Influenza-associated hospitalizations and mortality in Hong Kong, 1998-2015

Abstract

BACKGROUND: Influenza viruses circulate each year, causing infections and disease in all age groups. A small fraction of infections are severe, requiring hospitalization, and some infections can be fatal. Apart from deaths caused by primary viral pneumonia, influenza can also lead to secondary bacterial infections, and can exacerbate underlying medical conditions such as cardiovascular disease. We used statistical models to estimate the burden of influenza-associated excess hospitalizations and deaths in Hong Kong. METHOD: We combined outpatient surveillance data on influenza-like illnesses (ILI) and laboratory detections of influenza into a proxy for influenza virus activity, denoted ILI+, by multiplying the rate of ILI consultations per 1000 consultations with the proportion of laboratory specimens testing positive for each type/subtype of influenza. We applied linear regression models to investigate the association between influenza activity as proxied by ILI+ and weekly hospitalization rates or mortality rates coded under various causes. Statistical models were fitted in a Bayesian framework. RESULTS: Using regression analysis, we estimated that influenza was associated with 7.60 (95% credibility interval, CrI: 6.35, 8.71) excess respiratory deaths per 100,000 persons per year, and 200 (95% CrI: 185, 214) excess respiratory hospitalizations per 100,000 persons per year. Rates of influenza-associated excess respiratory mortality were much greater in adults ≥65y than in all other age groups, while rates of influenza-associated respiratory hospitalizations had a U-shaped relation with age, being greatest in those <1y and ≥65y, and lowest in those 16-44y. When examining the contribution of different types and subtypes of influenza virus, we found that influenza A(H3N2) had the greatest impact, contributing around half of the excess respiratory mortality and respiratory hospitalizations on average, with influenza B contributing the second largest average. There was an average of 240 (95% CrI: 170, 370) excess respiratory hospitalizations for every excess respiratory death in persons 45-64y, compared to 16 (95% CrI: 13, 19) excess respiratory hospitalizations for every excess respiratory death in persons ≥65y. CONCLUSION: Influenza causes a substantial burden of hospitalizations and deaths each year, with A(H3N2) having the greatest impact. Infections in most groups are rarely fatal, but the risk of mortality appears to be highest for infections in persons ≥65y.