Commentary to Gebel 2012: A quantitative review should apply meta-analytical methods—and this applies also to quantitative toxicological reviews

Abstract

Gebel (In Arch Toxicol 87(5):923-924, 2013) replied to my Letter to the Editor (Morfeld in Arch Toxicol 87(5):921, 2013) in which I criticized the quantitative review of inhalation rat studies on the association of granular biopersistent dust exposures and lung cancer risk (Gebel in Arch Toxicol 86(7):995-1007, 2012). My methodological comments were not understood. The Editors of the Archives invited me to detail and substantiate my criticism. The main issues are as follows: (1) A quantitative summary of the study results was performed without weighting for precision of the single studies (the published unweighted synthesis is potentially biased). (2) No heterogeneity assessment was performed before combining the findings (it is unclear whether overall summaries are sensible). (3) Correlation coefficients were used (correlations are distorted estimates of exposure-response and misleading). (4) An incomplete input data table was published (no transparent reporting, no replication possible for the reader). The quantitative synthesis by Gebel (In Arch Toxicol 86(7):995-1007, 2012) does not fulfil the usual requirements of a scientific quantitative review and should be replaced by an appropriate meta-analysis.