Random effects model with only p-values

Hello statisticians,

I’m working on a systematic review of the risk factors that predict post-operative infection.

There are a number of risk factors e.g. Age, sex, weight, etc etc. I am attempting to synthesize the results from multiple studies to give an overall measure of effect for each risk factor. I’ve been reading on random vs fixed effects models but have only encountered conceptual descriptions. I’m trying to learn how to do the actual analysis (I use R) to get an overall measure of effect in different scenarios:

  1. For some of the risk factors, the OR, CI and p-value for both univariate and multivariate analysis is stated in all papers

  2. For some risk factors, I only have the multivariate OR, CI and p-value.

  3. For other risk factors, only the p-value is available

With thanks

Please have a look here for prognostic reviews: https://methods.cochrane.org/prognosis/tools

1 Like