Top POEMs of 2012: Infectious Disease

Honey Improves Nocturnal Cough in Children

Clinical question: Can honey decrease nighttime cough and improve sleep in children with upper respiratory tract infection?

Bottom line: A teaspoonful of honey, given alone or with a noncaffeinated liquid before bedtime, decreases cough frequency and severity while improving the sleep of both parents and the child with acute cough. Placebo also works, but not as well. Both (honey or placebo) give parents an active role in their child's well-being while not exposing the child to potentially harmful medicines. (LOE = 1b)

Reference: Cohen HA, Rozen J, Kristal H, et al. Effect of honey on nocturnal cough and sleep quality: a double-blind, randomized, placebo-controlled study. Pediatrics 2012;130(3):465-471.

Study design: Randomized controlled trial (double-blinded)

Funding source: Industry + govt

Allocation: Concealed

Setting: Outpatient (primary care)

Synopsis: There are 181 compounds in typical honey (who knew?). The Israeli investigators conducting this study recruited 300 children 1 year to 5 years of age (median age = 29 months) presenting to pediatric practices with nocturnal cough for less than 7 days. The parents completed a questionnaire regarding their assessment of the child's cough and sleep difficulty on the previous night. Children with a severity score of at least 3 (out of a possible 7) were randomized, using concealed allocation, to a receive a single dose of 1 of 3 types of honey or a sweet-tasting placebo (date extract) at bedtime. Parents of 90% of the children completed a follow-up questionnaire the next day. Cough frequency, as reported by the parents, was significantly less in all 4 groups (including the placebo group). But the decrease was significantly more in all 3 honey groups. Similarly, cough severity, "bothersomeness," children's and parents' sleep, and combined symptom scores were significantly improved with honey as compared with placebo. This was only a single-dose study and the significant improvement with placebo emphasizes the effect of a parent's active role on their impression of their child's symptoms. A Cochrane review also found honey to be better than no treatment and perhaps better than diphenhydramine (Benadryl). There were significantly more dropouts in the children randomized to receive eucalyptus honey or citrus honey, which the authors speculate might be due to the strong taste (the third honey was produced from a mint-pollen honey and had a milder taste). The mechanism of action is unknown but may be a central effect due to influence on sensory nerves that initiate cough.

Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine
Tufts University
Boston, MA

Rules for Strep Throat Validated

Clinical question: Are strep throat decision rules effective in ruling out children and adults who do not need to be tested for group A beta-hemolytic Streptococcus as a cause of their sore throat?

Bottom line: Two commonly used "strep scores" are as effective as advertised for determining children and adults with a low likelihood of streptococcal pharyngitis. Their proper use, advocated in the U.S. by the Centers for Disease Control, can decrease costs by avoiding unnecessary testing. (LOE = 1a)

Reference: Fine AM, Nizet V, Mandl K. Large-scale validation of the Centor and McIsaac scores to predict Group A Streptococcal pharyngitis. Arch Intern Med 2012;172(11):847-852.

Study design: Decision rule (validation)

Funding source: Government

Setting: Population-based

Synopsis: To test two different "strep scores," the authors analyzed data collected from 206,870 patients 3 years or older who presented with a sore throat to one of 581 "Minute-Clinics," in-pharmacy health clinics providing acute care for minor illness. Clinicians in these clinics follow, with >99% adherence, an acute pharyngitis protocol that requires collecting of signs and symptoms before rapid antigen testing for Group A beta-hemolytic Streptococcus. The authors compared the test characteristics for the Centor and McIsaac rules, two scoring systems that estimate the likelihood of strep based on clinical presentation. For patients 15 years or older, 7% with a Centor score of 0 and 12% with a score of 1 had strep (overall prevalence of strep was 23%). For patients 3 years or older, 8% with a McIsaac score of 0 and 14% with a score of 1 had strep throat. Both of these sets of percentages are somewhat higher than the results of earlier studies used to validate these rules. Centor rule -- Give one point for each of the following: fever, absence of cough, presence of tonsillar exudate; and swollen, anterior cervical nodes. Scores of 0 or 1 do not need to be tested or treated, according to the Centers for Disease Control. The McIsaac rule adds 1 point for children <15 years old and subtracts a point for adults >45 years.

Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine
Tufts University
Boston, MA

Azithromycin Associated with Increased Risk of CV Death

Clinical question: Does use of azithromycin increase the risk of cardiovascular death?

Bottom line: For every 1 million courses of azithromycin (Zithromax) that are prescribed to adults, there are an additional 49 deaths (number needed to treat to harm [NNTH] = 20,400), mostly due to sudden cardiac death. The increase in risk is even greater among those at high baseline risk for cardiovascular death (NNTH = 4081). One more reason to avoid inappropriate use of antibiotics, and to use amoxicillin instead of azithromycin when appropriate. (LOE = 2b)

Reference: Ray WA, Murray KT, Hall K, Arbogast PG, Stein CM. Azithromycin and the risk of cardiovascular death. N Engl J Med 2012;366(20):1881-1890.

Study design: Cohort (retrospective)

Funding source: Government

Setting: Population-based

Synopsis: Previous reports have linked macrolides to ventricular arrhythmias and sudden cardiac death, and there have been a number of case reports of torsades de pointes in patients given azithromycin. The authors used the Tennessee Medicaid database that links information about patients and their diagnoses with prescribing data. They identified patients who had been given azithromycin, and used propensity scoring to match them with similar patients who had not received an antibiotic (4 control patients for every case) or who had received an alternative antibiotic (levofloxacin, amoxicillin, or ciprofloxacin). Participants were limited to adults, aged 30 years to 74 years, without life-threatening noncardiovascular illness, recent hospitalization, or nursing home residence. The primary outcome was cardiovascular death, and secondary outcomes were sudden cardiac death and all-cause mortality. They found an increase in all-cause mortality (57 vs 106 per 1 million courses of antibiotic; hazard ratio [HR] = 1.85; 95% CI 1.25 - 2.75), total cardiovascular deaths (HR = 2.88; 1.79 - 4.63) and sudden cardiac death (HR = 2.71; 1.58 - 4.64). This increase in risk was not seen with the other antibiotics. The increase in risk was consistent across cardiovascular risk groups, so the absolute increase in death was greatest among those at the highest baseline risk (245 additional deaths per million courses of the drug).

Mark H. Ebell, MD, MS
Associate Professor
University of Georgia
Athens, GA

POEMs are provided by Essential Evidence Plus, a point-of-care clinical decision support system published by Wiley-Blackwell. For more information, please see