Selective reporting of adjusted estimates in observational epidemiology studies: reasons and implications for meta-analyses

Abstract

For meta-analyses of observational epidemiology studies, unadjusted and adjusted study estimates are often extracted. However, there is evidence of selective reporting of adjusted study estimates. We investigate adjustment reporting bias, examining the reasons why some studies do not contribute an adjusted estimate to a meta-analysis. Ten published meta-analyses were re-analysed to assess evidence of adjustment reporting bias and over 100 primary studies were read to investigate why they did not contribute an adjusted estimate to a meta-analysis. Selective reporting of adjusted estimates may lead to a bias in some meta-analyses when adjusted study estimates are not reported because univariate analyses indicated a non-significant effect. We recommend that unadjusted and adjusted study estimates be extracted for a meta-analysis. If adjusted estimates cannot be obtained, the reasons for this should be investigated and sensitivity analyses could be used to assess the impact of this on the meta-analysis.