HIV prevalence in blood donors and recipients in Pakistan: a meta-analysis and analysis of blood-bank data.

BACKGROUND The first instances of HIV-antibody detection in donated blood in Pakistan were reported in 1988. Since then, documentation of HIV in blood donors and of rates of transmission via transfusion has been limited. Previously assumed to have a low prevalence, HIV is an increasing health concern in Pakistan. Since there is no national, centralized blood-banking system, there are no reliable data on which to base estimated risks of transfusion-associated HIV infection. This study was therefore conducted to estimate the prevalence of HIV in blood donors and recipients in Pakistan between 1988 and 2012. METHODS Meta-analyses were undertaken of reported prevalences of HIV in blood donors and recipients published during 1988-2012. Papers were identified by searching PubMed, Google, CINAHL and PakMediNet and the websites of the World Health Organization, the national HIV/AIDS Surveillance Project and the National AIDS Control Programme of Pakistan. In addition, the 1998-2012 records of the Aga Khan University blood bank were analysed. RESULTS The 254 abstracts identified at the preliminary search were reviewed and, after removal of duplications, case-reports, editorials and reviews, 32 papers were selected that met the inclusion criteria. All studies that reported on HIV antibodies in blood donors/recipients were included, irrespective of the methodology used. Since seroconversion had only been confirmed through supplemental testing in a few papers, the results were analysed separately for reports based on screening only and confirmed cases. A total of 142 of 2 023 379 blood donors and 4 of 3632 blood recipients were HIV positive, giving an overall pooled seroprevalence of 0.00111% in blood donors and 0.00325% in blood recipients. The annual prevalences of HIV in donors at the Aga Khan University blood banks were similar, ranging from 0.013% to 0.116%. CONCLUSION Very few reports on HIV in blood donors in Pakistan could be retrieved, and the overall pooled prevalence is low. However, the limited data and confounding factors mean that that these results may significantly underestimate the true situation. It is recommended that a complete survey of blood banks should be conducted throughout the country, in order to provide a more reliable estimate of the risk of transfusion-associated HIV infection in Pakistan.