Surgical infections surveillance: results of a six-month incidence study in two Italian hospitals.

In a six-month incidence study of surgical wound infections (SWI) in two Italian hospitals, 1,019 surgical patients, in three general surgery wards, and 433 surgical patients in one orthopedics ward were studied. For the SWI surveillance, the DANOP-DATA system was used: this microcomputer program was developed by Danish authors and tested in a European multicenter study coordinated by the World Health Organization in 1989. Two Italian hospitals participated in the multicenter study. The overall infection rate was 1.2 per 100 operations in orthopedics and 4.9/100 in general surgery. The risk of infection increased with age (RR = 2.06; 95% CL = 1.20-3.53), wound class (RR = 3.38; 95% CL = 1.97-5.8), length of pre-operative stay (RR = 2.71; 95% CL = 1.54-4.74), and duration of operation (RR = 2.59; 95% CL = 1.48-4.54). The infection rates ranged from 3.7 to 7.3/100 among the three general surgery wards; this variability by ward was only partially explained by differences in the age distribution of in-patients, wound class, duration of operation and length of pre-operative stay. When all these risk factors were simultaneously taken into account using a logistic regression model, the odds radio, comparing one of the three general surgical wards with the other two, was still 2.29 (95% CL = 1.23-4.26). The observed variability can be attributed to differences, among the participating wards, in the case-mix of patients treated and/or to differences in the quality of infection control programs implemented.