Background: Infectious disease testing has improved but viral infection transmission through transfusion continues to occur. Residual risk, however, is now so low in some countries that it can only be estimated by mathematical modeling. With hierarchical Bayesian methods, this study estimates the residual risk of transfusion-transmitted infections in Santa Catarina, Brazil.
Study design and methods: Data from the six state blood collection services covering Santa Catarina between 1998 and 2002 were used. Information was obtained on donor profiles, screening and confirmatory test results, and records of all allogeneic blood donations for repeat donors. Residual risk estimates of hepatitis B virus (HBV), human immunodeficiency virus (HIV), hepatitis C virus (HCV), and human T-cell leukemia virus (HTLV) were separately derived from posterior distributions of incidence rates and preseroconversion window-period lengths.
Results: Estimated risks of a donation infectious for HBV and HIV entering the blood supply are 1:10,700 (95% confidence interval [CI], 1:4900-1:28,200) and 1:26,200 (95% CI, 1:14,800-1:64,100) donations, respectively. Estimated risks for HCV and HTLV were 1:19,300 (95% CI, 1:10,400-1:44,800) and 1:116,300 (95% CI, 1:40,200-1:1,000,000) donations, respectively. HBV risk is 1.8 times greater than HCV risk, 2.4 times greater than HIV risk, and 10.8 times that of HTLV. Actual risks would be lower due to immune recipients and subinfectious levels of undetected viremia.
Conclusion: The major factor contributing to risk differences between Santa Catarina and countries with similar testing regimes is the much higher source population diseases prevalence. Payoff for nucleic acid testing would be low, thus additional investment in safety should be based on studies of the cost-effectiveness of different strategies to reduce overall transmission.