Comparative Effectiveness of Medical Interventions in Adults Versus Children
Received 14 August 2009; received in revised form 15 January 2010; accepted 9 February 2010. published online 03 May 2010.
Objective
To estimate the comparative effectiveness of medical interventions in adults versus children.
Study design
We identified from the Cochrane Database of Systematic Reviews (Issue 1, 2007) meta-analyses with data on at least 1 adult and 1 pediatric randomized trial with binary primary efficacy outcome. For each meta-analysis, we calculated the summary odds ratio of the adult trials and the pediatric trials, respectively; the relative odds ratio (ROR) of the adult versus pediatric odds ratios per meta-analysis; and the summary ROR across all meta-analyses. ROR <1 means that the experimental intervention is more unfavorable in children than adults.
Results
Across 128 eligible meta-analyses (1051 adult and 343 pediatric trials), the summary ROR did not show a statistically significant difference between adults and children (0.96; 95% confidence intervals, 0.86 to 1.08). However, in all meta-analyses except for 1, the individual ROR's 95% confidence intervals could not exclude a relative difference in efficacy over 20%. In two-thirds, the relative difference in observed point estimates exceeded 50%. Nine statistically significant discrepancies were identified; 4 of them were also clinically important.
Conclusions
Treatment effects are on average similar in adults and children, but available evidence leaves large uncertainty about their relative efficacy. Clinically important discrepancies may occur.
ABSTRACT