Estimation of the quality of life effect of seasonal influenza infection in the UK with the internet-based Flusurvey cohort: an observational cohort study

Abstract

BackgroundDespite the public health effect of seasonal influenza in the UK, population-wide estimates of the quality-adjusted life-year (QALY) loss associated with influenza do not exist. Such estimates are needed to quantify the overall disease burden, compare it with other diseases, and assess the cost-effectiveness of control measures. Since 2009, the Flusurvey—an internet-based community surveillance system for influenza-like illness—has been operating in the UK. UK residents are recruited through print and broadcast media, social media, and word of mouth. There were about 6000 registered participants during the 2012–13 season, during which we used the Flusurvey cohort to measure QALY loss associated with illness. MethodsEach week, participants were asked to report any symptoms and their health status. Participants were asked to score their health status on a scale of 0–100 (0 representing the worst health imaginable and 100 the best); this method has provided estimates of QALY loss that are consistent with those obtained with the more traditional EQ-5D during the 2009 influenza pandemic. The analysis presented is based on participants who completed the health score at least once when symptomatic and at least once when free of symptoms, allowing the effect of illness on quality of life to be measured in individuals. Results were stratified by age group (0–17, 18–44, 45–64, and ≥65 years) and by severity of symptoms (acute respiratory infection [ARI] and influenza-like illness [ILI]; appendix). FindingsIllness lasted longer for cases of ILI (median 9 days, IQR 6–15) than for ARI (6 days, 4–10) and increased somewhat with age (appendix). The median baseline health score, reported when individuals were free from symptoms, was 90 (IQR 83–95). Individuals with ILI reported low health scores (median 60, IQR 43–75; 1837 episodes); a small reduction was reported by those with ARI (80, 70–85; 1297 episodes). The scores were similar in different age groups. The mean loss of quality-adjusted life was about 2·0 days (SD 2·5) for ILI cases (median 1·1 days, IQR 0·5–2·6) and 0·6 days (SD 0·9) for ARI cases (median 0·3 days, IQR 0·1–0·7). The quality-adjusted life-day (QALD) loss was thus over three times higher for ILI than for ARI. The mean estimated ILI QALD loss was close to that estimated for influenza during the 2009 influenza pandemic (2·0 vs 2·7 days). QALD loss increased with age for both ARI and ILI (appendix) and was significantly higher in individuals experiencing fever (p<0·0001). There was weak evidence for an increase of QALD loss in smokers (p=0·07) but no association with being in a risk group (p=0·15) or being vaccinated (p=0·73). InterpretationThis internet-based approach opens the way for standardised and refined measure of the public health effect of seasonal influenza and other respiratory infections in the UK and elsewhere. However, the study has limitations: it was a non-random, self-selecting sample, in which children were under-represented, and the self-reported symptoms were not validated by a medical practitioner. FundingThe Flusurvey is funded by Epiwork as part of the EU Seventh Framework Programme. AC and SF are both funded by a Medical Research Council Career Development Award in Biostatistics. KE is funded by a Career Development Fellowship supported by the National Institute for Health Research. The views expressed in this publication are those of the authors and not necessarily those of the funders.