Epidemiology of infectious encephalitis, differences between a prospective study and hospital discharge data

The French epidemiology of infectious encephalitis has been described in a 2007 prospective study. We compared these results with available data (demographic features, causative agents, case-fatality ratio) obtained through the French national hospital discharge 2007 database (PMSI), in order to evaluate it as a surveillance tool for encephalitis. Causative agents were identified in 52% of cases in the study, and 38% in PMSI (P<0.001). The incidence of encephalitis in France in 2007 was estimated as 2.6 cases/100 000 inhabitants. HSV and VZV were the most frequent aetiological agents in both databases with similar rates. Listeria monocytogenes and Mycobacterium tuberculosis were less frequent in PMSI than in the study (Listeria: 2% vs. 5%, P=0.001; Mycobacterium: 2% vs. 8%, P<0.001). The case-fatality ratios were similar, except for Listeria (46% in the study vs. 16%). Nevertheless, despite the absence of case definitions and a possible misclassification weakening PMSI data, we suggest that PMSI may be used as a basic surveillance tool at a limited cost.