[Relationship between nosocomial infection and hospital mortality. Multicenter study]

Hospital mortality related to infections acquired in the hospital setting has not been well studied in Spain. We carried out a study of seven hospitals in order to assess and quantify the problem.

The study period included three months of observations (between November 1, 1989 and January 31, 1990), and data pertaining to all deaths of patients hospitalized for a minimum of 24 hours were collected. The number of people admitted within the study period was 16,025, and the number of deaths registered and included in our study was 488 (3%). The data were obtained from the patient's medical history one week after death as well as from the hospital physicians on the case. In order to quantify the interobserver variability derived from the classification criteria, the simple kappa index was calculated and averaged to form an ordinal scale.

216 (44.3%) of 488 deaths included in our study had no infection, 138 (28%) had an infection no-hospital-associated, and 134 (27%) had nosocomial infection (50-10%--"causally related to death", 59-12%--"contributing to death", and 25-5%--"not related to death"). The lower respiratory tract infections, bacteremias and surgical wound infections were the most related to cause of death. Staphylococcus aureus was the pathogen most frequently associated with the infections found at the time of death.

Those patients admitted with non-fatal diseases made up the greatest percentage (39.9%) of deaths from nosocomial infections. The infection was considered the direct cause of death in 18.8% of these cases, although the differences found had no statistical significance.