Objective: to assess the risk of surgical wound infection and hospital acqu
ired infections among patients with and without adequate antibiotic prophyl
axis. Also, to provide models to predict the contributing factors of hospit
al infection and surgical wound infection. Design: survey study. Prospectiv
e cohort study over 14 months, with data collected by a nurse and a epidemi
ologist through visits to the surgical areas, a review of the medical recor
d and consultation with the medical doctor and nurses attending the patient
s. Setting: Two hundred and fifty bed, general hospital serving Puertollano
(Ciudad Real), population - 50,000. Results: between February 1998 and Apr
il 1999, 754 patients underwent surgery, 263 (34.88%) received appropriate
peri-operative prophylaxis while 491 (65.12%) received inadequate prophylax
is. For those who received adequate antibiotic prophylaxis, the percentage
of nosocomial infection was 10.65% compared with the group who received ina
dequate prophylaxis in which the percentage of nosocomial infection was 33.
40%. The relative risk of nosocomial infection was, therefore, 4.21 times h
igher in the latter group (confidence intervals 95%: 2.71-6.51). A patient
in the inadequate prophylaxis group had a 14.87% chance of wound infection
while a patient in the adequate prophylaxis group had a 4.56% chance of wou
nd infection. The relative risk of wound infection was 3.65 times higher in
the group that received inadequate prophylaxis (confidence intervals 95%:
1.95-6.86). The final regression logistic model to assess nosocomial infect
ion incorporated seven prognostic factors: age, length of venous periferic
route, vesicle catheter, duration of operation, obesity, metabolic or neopl
asm diseases and adequate or inadequate prophylaxis. When we incorporated t
hese variables in the multi-factorial analysis we found that the relative r
isk of developing nosocomial infection was 2.33 times higher in the group w
hich received inadequate prophylaxis. When we applied the second logistic m
ultiple regression model (wound infection), we discovered that the probabil
ity of developing surgical wound infection was 2.32 times higher in the gro
up which received inadequate prophylaxis as opposed to the group, which rec
eived adequate prophylaxis. The goodness of fit (Hosmer-Lemeshow test) show
ed a correct significance in all models. Conclusions: a multi-factorial ana
lysis was applied to identify the high-risk patients and the risk factors f
or postoperative infections. Through the application of these multiple regr
ession logistic models, we conclude that the correct antibiotic prophylaxis
is effective and will subsequently reduce postoperative infection rates, e
specially in high-risk patients. Therefore, the choice of antimicrobial age
nt should be made on the basis of the criteria of hospital committee.