BackgroundWe examined the added value of serologic testing for estimating influenza virus infection incidence based on illness surveillance with molecular testing versus periodic serologic testing. MethodsPregnant persons unvaccinated against influenza at <28 weeks gestation were enrolled before the 2017 and 2018 influenza seasons in Peru and Thailand. Blood specimens were collected at enrollment and ≤14 days postpartum for testing by hemagglutination inhibition assay for antibodies against influenza reference viruses. Seroconversion was defined as a ≥4-fold rise in antibody titers from enrollment to postpartum with the second specimen's titer of ≥40. Throughout pregnancy, participants responded to twice weekly surveillance contacts asking about influenza vaccination and influenza-like symptoms (ILS). A mid-turbinate swab was collected with each ILS episode for influenza real-time reverse transcription polymerase chain reaction (rRT-PCR). ResultsOf 1,466 participants without evidence of influenza vaccination during pregnancy, 296 (20.2%) had evidence of influenza virus infections. Fifteen (5.1%) were detected by rRT-PCR only, 250 (84.4%) by serologic testing only, and 31 (10.5%) by both methods. ConclusionsInfluenza virus infections during pregnancy occurred in 20% of cohort participants; >80% were not detected by a broad illness case definition coupled with rRT-PCR.
Read full abstract