Causes of Early and Late Mortality Following Invasive Pneumococcal Disease in Hull and East Yorkshire, 2007-2009
Although considered an acute illness, evidence suggests increased long term mortality following invasive pneumococcal disease (IPD). There is a lack of evidence regarding the underlying causes, but work in pneumonia suggests an increase in cardiovascular deaths. We conducted a retrospective cohort study to examine the causes of death in the two years following IPD.
All adult cases of IPD presenting to Hull and East Yorkshire Hospitals (1400 bed teaching hospital), 2007 to 2009 were identified via the Dept. of Microbiology's electronic-database. Medical records were reviewed and death certificates obtained for patients who died within 2 years of an episode of IPD. Causes of death within 30 days (early) were compared to those between 30 days and 2 years (late). Late deaths were compared to causes of death in our local population (N = 5819) in 2010.
207 patients (50.7% male, mean age 64.9) were included. Most patients (80%) had pneumonia. Mortality at 30 days following IPD was 20.8% (n=43) and 38.6% (n=80) by 2 years. Within 30 days, infection was the primary cause of death in 60.5% and contributed to 83.72%. Infection remained an important cause of late death, being the primary cause in 24.3% and contributory in 45.9%. Of late deaths due to infection, 89% were due to respiratory infection. Malignancy was the primary cause of most late deaths (48.6%). A higher proportion of deaths between 30 days and 2 years were caused by respiratory disease (including respiratory infection) and malignancy compared to overall deaths locally (27.0% vs. 15.7% and 48.6% vs. 29.0%, respectively), but less were due to cardiovascular disease (10.8% vs. 32.6%).
The majority of early deaths following IPD are due to infection, which remained an important primary and contributory cause of death during follow-up. Respiratory diseases (including pneumonia) were responsible for a greater proportion of late deaths compared to the general population. This work does not support the hypothesis that IPD increases late cardiovascular deaths with the majority of non-infective deaths being due to malignancy. Whether late deaths can be prevented by intervention (e.g. conjugate vaccine) after the IPD episode is unclear, but should be explored.
V. Allgar, None
G. Barlow, None