Incidence calculations
Incidence was calculated for all symptomatic hMPV infections in addition to stratifying by age, sex, and hMPV-associated ALRI from 2011–2016. Person-time was calculated as the amount of time starting at participant enrollment to December 30, 2016, or withdrawal from the study. Withdrawal for those who were lost to follow-up was calculated as the midpoint of the date of last contact and the date recorded by study personnel as lost to follow-up. Incidence was calculated using generalized linear models with Poisson distributions, participant age was calculated on a weekly basis.
To assess if there were significant differences in hMPV-associated ALRI reported by sex, we used Generalized Estimating Equations (GEE) assuming a dependence on the individual across the sampling weeks. Weeks when hMPV-associated ALRI was not reported was assumed to be negative for hMPV-associated ALRI.