BACKGROUND: Recent studies have shown that a short course of a macrolide or cephalosporin is just as effective as 10 days of penicillin. However, there has been concern that a shorter course of antibiotics would lead to an increase in poststreptococcal sequelae. This study was undertaken to determine if a shorter course of antibiotics would lead to an increased risk of acute rheumatic fever and glomerulonephritis.
POPULATION STUDIED: This study included a total of 5318 children aged 1 to 18 years from 137 pediatric practices in Germany. Inclusion criteria included temperature of at least 38°C and 1 or more of the following: exudate of the pharynx or tonsils, erythema of the pharynx or tonsils, and cervical adenopathy. They also had to have a positive rapid antigen test for Group A beta-hemolytic streptococcus (GABHS). Exclusion criteria included a negative culture for GABHS, recent antibiotic use, and a history of rheumatic fever or glomerulonephritis in the patient or a person in the same household. The racial make-up of the patients is not given, and this may limit the generalizability of this study to family physician practices in the United States.
STUDY DESIGN AND VALIDITY: This was a randomized open-label design in which patients were randomized to either 10 days of penicillin V (50,000 IU/kg/day with a maximum of 2,250,000 IU/day) or 5 days of one of the following: amoxicillin clavulanate, ceftibuten, cefuroxime axetil, loracarbef, clarithromycin, or erythromycin estolate. The patients were evaluated 2 to 4 days after the end of treatment, 7 to 9 days after the end of treatment, at 7 to 8 weeks, and at 6 and 12 months. Rheumatic fever was diagnosed on the basis of the updated Jones criteria and glomerulonephritis based on the pediatrician’s assessment. The number of patients followed up for 12 months (n=4077) was a little short of the sample size required (n=4710) to detect a difference of 3 per 1000 versus 9 per 1000 in the incidence of rheumatic fever or glomerulonephritis. Other limitations of the study include the open-label design and the failure to describe the method of allocation in more detail. Since the patients and physicians knew the group assignment, there may have been an expectation that those receiving the newer, “better” drugs would do better clinically. The study was supported by the companies making these newer drugs.
OUTCOMES MEASURED: The primary outcome was the probability of rheumatic fever or glomerulonephritis in the study groups. Secondary outcomes included clinical response as judged by the investigators, resolution of symptoms, microbiologic cure, and safety.
RESULTS: Three patients developed rheumatic fever, and 2 had glomerulonephritis during the study period (4 in the larger 5-day group). Analysis of the individual cases led the authors to conclude that none of the sequelae were related to the case of streptococcal pharyngitis treated in the study. The clinical response rates were similar at the first follow-up visit (94.5% vs 93.4%). At the 7-to 9-day visit 39.7% of the 5-day patients and 45.6% of the 10-day patients still had symptoms (P <.001). The recurrence rates between the 2 groups were 24.4% in the 10-day group and 21.9% in the 5-day group (P <.03). They also found a 6% resistance rate and 6.8% intermediate resistance rate to macrolides among the isolates in their study.
Although somewhat underpowered, this study supports the hypothesis that a 5-day course of one of the study antibiotics in children is as effective as a 10-day course of penicillin V in the treatment of GABHS pharyngitis. However, the widespread use of broad-spectrum agents for a common infection is a significant concern in an age of increasing bacterial antibiotic resistance. For patients who have documented streptococcal pharyngitis and are allergic to penicillin, the use of a 5-day course of one of the study antibiotics is reasonable.