Multiplevariate Logistic Analysis Of Adhesive Intestinal Obstruction

Acute intestinal obstruction is a common clinical acute abdomen, which is second only to the biliary system diseases and appendicitis. With the improvement of living standards and social security, the aging of the population and surgical indications relaxed, abdominal surgery is increasing year by year, and the risk of postoperative intestinal obstruction is increasing year by year. Adhesive intestinal obstruction(AIO) is formed after abdominal surgery, inflammation and trauma, and acute obstruction of the bowel is caused by adhesions. Intestinal contents can not operate normally and smoothly through the intestinal tract to the anus. Clinicians concern the diagnosis and treatment of AIO. The treatment includes non-surgical treatment and surgical treatment. For the patients invalidly with non-surgical treatment, in order not to occur the strangulating intestinal obstruction, bowel resection, systemic pathological physiological changes and even death, when to take surgical treatment is essential. Clinical studies to determine the timing of surgery in the domestic and international standards are not uniform. In this paper, the medical history, signs and various clinical indicators of AIO in our hospital were analyzed retrospectively. We want to seek an effective unified model to determine the timing of surgery.Objective:To analyse and research the main risk factors of the necessity of operative treatment in the AIO patients with operational histories,to establish the logistic regression model and to evaluate the accuracy,sensitivity and specificity for predicting the necessity of operative treatment for AIO.Methods:All the423patients selected in the study were the cases admitted with AIO in The Second Hospital of Tianjin Medical University from January2007to June2012. Case-control study was used to retrospectively investigate the clinical data of AIO cases. Firstly, single factor analysis was used to analyse a number of factors that may affect surgical indications. Then single logistic analysis was used in the statistically significant factors. At last to use multivariate logistic regression analysis, to get the logistic multivariate predictive model and to evaluate its accuracy, sensitivity and specificity. Results:11clinical variables,including wether the first adhesion, length of period of expectant treatment, abdominal pain continue or aggravation, severe abdominal distension, fever, heart rate, hit a swollen colon loops, peritonitis, decreased bowel sounds, ultrasound showed ascites and preoperative count of leukocyte were selected through single faetor analysis. By Logistic regression analysis, it was revealed that there were eight significant factors for the necessity of operative treatment as wether the first adhesion, length of period of expectant treatment,abdominal pain continue or aggravation,hit a swollen colon loops, peritonitis, decreased bowel sounds, ultrasound showed ascites and preoperative count of leukocyte. Logistic regression equation: P=ex/*1+ex),x=-5.357+1.233X3+0.584X4+2.159X6+1.197X10+1.547X11+0.993X12+1.361X13+1.300X14. The accuracy, specificity and sensitivity of the Logistic model were88.89%(376/423),95.51%(319/334),64.04%(57/89).Conelusion:Wether the first adhesion, length of period of expectant treatment, abdominal pain continue or aggravation, hit a swollen colon loops, peritonitis, decreased bowel sounds, ultrasound showed ascites and preoperative count of leukocyte were the main risk factors of the necessity of operative treatment in AIO. The Logistic regressive model based on these factors is reliable in predicting the Necessity of Operative Treatment.