Estimation of Tx Outcome Impacted by Meta-Analysis Strategy
(HealthDay News) — Estimation of treatment outcome varies with the meta-analysis strategy, according to a study published in the August 13 issue of the Journal of the American Medical Association.
Agnes Dechartres, MD, PhD, from the Centre de Recherche Epidémiologie et Statistique, INSERM U1153 in Paris, and colleagues compared treatment outcomes estimated by meta-analysis of all trials and alternative analytic strategies. For each meta-analysis, the difference in treatment outcomes between meta-analysis of all trials and alternative strategies (single most precise trial and various limited meta-analyses) was expressed as a ratio of odds ratios (ROR).
The researchers found that there were larger treatment outcomes in the meta-analysis of all trials vs. the single most precise trial (combined ROR, 1.13 and 1.03 for subjective and objective outcomes, respectively). In 51% of meta-analyses of subjective outcomes and 39% of meta-analyses of objective outcomes there was a substantial difference in treatment outcomes between the strategies. For subjective and objective outcomes, the combined ROR for meta-analysis of all trials was, respectively, 1.08 (95% confidence interval, 1.04–1.13) and 1.03 (1.00–1.06) compared with meta-analysis of the 25% largest trials; 1.17 (1.11–1.22) and 1.13 (0.82–1.55) compared with limit meta-analysis; and 0.94 (0.86–1.04) and 1.03 (1.00–1.06) compared with meta-analysis restricted to trials at low risk of bias.
"This instability in findings can result in major alterations in the conclusions derived from the analysis and underlines the need for systematic sensitivity analyses," the authors write.