Incidence-window period (IWP) and nucleic acid amplification testing (NAT) are the most widely used methods to estimate the incidence of transfusion-transmitted infections, including the hepatitis C virus (HCV) infection. The majority of HCV-infected people live in low and middle-income countries with only a serological screening of blood donors. A modification of the IWP method was proposed to include probabilistic case definition (PCD) of repeat blood donors whose prior donation was HIV-negative and occurred before the study start. First-time donor incidence was estimated by multiplying repeat donor incidence with repeat-to-first-time-donor prevalence ratio. Both methods were tested with artificial data with varying HCV prevalence (low, high), donor type (repeat, first-time) and donor follow-up (5 and 10 years). A real data example from the Santa Catarina state in southern Brazil was also provided. In the artificial data, the bias reduction of IWP-PCD compared to IWP incidence ranged between 95 and 100%. Real data analysis showed the HCV incidence per 100.000 person-years of 20.50 with 95% confidence interval (CI) of 7.52-44.62, and 6.84 (95% CI 0.82-24.70) for the IWP and IWP-PCD method, respectively, among repeat donors. Overall IWP incidence on the same scale by IWP was about threefold the IWP-PCD estimate of 32.5. Large bias reduction of the IWP incidence estimates due to IWP-PCD method with artificial data for both low and high prevalence scenario indicates a great potential of the latter method for blood donor screening in settings which rely only on serological testing. A relatively small number of donations, such as those found in regional blood banks, could be suitable for the IWP-PCD incidence and residual risk estimation. Recent changes in incidence for monitoring and surveillance could be detected more promptly than by widely used prevalence trends.
Buy this Article