Unadjusted (bivariate) and multivariable ordinal logistic regression analysis of predictors of disease prevalence for ordinal outcome in three categories of increasing disease severity (No CIN, CIN1, CIN2, ≥CIN3) confirmed by composite colposcopic-histopathological diagnosis among HIV-infected wome

Abbreviations: OR: Odds ratio; CI: Confidence intervals; INR: Indian Rupees; BMI: Body Mass Index; ART: Antiretroviral therapy; STI: Sexually transmitted infection; WHO: World Health Organization.
*The number of covariates that could be included in the multivariable regression model was dependent on the number of event outcomes for that analysis to avoid over-fitting of the models. Hence only 10 covariates were included in the multivariable model. Covariates were deemed scientifically important and selected a priori along with covariates that were significant in unadjusted (bivariate) analysis.
∧The exchange rate at the time of the study was approximately 45 INR (Indian Rupees) per US$ such that Indian Rupees 1,000 are the equivalent of US$ 22.