Performance and coverage of HIV interventions for injection drug users: Insights from triangulation of programme, field and surveillance data from Pakistan

Introduction

Nearly 20% of the estimated 84,000 injecting drug users in Pakistan are HIV infected. Non-governmental organisations have implemented HIV interventions for IDUs in 7 cities in Pakistan. Here we report on the performance, coverage and costs of these interventions.

Methods

National HIV bio-behavioural surveillance data were used to measure effectiveness of interventions by comparing HIV prevalence and behavioural trends in intervention and non-intervention cities. Coverage was measured by comparing the supply of syringes with the total syringe need and intervention cost per IDU served per year was calculated.

Results

The NGOs registered 20,640 IDUs (original targets: <10,000); provided 66% of new syringes for all registered IDUs and 75% for all estimated street-based IDUs. This compared to a national coverage of about 13%. Intervention cities had higher baseline HIV prevalence, reflecting their choice as intervention sites. More IDUs from intervention cities (59% vs. 27%) reported always using a clean syringe. Condom use with last sexual partner (24% vs. 11%) and HIV prevention knowledge were also higher amongst this group (all at p < 0.001). HIV prevalence in intervention cities remained unchanged in Faisalabad (13%) and Quetta (10%) but increased in Karachi (26-30%) and Lahore (4-7%). Coverage of sterile syringes for intervention cities was 30% compared to 13% nationwide. However within city, coverage varied from 30 to 99%. The costs of services varied widely by NGOs from USD 146 to 403.

Conclusions

IDUs interventions are performing well in some Pakistani cities. However, considerable expansion is needed to increase nationwide coverage.