Repeated prevalence investigations on nosocomial infections for continuous surveillance.

In order to obtain an overview for the planning of further infection control activities, nine repeated prevalence studies were performed at monthly intervals. These occurred in the surgical units of eight medium-sized German hospitals. A total of 4984 surgical patients were investigated, the number of patients observed in each hospital varied from 365 to 913 patients, an average of 69.2 patients per prevalence study per hospital. A total of 212 nosocomial infections were found, the majority being surgical site (43.9%) and urinary tract infection (33.0%). The overall prevalence rate was 4.0%. More than four repeated investigations had only a minor influence on the 95% confidence intervals, and a follow-up of late microbiological reports increased the prevalence rate by only 7.5%. However, it was very useful to record the presence of urinary catheters on the prevalence day and also the preceding days; for instance, a device-associated prevalence of 7.8 urinary tract infections per 100 patients with urinary catheters was found on the day of investigation. In order to evaluate the situation in one's own surgical department by prevalence studies and for reasons of cost-effectiveness, the workload can be limited to four repeated studies in most hospitals. A further follow-up of later microbiological reports is not recommended, and it seems useful to concentrate on patients with indwelling devices.