Incidence of human immune deficiency virus among injection drug users enrolled in a harm reduction program at Karachi

Date of Award


Document Type


Degree Name

Master of Science in Epidemiology & Biostatistics (MSc Epidemiology & Biostats)


Community Health Sciences


Background HIV in Pakistan has transitioned from low to concentrated level epidemic (>5%) among injection drug users (IDUs). Although the HIV prevalence among general population is still low (<0.1%), yet the prevalence is higher among high risk groups (HRG). We have sufficient surveillance data on prevalence but there is dearth of data on incidence rate of HIV. Since the epidemic in the country is mainly driven by the IDUs, the current study captured the incidence of HIV in this group. Objectives The primary objective was to estimate the HIV incidence rate among IDUs enrolled in harm reduction program at Karachi. The secondary objectives were to estimate the risk factors associated with HIV sero-conversion and predictors of loss to follow up among IDUs in the harm reduction program. Methods A quantitative and qualitative study was conducted among IDUs registered in a harm reduction program in Karachi from March-December 2009. We used descriptive cohort design for incidence rate and nested case-control study for factors associated with HIV sero-conversion. We recruited 474 (IDUs) who were HIV sero-negative at the baseline. The cohort was re-tested for HIV status from April-June 2011. For factors associated with HIV sero-conversion, the HIV positive IDUs were taken as cases and HIV negative as controls. Focus group discussions (FGDs) with IDUs and in-depth interviews with the program's implementers were also conducted to gather perceptions regarding reasons for HIV sero-conversion and loss to follow-up. Findings The incidence rate of HIV among IDUs was 13 per 100 person-years (95% CI: 10.92 — 15.67). The mean age of the clients was 31+8 years. In multivariate regression analysis unemployment (AOR: 2.4, 95% CI 1.30-4.56, p-value <0.01), Injecting drugs at streets (AOR: 5.29, 95% CI 1.36-20.59, p-value 0.01), sharing of syringes (AOR: 5.7, 95% CI 2.89-11.37, p-value <0.01), not getting syringes from drop-in center (AOR: 5.0, 95% CI 2.73- 9.17, p-value <0.01); frequency of daily drug use (AOR: 2.0, 95% CI 1.40-2.85, p-value <0.01), were significantly associated with HIV sero-conversion, after adjusting for other variables in the model. In multivariate analysis for the predictors of loss to follow-up, the enrollment of the clients through sources other than the outreach (AOR: 4.7, 95% CI 2.74-7.93, p-value <0.01), being foreigner (AOR: 13, 95% CI 4.86-39.57, p-value <0.01), being non-Sindhi (AOR: 2.6, 95% CI 1.52-4.50, p-value <0.01), being illiterate (AOR: 2.3, 95% CI 1.34-4.11, p-value 0.003), history of incarceration (AOR: 1.8, 95% CI 1.05-3.00, p-value 0.031), being physically disable (AOR: 22.5, 95% CI 1.92-262.31, p-value 0.01), and getting syringes from pharmacies and friends (AOR: 2.0, 95% CI 1.06-3.86, p-value 0.03) were significantly associated with loss to follow-up, after adjusting for other variables in the model. In qualitative analysis, themes were generated and text was coded on the themes. The perceptions regarding reasons for sero-conversion were consistent with the quantitative analysis. In contrast, completely different predictors of loss to follow up were identified in the qualitative arm of the study. Conclusion Although the study included the IDUs associated with a harm reduction program, the incidence HIV was high, the situation calls for public health and policy measures to curb the epidemic

This document is available in the relevant AKU library