Evaluation of HIV testing of the blood supply in Kenya

Objective: To identify ways to prevent transmission of HIV by blood transfusion in Kenya. Methods: From April through September 1994, we collected demographic and laboratory data on all blood donors in 5 government hospitals in western Kenya and I in Nairobi. Donations were tested for HIV-1 antJbody by routine hospital screening, and results were later compared with reference laboratory test results. The study team supplied HIV test kits to the hospitals if they were unavailable. Results: Of the 1883 Wood donations made in the 6 hospitals during the study period. 123 (6.5%) were HIV positive. Seroprevalence varied among hospitals, ranging from 2.0% to 20.6%. Seroprevalence was significantly higher among persons making a directed donation for a family member (9.7%) than among volunteer donors recruited by campaigns ( 1.7%, odds ratio = 6.3.95% confidence interval = 3.3. I13). Because of a national shortage of test kits during the study period, 29% of donations were screened by kits supplied by the study team. Overall, ï% of all blood transfusions were séropositive but not removed by hospital screening. Ninety-six percent of donations were screened by the hospitals, but one hospital screened only 56% of donations. Among units screened, hospitals accurately identified 81 % of séropositive donations. Conclusions: The Kenya Ministry of Health conducted this collaborative study and subsequently developed the following initiatives to reduce thehighriskofHIV transmission: ( I ) recruiting low risk donors, (2) improving procurement and distribution of test kits, (3) promoting universal screening of donors, and (4) retraining personnel in laboratory techniques and record keeping.