Eleven RCTs were included in the review (n=2,389 patients in intention-to-treat population; sample size ranged from 60 to 520). Four trials were double-blinded and one was single-blinded. Three trials scored more than 2 out of 5 on the Jadad scale; the other trials scored 2.
Primary outcome (eight RCTs, n=1,607 patients): Short-course antibiotic regimens were associated with a statistically significant reduction in microbiological eradication of group A beta-haemolytic streptococcus (GAS) from the throat at the end of treatment compared with long-course regimens (OR 0.49, 95% CI 0.32 to 0.74). No significant heterogeneity was found. Short courses (lasting five to seven days) of penicillin V were associated with lower rates of microbiological eradication than long courses (OR 0.36, 9%% CI 0.13 to 0.99; three RCTs, n=500 patients). There was no statistically significant difference for short versus long courses of cephalosporins (four RCTs, n=1,018 patients). Short-course regimens were associated with significantly lower rates of microbiological eradication than long courses in trials of predominantly children or adolescents (OR 0.63, 95% CI 0.40 to 0.98; six RCTs, n=1,258 patients) and in blinded RCTs (OR 0.31 (95% CI 0.16 to 0.58; three RCTs, n=469 patients).
Secondary outcomes: Short-course regimens were associated with lower rates of clinical success (OR 0.49, 95% CI 0.25 to 0.96; five RCTs, n=1,217 patients) and higher rates of bacteriological recurrence (OR 3.02, 95% CI 1.06 to 8.56; three RCTs, n=698 patients) than long-course regimens, but there was no significant difference in bacteriological relapse (five RCTs, n=981 patients) or adverse events (three RCTs, n=879 patients).
No significant heterogeneity was found for the above analyses.
Results for other analyses were reported in the paper.