If a study provided several effect measures, such as unadjusted and adjusted effect measures, the most completely adjusted effect measure was used.
Only two studies provided adjusted effect estimates [14], [16].
HIV incidence was higher in women with bacterial vaginosis (adjusted effect 1.57, 95%CI 1.26, 1.94, I2 19.0%) and Trichomonas vaginalis (adjusted effect 1.64, 95%CI 1.28, 2.09, I2 0.0%).
We used multilevel logistic regression to obtain adjusted effect estimates.
Influence of removing studies one by one on adjusted effect estimates of patients requiring supplementary analgesia.
Additional file 2 provides adjusted effect sizes for all LPS items.
In the case of confounding, adjusted effect estimates will also be reported.
We used the generic inverse variance method to pool maximally adjusted effect estimates.
In the primary meta-analyses, we will use the maximally adjusted effect sizes and 95% CIs.
(eTable 5, http://links.lww.com/EDE/A805, shows minimally and fully adjusted effect estimates for comparison).
We also adjusted effect estimates for educational attainment prior to cannabis use to address the issue of reverse causation.
