Excess mortality between 2007 and 2014 among patients with Clostridium difficile infection: a French health insurance database analysis

Background: The impact of Clostridium difficile infection (CDI) on mortality is controversial.

Aim: To assess excess mortality due to CDI in France.

Method: Two cohorts of patients with CDI and a cohort of matched controls were extracted from a 1% representative sample of subjects covered by the general health insurance system in France (Echantillon Généraliste de Bénéficiaires database, 660,000 patients). The CDI patients were hospitalized with CDI as a principal diagnosis or an associated diagnosis between 2007 and 2014, but not in 2006. Controls were patients hospitalized between 2007 and 2014 but not hospitalized with CDI between 2006 and 2014. The one-year incidence of deaths between 2007 and 2014 was estimated and compared with that of a propensity score (PS)-matched control group with no CDI (two controls per case). The PS was calculated with the following variables: age; sex; Charlson Comorbidity Index score; duration of stay; year of index stay; and main comorbidities. Cox and Poisson models were used to estimate the increased risk of death while adjusting for PS. Sensitivity analyses (timeframe, diarrhoea, recurrent hospitalization for CDI) were used to explore the robustness of the results.

Findings: In total, 482 patients who had been infected with C. difficile were matched with 964 controls. A significantly higher risk of death was observed among the subjects with CDI, with a non-adjusted hazard ratio of 1.65 [95% confidence interval (CI) 1.33-2.04] and an adjusted ratio of 1.58 (95% CI 1.27-1.97). The adjusted relative risk of death was 1.78 (95% CI 1.18-2.70]) at 28 days, 1.52 (95% CI 1.17-1.98) at three months, 1.52 (95% CI 1.20-1.93) at six months and 1.64 (95% CI 1.32-2.03) at 12 months. Sensitivity analyses produced similar results; the hazard ratio ranged from 1.53 to 1.86, and was always statistically significant.

Conclusion: CDI is responsible for excess mortality after taking age, sex, comorbidities and length of hospital stay into account.