Six RCTs (3,927 children) were included in the review. Three trials were of respiratory syncytial virus immune globulin therapy (1,096 children, 533 randomised to treatment groups) and three were of palivizumab therapy (2,831 children, 1663 randomised to treatment groups). The median Jadad score was 4 points (range 1 to 5).
Children who received prophylactic respiratory syncytial virus immune globulin or palivizumab were approximately half as likely to be hospitalised compared with children who received placebo or no treatment (RR 0.53, 95% CI 0.43 to 0.66). The effect was of similar magnitude for respiratory syncytial virus immune globulin compared with placebo or no treatment (RR 0.59, 95% CI 0.42 to 0.83), as for palivizumab compared with placebo or no treatment (RR 0.50, 95% CI 0.38 to 0.66). The number-needed-to-treat with palivizumab to prevent one hospital admission was 20 (95% CI 14 to 33).
Treatment with a prophylactic agent (respiratory syncytial virus immune globulin or palivizumab) was associated with a significant reduction in intensive care unit admission (RR 0.39, 95% CI 0.21 to 0.70); in the stratified analysis this effect was stronger for palivizumab than for respiratory syncytial virus immune globulin therapy. There was a suggestion that treatment was associated with a lower risk of respiratory syncytial virus infection (RR 0.78, 95% CI 0.60 to 1.01), although this was not statistically significant. Neither agent was associated with mechanical ventilation or all-cause mortality.
There was a suggestion that the effect of prophylaxis on the risk of hospitalisation for respiratory syncytial virus was more effective in babies born premature, than in those with chronic lung disease or congenital heart disease.
The authors stated that adverse effects were rarely associated with palivizumab. However, respiratory syncytial virus immune globulin was associated with fluid overload (three RCTs), decreased oxygen saturation or cyanosis (two RCTs), fever (two RCTs) and respiratory distress (two RCTs).
The authors stated that there was no evidence of publication bias.