Surgical site infection of 7301 traumatologic inpatients (divided in two sub-cohorts, study and validation): modifiable determinants and potential benefit.

Abstract

The widespread use of peri-surgical chemoprophylaxis is decreasing the incidence of surgical site infection in Traumatology, which nevertheless remains as a problem amenable to further reduction. We have tried to identify modifiable determinants of the avoidable incidence of surgical site infection, and to estimate the potential benefit from implementing adequate measures targeting those determinants. In a firts sub-cohort of 5320 traumatologic patients with a post-surgical stay of more than 2 days, in Madrid's La Paz Hospital, between 1991 and 1996. The epidemiological surveillance was prospective during their hospital stay but also includes data on readmissions due to infection so as to analyse the actual incidence of surgical site infection, both before and after discharge. Bivariate and multivariate (multiple logistic regression model) analyses of risk factors for surgical site infections have been performed. In this sub-cohort, 212 patients (3.9%) suffered some type of nosocomial infection. The incidence of surgical site infection before discharge in cases of clean surgery was 1.6%, rising to 2% when cases readmitted for infection after discharge were considered. The multivariate analysis of risk factors for surgical site infection identified the following: contaminated ('dirty') surgery (OR: 10.5), inadequate chemoprophylaxis (OR: 1.5) and a pre-surgical stay of more than 4 days (OR: 1.6). Next, a second sub-cohort, consisting of 1981 patients, treated between 1997 and 1999, was analysed to validate the results of the first multivariate analysis. The validation model (and the global cohort with 7301 patients) corroborates the importance of the same three factors. Last, we calculated that controlling two modifiable factors, pre-surgical stay and peri-surgical chemoprophylaxis, could avoid 56% of surgical site infections.