The effects of clinical and statistical heterogeneity on the predictive values of results from meta-analyses W.G. Melsen, M.C.J. Bootsma, M.M. Rovers, M.J.M. Bonten Clinical Microbiology and Infection Volume 20, Issue 2, Pages 123-129 (February 2014) DOI: 10.1111/1469-0691.12494 Copyright © 2014 European Society of Clinical Infectious Diseases Terms and Conditions
FIG. 1 Results of the Monte Carlo simulation study. The associations between the I2-statistic and the probability (%) that the true average relative risk (a) and the result of the reference study (b) are contained in the 95% CI of the meta-analysis, according to different levels of heterogeneity. The dotted line represents the average I2-statistic. The coloured lines represent the probabilities with a fixed-effects model (red) and a random-effects model (blue). MH, calculations based on a fixed-effects model with the Mantel–Haenszel method; RMH, calculations based on a random-effects model with the Mantel–Haenszel method. Clinical Microbiology and Infection 2014 20, 123-129DOI: (10.1111/1469-0691.12494) Copyright © 2014 European Society of Clinical Infectious Diseases Terms and Conditions
FIG. 2 Clinical scenarios. Clinical Microbiology and Infection 2014 20, 123-129DOI: (10.1111/1469-0691.12494) Copyright © 2014 European Society of Clinical Infectious Diseases Terms and Conditions