WANG Xiuhua, ZHI Hongmin, ZHANG Yujuan
Objective The incidence rate of surgical site infection of surgical patients in a hospital was calculated, the related factors of postoperative infection were found out, and the probability of postoperative infection of surgical patients was predicted scientifically. Methods Research object selected a top three teaching hospitals in 2012—2015 postoperative patients, 275 cases patients with surgical site infection cases, the other in accordance with the 1∶1 ratio to choose 266 patients as control group, postoperative infection cases retrospective survey study, statistical detection pathogenic bacteria and surgical site infection situation, to probe into comprehensive factors of surgical site infection combined with statistical analysis data to construct artificial network model of postoperative infection. Results The Results showed that surgical site infection rate was 0.35%. The Results of multi-factor analysis showed that the independent risk factors for surgical site infection were the type of operation, the presence of primary disease, the type of incision, the ASA grade, the patient's age, the duration of operation, and the number of cases (OR=11.043, 9.587, 2.136, 1.818, 1.299, 1.293, 1.041, P<0.05). Finally, the operation resultsof the network model are determined as follows: the average error is 0.040% and the network error rate is 0.038%. Area under the curve (ROC) is greater than 0.9. Conclusion The top three risk factors of surgical site infection were surgical type, primary disease and incision type. The prediction Results of the artificial network model for the risk of ssis were satisfactory, and the related prediction information platform system was established to provide guidance for the clinical hospital infection control decision.
