1
$\begingroup$

I'm working on a meta-analysis and one study provides odds ratios without standard errors, confidence intervals or exact p-values.

The only information provided is odds ratios, whether the p-value was greater than or less than 0.05, and the study sample size.

Is there any way I can calculate a standard error from an odds ratio from the effect size and sample size?

$\endgroup$

1 Answer 1

2
$\begingroup$

Simple answer is No.

Here's some options:

  1. contact author for missing SD.

  2. use the largest SE from other studies in the same analysis then calculate the SD based on the reported N. This will allow you to use yhe study while giving it weight in the analysis as the least weighted study.

  3. use an SE from a similar study in the same analysis (e.g. similar sample size, intervention details, etc.)

In all cases conduct a sensitivity analysis removing this study to test the robustness of the analysis and effect of the assumptions.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.