Tajeu et al. (1) reviewed all 2010 issues of Obesity and International Journal of Obesity to investigate how often odds ratios were misinterpreted as relative risks (as odds and probabilities are not equivalent in situations where the outcome is not rare).
Of the 855 articles examined, 62 (7.3%) presented, and these were presented incorrectly in 23.2% of the articles. Clinical articles presented ORs more correctly than social science or basic science articles. Almost one-quarter of the studies that presented ORs had misinterpreted them.
The authors conclude that researchers should carefully present interpretable measures of association including RRs and risk differences to minimize confusion and misrepresentation of results.
1. Tajeu GS, Sen B, Allison DB, Menachemi N. Misuse of odds ration in obesity literature: an empirical analysis of published studies. 2012; 20:1726-1731.