We used several methods for quality check of the meta-analysis (Supplementary File 2). Baujat plots, influence plots & the leave-one-out method of sensitivity analysis were used to look for studies with prevalence estimates that were outliers, along with their weightage. Data extraction from these studies were double-checked. If the study was found to be an outlier with a considerable effect on the pooled prevalence estimate for a particular comorbidity, then it was excluded from the analysis of only that comorbidity.
