, risk of permanent harm

Bell’s palsy, defined as an acute peripheral facial weakness of unknown cause, has an annual incidence of 20 to 32 per 100,000. Most patients recover completely, with or without treatment, but 20% to 30% can have permanent facial weakness or paralysis.

The time to resolution is a quality of life issue for those in whom disease does not resolve spontaneously.⁶

We think that this study provides convincing evidence that acyclovir is not indicated for Bell’s palsy and that corticosteroids are.

STUDY SUMMARY: 10-day treatment, starting promptly

This double-blind, placebo-controlled, randomized, multifactorial trial compared recovery of facial nerve function for patients randomized to receive 10 days of treatment with prednisolone (25 mg twice daily), acyclovir (400 mg 5 times daily), both agents, or placebo (lactose).

Inclusion criteria

Patients had to be at least 16 years of age (average age=44), with unilateral facial nerve weakness of no identifiable cause (eg, a diagnosis of Bell’s palsy). They were recruited mostly through their family doctors (75%) but also through emergency rooms and dental offices, and were referred to otolaryngologists at 17 Scottish hospitals within 72 hours.

The degree of initial facial paralysis was moderate to severe, based on the House-Brackmann scale, a widely used system for grading recovery from facial nerve paralysis. After the onset of symptoms, most patients (53.8%) initiated treatment within 24 hours, 32.1% within 48 hours, and 14.1% within 72 hours. Patients were assessed at baseline, 3 months, and 9 months.

Exclusion criteria

Exclusion criteria included pregnancy, breastfeeding, uncontrolled diabetes (Hb A_1c >8.0%), peptic ulcer disease, suppurative otitis media, herpes zoster, multiple sclerosis, systemic infection, sarcoid or other rare disorder, and inability to give informed consent.

FIGURE
Facial weakness or paralysis may be permanent

Primary outcome: Complete recovery

The study was designed to test the effectiveness of prednisolone and acyclovir’s effects on facial nerve recovery. The House-Brackmann scale was used to score recovery. The scale divides patients into 1 of 6 categories depending on the severity of facial nerve dysfunction, with grade 1 describing normal function and grade 6 indicating total paralysis.

The scale was applied to photographs of patients taken while smiling, raising eyebrows, at rest, and closing eyes. The photographs were assessed and graded independently by 3 experts: an otolaryngologist, a neurologist, and a plastic surgeon. They were unaware of the study group assignment or stage of assessment.

Of 496 patients who completed the study, 357 recovered fully at 3 months, with no further treatment needed. Of the remaining patients, 80 had fully recovered at 9 months and 59 still had some facial-nerve deficit. At 3 months, there was a significant difference in recovery rates in prednisolone comparison groups: 83% with prednisolone vs 63.6% without prednisolone, a difference of 19.4 percentage points (95% confidence interval [CI], 11.7 to 27.1; P<.001, number needed to treat [NNT]=5). There was no significant difference in recovery rates in acyclovir comparison groups: 71.2% with acyclovir vs 75% not treated with acyclovir, a difference of 4.5 % percentage points (95% CI, –12.4 to 3.3; unadjusted P=.30; adjusted P=.50). At 9 months, the rates of complete recovery were 94.4% in prednisolone treated groups vs 81.6% in no prednisolone treatment groups (NNT=8) (TABLE).

TABLE
Complete recovery was significantly higher in the group that received prednisolone without an antiviral

Adverse events

Adverse events included an expected range of minor symptoms associated with use of prednisolone and acyclovir, such as dizziness and vomiting. During the study, 3 patients died under circumstances unrelated to treatment: 2 were receiving double placebo and 1 received only acyclovir.¹

WHAT’S NEW: A treatment based on patient-oriented evidence

Neither corticosteroids nor antivirals are new treatments for Bell’s palsy. What is new is that we know what works (corticosteroids) and what does not work (antivirals). This randomized controlled trial finally gives us the evidence on patient-oriented outcomes that we need to make confident recommendations, primarily because it enrolled twice as many patients as all trials compiled for the Cochrane systematic reviews on this topic.^7,8

As an interesting side note, this is a good case study of how pathophysiologic reasoning sometimes leads us to good medical practice (corticosteroids in this case) and sometimes does not (antivirals in this case).

Isn’t it good to know that we can actually help patients with Bell’s palsy with corticosteroids and that antivirals are not necessary?

CAVEATS: Valacyclovir

Hato et al,⁹ in a Japanese study, showed that valacyclovir reaches a level of bioavailability that is 3 to 5 times more than acyclovir and may add some benefit to recovery when used in conjunction with prednisolone, particularly in more severe cases of Bell’s palsy.

The Hato study was a prospective, multicenter, randomized, placebo-controlled study that investigated the effects of valacyclovir (1000 mg/d for 5 days) and prednisolone in comparison with the effects of placebo and prednisolone for the treatment of Bell’s palsy.

The study outcomes included complete recovery from palsy; patients were followed until recovery occurred or more than 6 months in cases with severe prognosis. The patients in the Hato study had an average Yanagihara score of 15 when rating their facial palsy (which falls between House–Brackmann grades 4 and 5).

The overall rate of recovery of those treated with valacyclovir and prednisolone (96.5%) was significantly better (P<.05) than the rate among those treated with placebo and prednisolone (89.7%). In cases of complete or severe palsy, the rates of patients treated with both agents vs prednisolone alone who recovered were 95.7% (n=92) and 86.6% (n=82) (P<.05; NNT=11).

One big difference between the Sullivan and Hato studies is that the patients recruited for the Hato study had much more severe facial palsy (rated between 4 and 5) than in the Sullivan study (average=3.6), which suggests that there may be a use for valacyclovir in treating patients with complete facial palsy.⁶ Patients were all recruited from tertiary care centers as opposed to mainly from primary care settings as in the Sullivan study, consistent with the greater severity of cases in the Hato study.

Outcome assessors were not blinded to treatment assignments or stage of assessment in the Hato study, raising major concerns about the validity of the findings given the nature of facial paralysis as an outcome measure. We find the Sullivan study a more rigorous and convincing study. Nonetheless, future research may verify their findings and support the use of valacyclovir in the most severe cases of Bell’s palsy. For now, we are not convinced.

CHALLENGES TO IMPLEMENTATION: Easy to put into practice

Thankfully, some changes in practice are easy to implement. This is one of them. For those who prefer to prescribe prednisone, the dose of prednisolone used in the study, 25 mg bid, is equivalent to 60 mg of prednisone.

PURL surveillance system methodology
This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) surveillance system methodology.

Illustrative case

A 45-year-old man presents to your outpatient clinic within 24 hours after onset of left-sided facial nerve weakness and inability to smile on one side of his face. He asks for a therapy to help improve his symptoms quickly, as his daughter is getting married in a few months, and he will be in the wedding pictures.

Is there a treatment that will hasten his complete recovery?

Background: Insufficient statistical power until now

Many of us treat patients with Bell’s palsy with both corticosteroids and antiviral medications, such as acyclovir or valacyclovir, largely on the basis of pathophysiologic reasoning, because we’ve had no clear guidance from outcome studies. Until now, outcome studies have had mixed findings, and have been inconclusive.² Most outcome studies have lacked the statistical power to either detect or to rule out potential benefits convincingly. The study by Sullivan and colleagues is the first to have a sufficiently large study sample from which to draw more definitive conclusions based on patient-oriented outcomes.

Steroid plus antiviral makes sense, pathophysiologically

Corticosteroids are thought to decrease inflammation of the facial nerve during an episode of facial paralysis. Some have postulated that herpes simplex virus type I may be a cause of facial nerve paralysis, hence the treatment with antivirals.³

Most of our PURL surveillance system clinician reviewers said that they prescribe both corticosteroids and antivirals.

Guidelines: “probably, possibly”

For example, a report of the Quality Standards Subcommittee of the American Academy of Neurology concluded that benefit from both steroids and antivirals has not been well established in patients with Bell’s palsy. However, the report states that evidence suggests that steroids are safe and probably effective, while antivirals are also safe and possibly effective.⁴

In contrast, UpToDate suggests treating all patients seen within a week of symptom onset with corticosteroids (prednisone 60–80 mg daily) plus valacyclovir (1 g 3 times daily) for 1 week.⁵

CLINICAL CONTEXT: Quality of life, risk of permanent harm

Bell’s palsy, defined as an acute peripheral facial weakness of unknown cause, has an annual incidence of 20 to 32 per 100,000. Most patients recover completely, with or without treatment, but 20% to 30% can have permanent facial weakness or paralysis.

The time to resolution is a quality of life issue for those in whom disease does not resolve spontaneously.⁶

We think that this study provides convincing evidence that acyclovir is not indicated for Bell’s palsy and that corticosteroids are.

STUDY SUMMARY: 10-day treatment, starting promptly

This double-blind, placebo-controlled, randomized, multifactorial trial compared recovery of facial nerve function for patients randomized to receive 10 days of treatment with prednisolone (25 mg twice daily), acyclovir (400 mg 5 times daily), both agents, or placebo (lactose).

Inclusion criteria

Patients had to be at least 16 years of age (average age=44), with unilateral facial nerve weakness of no identifiable cause (eg, a diagnosis of Bell’s palsy). They were recruited mostly through their family doctors (75%) but also through emergency rooms and dental offices, and were referred to otolaryngologists at 17 Scottish hospitals within 72 hours.

The degree of initial facial paralysis was moderate to severe, based on the House-Brackmann scale, a widely used system for grading recovery from facial nerve paralysis. After the onset of symptoms, most patients (53.8%) initiated treatment within 24 hours, 32.1% within 48 hours, and 14.1% within 72 hours. Patients were assessed at baseline, 3 months, and 9 months.

Exclusion criteria

Exclusion criteria included pregnancy, breastfeeding, uncontrolled diabetes (Hb A_1c >8.0%), peptic ulcer disease, suppurative otitis media, herpes zoster, multiple sclerosis, systemic infection, sarcoid or other rare disorder, and inability to give informed consent.

FIGURE
Facial weakness or paralysis may be permanent

Primary outcome: Complete recovery

The study was designed to test the effectiveness of prednisolone and acyclovir’s effects on facial nerve recovery. The House-Brackmann scale was used to score recovery. The scale divides patients into 1 of 6 categories depending on the severity of facial nerve dysfunction, with grade 1 describing normal function and grade 6 indicating total paralysis.

The scale was applied to photographs of patients taken while smiling, raising eyebrows, at rest, and closing eyes. The photographs were assessed and graded independently by 3 experts: an otolaryngologist, a neurologist, and a plastic surgeon. They were unaware of the study group assignment or stage of assessment.

Of 496 patients who completed the study, 357 recovered fully at 3 months, with no further treatment needed. Of the remaining patients, 80 had fully recovered at 9 months and 59 still had some facial-nerve deficit. At 3 months, there was a significant difference in recovery rates in prednisolone comparison groups: 83% with prednisolone vs 63.6% without prednisolone, a difference of 19.4 percentage points (95% confidence interval [CI], 11.7 to 27.1; P<.001, number needed to treat [NNT]=5). There was no significant difference in recovery rates in acyclovir comparison groups: 71.2% with acyclovir vs 75% not treated with acyclovir, a difference of 4.5 % percentage points (95% CI, –12.4 to 3.3; unadjusted P=.30; adjusted P=.50). At 9 months, the rates of complete recovery were 94.4% in prednisolone treated groups vs 81.6% in no prednisolone treatment groups (NNT=8) (TABLE).

TABLE
Complete recovery was significantly higher in the group that received prednisolone without an antiviral

Adverse events

Adverse events included an expected range of minor symptoms associated with use of prednisolone and acyclovir, such as dizziness and vomiting. During the study, 3 patients died under circumstances unrelated to treatment: 2 were receiving double placebo and 1 received only acyclovir.¹

WHAT’S NEW: A treatment based on patient-oriented evidence

Neither corticosteroids nor antivirals are new treatments for Bell’s palsy. What is new is that we know what works (corticosteroids) and what does not work (antivirals). This randomized controlled trial finally gives us the evidence on patient-oriented outcomes that we need to make confident recommendations, primarily because it enrolled twice as many patients as all trials compiled for the Cochrane systematic reviews on this topic.^7,8

As an interesting side note, this is a good case study of how pathophysiologic reasoning sometimes leads us to good medical practice (corticosteroids in this case) and sometimes does not (antivirals in this case).

Isn’t it good to know that we can actually help patients with Bell’s palsy with corticosteroids and that antivirals are not necessary?

CAVEATS: Valacyclovir

Hato et al,⁹ in a Japanese study, showed that valacyclovir reaches a level of bioavailability that is 3 to 5 times more than acyclovir and may add some benefit to recovery when used in conjunction with prednisolone, particularly in more severe cases of Bell’s palsy.

The Hato study was a prospective, multicenter, randomized, placebo-controlled study that investigated the effects of valacyclovir (1000 mg/d for 5 days) and prednisolone in comparison with the effects of placebo and prednisolone for the treatment of Bell’s palsy.

The study outcomes included complete recovery from palsy; patients were followed until recovery occurred or more than 6 months in cases with severe prognosis. The patients in the Hato study had an average Yanagihara score of 15 when rating their facial palsy (which falls between House–Brackmann grades 4 and 5).

The overall rate of recovery of those treated with valacyclovir and prednisolone (96.5%) was significantly better (P<.05) than the rate among those treated with placebo and prednisolone (89.7%). In cases of complete or severe palsy, the rates of patients treated with both agents vs prednisolone alone who recovered were 95.7% (n=92) and 86.6% (n=82) (P<.05; NNT=11).

One big difference between the Sullivan and Hato studies is that the patients recruited for the Hato study had much more severe facial palsy (rated between 4 and 5) than in the Sullivan study (average=3.6), which suggests that there may be a use for valacyclovir in treating patients with complete facial palsy.⁶ Patients were all recruited from tertiary care centers as opposed to mainly from primary care settings as in the Sullivan study, consistent with the greater severity of cases in the Hato study.

Outcome assessors were not blinded to treatment assignments or stage of assessment in the Hato study, raising major concerns about the validity of the findings given the nature of facial paralysis as an outcome measure. We find the Sullivan study a more rigorous and convincing study. Nonetheless, future research may verify their findings and support the use of valacyclovir in the most severe cases of Bell’s palsy. For now, we are not convinced.

CHALLENGES TO IMPLEMENTATION: Easy to put into practice

Thankfully, some changes in practice are easy to implement. This is one of them. For those who prefer to prescribe prednisone, the dose of prednisolone used in the study, 25 mg bid, is equivalent to 60 mg of prednisone.

PURL surveillance system methodology
This study was selected and evaluated using FPIN’s Priority Updates from the Research Literature (PURL) surveillance system methodology.