The effectiveness of seasonal influenza vaccine in Hong Kong

Abstract

BACKGROUND: Influenza vaccination is the most effective intervention to reduce influenza infection and transmission . Vaccine effectiveness (VE) can vary due to factors such as matching between vaccine strains and prevailing strains, timing of vaccination, age and other characteristics of the vaccine recipients . In Hong Kong, there is limited information on influenza VE in the community . METHOD: We conducted a test-negative study between November 2013 and May 2015 . Patients aged at least 6 months of age presenting with at least two symptoms of acute upper respiratory tract infection within 72 hours of onset to local private outpatient clinics, university health clinics or private hospital outpatient clinics were recruited and tested for influenza virus by RT-PCR . Patients testing positive for influenza were defined as test-positive cases while those meeting with the same inclusion criteria but testing negative were controls . VE was estimated as one minus the odds ratio of vaccination among cases versus controls, adjusted for age and sex, and matching by fortnight of recruitment, using conditional logistic regression . RESULTS: Three influenza epidemics were included in this study during the year of 2013/14 and the winter of 2014/15 . Among 20 local private outpatient clinics, 1 university health clinic and 2 private hospital outpatient clinics, we recruited 1247 subjects of whom 365 (29 .3%) tested positive for influenza A and B virus by RT-PCR . The overall VE against laboratory-confirmed influenza A and B was estimated to be 42 .6% (95%CI: 8 .2%, 64 .1%) . Across two influenza epidemics of 2013/14, 126 of 613 (20 .6%) patients tested positive for influenza and the estimated VE was 47 .7% (95%CI: -10 .0%, 75 .1%) . During the first epidemic of 2014/15, on the other hand, 239 out of 634 (37 .7%) test-positive cases were enrolled and the estimated VE was 39 .7% (95%CI: -10 .6%, 67 .1%) . CONCLUSION: In summary, VE against laboratory-confirmed influenza was moderate overall . Lower estimated VE in 2014/15 is consistent with the mismatch between circulating and vaccine strains for the year . The relatively small sample size led to wide confidence intervals, yet results could still inform the general effectiveness of seasonal influenza vaccine for each epidemic and help estimate the overall protection provided by the influenza vaccination in the community .