Multiple-outcome meta-analysis of clinical trials.

摘要: When several clinical trials report multiple outcomes, meta-analyses ordinarily analyse each outcome separately. Instead, by applying generalized-least-squares (GLS) regression, Raudenbush et al. showed how to the outcomes jointly in a single model. A variant of their GLS approach, discussed here, can incorporate correlations among within treatment groups and thus provide more accurate estimates. Also, it facilitates adjustment for covariates. In our study need not all nor evaluate treatments. For example, meta-analysis may two or treatments (one ‘treatment’ be control) include randomized controlled that on any subset (of one more) interest. The analysis omits other these evaluated but are interest meta-analyst. proposed fixed-effects regression model, study-level treatment-arm-level covariates predictors outcomes. An rheumatoid arthritis data from second-line drug (used after initial standard therapies prove unsatisfactory patient) motivates applies method. Data 44 were used effectiveness injectable gold auranofin three tender joint count, grip strength, erythrocyte sedimentation rate. model quality duration trial baseline measures patients' disease severity activity trial. found was significantly effective than estimated coefficients, multiple-outcomes produced moderate changes values slightly smaller errors, separate models.