Twenty RCTs were included (n=882 children), with sample sizes ranging from 12 to 122. Six trials reported no adverse events; the remaining 14 reported 542 adverse events of 35 different types; the authors stated that the adverse events were not clearly defined. Trials met between 12 and 19 of the quality items (average 16.3) and the duration of follow-up was between six weeks and three years.
Botulinum toxin type A was associated with a statistically significant increase, compared with control, in the following adverse events: respiratory tract infection (RR 15.82, 95% CI 3.86 to 64.83; four RCTs); bronchitis (RR 11.74, 95% CI 2.31 to 59.59; three RCTs); pharyngitis (RR 7.50, 95% CI 1.78 to 31.61; three RCTs); asthma (RR 6.40, 95% CI 1.20 to 34.00; three RCTs); muscle weakness (RR 5.60, 95% CI 1.44 to 21.84; three RCTs); urinary incontinence (RR 5.30, 95% CI 1.20 to 23.52; three RCTs); falls (RR 5.17, 95% CI 1.74 to 15.36; five RCTs); seizures (RR 4.24, 95% CI 1.85 to 9.71; five RCTs); fever (RR 2.77, 95% CI 1.04 to 7.34; three RCTs); unspecified pain (RR 2.44, 95% CI 1.39 to 4.27; 10 RCTs); vomiting (RR 3.99, 95% CI 1.39 to 11.48; six RCTs); and viral upper respiratory tract infection (RR 5.91, 95% CI 1.07 to 32.46; three RCTs). Data were also presented for another 12 adverse events that were reported only in single trials.
Sensitivity analysis, using the two correction factors, suggested that all the adverse events were underestimated because no cases were reported in the exposed groups, in some of the RCTs.