Assessing the effectiveness of 3 surveillance methods for nosocomial infections in psychiatry

Methodological data concerning nosocomial infection surveillance in psychiatric institutions are scarce. Methods used in Ether type of hospitals are probably not adjusted to psychiatry because risk groups are different. Our purpose was to assess retrospectively the effectiveness of 3 surveillance methods in our hospital. We studied the results of a prevalence study, the results of surveillance according to clinical microbiology data, and according to prescriptions of antimicrobial agents. The prevalence survey gave poor information because of a low number of infections (n = 5; prevalence rate = 0.9 %). The 2 other methods can not be used separately. Surveillance according to clinical microbiology data showed a lack of sensitivity. Indeed, during the 6 months of the study, 38 nosocomial infections were detected by clinical microbiology data, whereas our hospital physicians wrote out 88 antibiotic prescriptions. Moreover, antimicrobial agent prescription data showed a lack of sensitivity and a lack of specificity. Indeed, on one hand, among 32 urinary tract infections detected by laboratory work, only 14 led to a prescription. On the other hand, it was not easy to consider as nosocomial, 20 rhino-pharyngeal infections and 25 cases of bronchopneumonia that led to a prescription. These 3 methods gave different but probably complementary results. According to us and with great care accorded to the quality of data collection and to the pertinence of infection diagnosis, a prospective study of incidence associating these 2 methods could allow to detect most nosocomial infections in our hospital. Such a study would be long and time-consuming. Regular sensitization and motivation of all concerned hospital staff would be necessary (environmental microbiology analysis, regular information in various wards).