Data analysis
We calculated the incidence per 1000 person-years of
laboratory-confirmed influenza and RSV-associated ARI and SARI by
dividing the total episodes by total person-time under observation.
Weeks of intermittent loss to follow up (due to missed swab collection
and/or missed surveillance contacts) were excluded from the person-time
denominator when calculating the incidence rates. Multi-variable Poisson
regression was used to calculate incidence rate ratios (IRR) between
participants with and without any underlying cardiopulmonary conditions,
adjusting for age, sex, influenza vaccination status (participants
receiving influenza vaccine in at least one of the two study years were
considered as vaccinated), weekly influenza activity (influenza
infections detected by weekly NIH surveillance, categorized into
quartiles), and number of hospital visits during the past two years as a
proxy for healthcare seeking behavior. All data analyses were conducted
using STATA software version 14.2 (StataCorp LP, College Station, TX,
USA).