Transfusion risk associated with recent arbovirus outbreaks in French Polynesia.

Background and objectives: French Polynesia, where dengue virus (DENV) has been present for a long time, experienced two successive outbreaks of Zika (ZIKV) and chikungunya viruses (CHIKV) between 2013 and 2015. To avoid the transmission of these viruses by transfusion, nucleic acid testing (NAT) has been in place for DENV since 2013 and for ZIKV and CHIKV during epidemics. The objective was to compare the estimated risk of viraemic blood donation with NAT results and to discuss the impact on the prevention of transfusion‐related infectious risk. Materials and methods: The average risks of viraemic blood donation were estimated per year for DENV, and during the epidemic periods for ZIKV and CHIKV, using the Biggerstaff and Petersen model based on the incidence rate, the mean length of viraemia and the frequency of asymptomatic infection. The estimated risks were compared with the number of viraemic blood donations detected by NAT. Results: According to the different assumptions, risks estimates ranged from 11·2 to 53·1/100 000 donations for DENV, 746 to 1924/100 000 for ZIKV and 1083 /100 000 for CHIKV. When compared to the number of donations collected during the study periods, these estimates match NAT results (5 blood donors reactive for DENV, 42 for ZIKV and 34 for CHIKV). Conclusion: The risks of viraemic blood donation were related to the viral incidence in the general population and concordant with NAT results. These findings suggest that the screening may be optimized by a targeted NAT implementation based on incidence data. [ABSTRACT FROM AUTHOR]