The metanalysis was done for the Sensitivity, Specificity, LR+, LR-, and DOR of each outcome. But the reviewer asked me to calculate and include the PPV and NVP too, however I don't think it is possible because the prevalence of the outcome (difficult laryngoscopy) is very variable (between 1,5 and 20% in the literature). What should I do?
should I calculated it with a median? or just not calculated it all.
I will put a picture of one metanalysis
Thanks
Hi David, to perform a meta-analysis of diagnostic tests, in addition to the graph you present (where the values farthest from 0 will represent the largest area under the ROC curve)
you must include a summary measure, I think this link (https://jbi-global-wiki.refined.site/space/MANUAL/3283911144/Appendix+9.3+Meta-analysis+equations+and+models) will help you better find a summary measure or the findings
No. PPV and NPV are dependent on the incidence in the study and have no meaning beyond the study itself. The vary. It is better to use likelihood ratio positive and likelihood ratio negative as these are independent of incidence
Yes Patrick. Completely agree it would be innaproppiate to put NPV and PPV because the high variability in prevalanece of the outcome. Tough telling the reviewer its not a good thing to do. Thanks!
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning