| Objective: Surgical site infection (SSI) after open surgery for colorectal lesions is a common and mostly inevitable complication due to the existence of a large number of bacteria in the intestinal lumen. It may lead to incision infection, prolonged hospitalization, and increased cost. Moreover, it may progress to general infection and cause the patients'death. Recently, with the growing drug resistance of bacteria, the treatment of SSI has become more difficult. The objective of this study was to describe the predictors of SSI after surgery for colorectal diseases so as to lower the morbidity and mortality of SSI to a great extent.Methods: Four hundred and ninety six patients who underwent colorectal operations from Jan 1st, 2008 to Dec 31st, 2008 were enrolled from the 2nd affiliated hospital of Dalian Medical University. SSI was diagnosed according to the standard guideline and the SSI patients were selected by case files and follow-up. The following predictor factors were analyzed: sex, age, smoking history, diabetes mellitus (DM) and other chronic diseases, intestinal obstruction, benign and malignant diseases, preoperative hospital day (HOD), operation site, blood transfusion, preoperative and intraoperative usage of antibiotics, postoperative duration of antibiotics. The patients were divided into two groups by random digits table with a ratio of 2:1. Of all patients, 330 were selected as model fitting samples, while the other 166 as qualification samples. Using SPSS13.0 statistical software, multivariate logistic regression analysis was performed to identify predictors of SSI. Specifically, the predictor factors were categorized in view of patient and operation-related risk factors and then fitted in a Logistic regression model to analyze the positive factors by multivariate logistic regression. Finally, qualification samples were introduced in the final regression model to predict SSI and identify fitting model. Results: Of 496 patients who underwent colorectal operation, 35 (7.06%) were diagnosed with SSI, male12 and female 23, aged from 49-85 years old (mean 67.1±9.61 ) . The mean preoperative HOD was 8.83±3.75(X±S). The postoperative duration of antibiotics was 6.45±2.33(X±S). 28 patients had superficial and 4 had deep tissue infection in incision site and 3 had organic infection within 4-25 days after operation. Based on the criteria that the independent variable was selected as patients'condition and the dependent variable was selected as SSI, a fitting model was created and the predicting factors of SSI included sex and DM. When the independent variable was treatment condition and the dependent variable was SSI, the predicting factors of SSI included pre- and intra-operative administration of antibiotics and preoperative hospital stay. Then the statistically significant factors were analyzed using a Backward-LR logistic regression method. Multivariate predictors of SSI showed the risk factors are as followed:DM (OR=2.137, 95% confidence interval [CI] 0.71-6.426, P<0.1), pre- and intra-operative administration of antibiotics (OR=6.225,95% CI 2.016-19.215, P<0.1), sex (OR=3.154,95% CI 1.259-7.901 , P<0.1), and preoperative hospital stay (OR=1.111,95% CI 0.973-1.269,P<0.1). Qualification samples introduced in the model identified that only 27 patients had unexpected predicting results which showed 83.6% of accuracy.Conclusions: Sex of the patients, DM, pre- and intra-operative administration of antibiotics and preoperative hospital stay are potentially modifiable risk factors for predicting SSI in colorectal operation. |
PDF Full Text Request