I tried to do a meta-analysis of diagnostic accuracy study. I found out that I have a publication bias based on the Deek's regression test. Thus, I want to supplement my analysis with P-curve analysis which I read from here. The author used the P-curve analysis with interventional studies.
I have tried to the P-curve analysis with a diagnostics odds ratio (DOR) model, however I am quite doubtful whether the P-curve analysis is appropriate in this meta-analysis.
I tried to find any example or paper that uses P-curve analysis with meta-analysis of diagnostic accuracy study, I have found none so far. Any opinion is appreciated.
Thanks
