I am conducting a meta-analysis and have one study that reports PR (Prevalence ratio and 95% CI). I would like to convert the PR to odds ratios and obtain corresponding confidence intervals. The studies were all cross-sectional and cohort studies, and there were no frequencies or percentages.
Dear Dung Phuong Dang ,
- Prevalence is the number or proportion of subjects with some outcomes (events) at a point in time (or period of time). Therefore, the Prevalence Ratio (PR) indicates how large is the prevalence of an outcome (event) in one group compares with another group.
- On the other hand, the Risk estimates the average probability of an outcome over a specified period of time among individuals at risk for having the outcome at the beginning of the follow-up period. So the Ratio of the Risks indicates how more or less likely one group of exposed subjects is to develop/acquire a health condition over the follow-up period compared with the other unexposed group.
Hence, the Prevalence Odds Ratio and Risk Odds Ratio could not be interchangeable.
You can read these articles:
Article Prevalence odds ratio versus prevalence ratio: choice comes ...
Article Prevalence odds ratio or prevalence ratio in the analysis of...
In your case, I think you can "estimate" OR from PR if you know the prevalence of disease and the prevalence of exposure; of course, this result would not be 100% as the genuine one because this is an estimate.
Article Relationship between prevalence rate ratios in cross-sectional studies
Hello Mr. Phong!
Thanks for your answer. I know the prevalence of disease and the prevalence of exposure, but then they were included in the multivariable analysis. I only extracted OR, PR, or RR and 95% CI data.
My advisor suggested converting PR to beta coefficient and then convert beta coefficient to OR. But this way I can't find any information.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis