Long-term survival after invasive pneumococcal disease - a matched cohort study using electronic health records in England

BackgroundInvasive pneumococcal disease (IPD) is associated with increased long-term mortality, but it is unclear if this can be explained by pre-existing comorbidities that predispose individuals to pneumococcal infection. MethodsWe conducted a matched cohort study comparing long-term survival (beyond 120 days after infection) in individuals with IPD and comparators without IPD. Cases were individuals in England aged [≥]65+ years with laboratory-confirmed IPD (2012-2019) identified through enhanced national surveillance. Comparators matched on age, sex, and calendar date were drawn from primary care electronic health records in Clinical Practice Research Datalink GOLD. We used Cox regression (stratified by matched set) to compare mortality in people with and without IPD, adjusting for relevant comorbidities, deprivation, and ethnicity. ResultsWe included 13,401 IPD cases and 67,005 comparators. After adjusting for comorbidities, deprivation, and ethnicity, we found increased all-cause mortality in IPD cases compared to comparators (hazard ratio 3.74, 95% CI: 3.50-3.99). The predicted median survival from adjusted models was 4.7 years (IQR: 2.9-7.4) for IPD cases and >11.9 (IQR : 8.7->11.9) for comparators. This increased mortality was consistent across subgroups defined by age, vaccination history, and comorbidity status (including diabetes, chronic respiratory disease, and chronic heart disease). ConclusionsIPD was associated with increased mortality at least 5 years after infection. These findings demonstrate the long-term consequences of IPD, emphasising the value of IPD prevention and the need for more research into clinical management of IPD survivors. Long-term mortality should be incorporated in cost-effectiveness analyses for pneumococcal vaccines. Article summaryCompared to people without infection, people who survived invasive pneumococcal disease in England had a 3.7-fold increase in mortality up to 10 years after infection. This increased mortality remained after accounting for age, sex, underlying comorbidities, deprivation, and ethnicity.