Journal of Medical Sciences (Jan 2020)

A superior odds ratio compared to the risk ratio when estimating moderator effects in meta-regression analyses of randomized controlled trials

  • Chih-Chien Chiu,
  • Chien-Fu Chen,
  • Po-Jen Hsiao,
  • Dung-Jang Tsai,
  • Hsueh-Lu Chang,
  • Wen-Hui Fang,
  • Wei-Teing Chen,
  • Jenq-Shyong Chan,
  • Min-Tser Liao,
  • Yi-Jung Ho,
  • Wen Su,
  • Ying-Kai Chen,
  • Hui-Han Hu,
  • Zheng-Zong Lai,
  • Chin Lin

DOI
https://doi.org/10.4103/jmedsci.jmedsci_133_19
Journal volume & issue
Vol. 40, no. 3
pp. 119 – 126

Abstract

Read online

Background: Moderator effect assessment is important in personalized medicine. We mathematically prove that the average summary value is actually nonlinearly to logRR, and we assess the bias from linear meta-regression on logRR via simulation. Methods: In the meta-analysis of randomized controlled trials, the moderator effect is generally evaluated by the linear meta-regression of the logarithmic risk ratio (RR) versus the average summary value of the entire study population. Conclusions: We recommend using linear meta-regression on logarithmic odds ratio (logOR) since it has been shown that the average summary value is actually linear to logOR.

Keywords