Preoperative prediction of the nature of gallbladder polypoid / 中华普通外科杂志

ABSTRACT

Objective To evaluate the risk factors of gallbladder neoplastic polyps and establish a Logistic regression equation.Methods We retrospectively collected clinicopathological data of 580 patients who underwent cholecystectomy for gallbladder polyps.Compared to the clinicopathological parameters between non-neoplastic polys and neoplastic polys,and adenomas and gallbladder cancer by univariate and multivariate analysis,we developed a Logistic regression equations to conduct diagnostic test.Results The univariate analyses showed that ultrasonic measuring gallbladder basal ＞ 0.72 cm,CEA ＞ 1.51 ng/ml,CA199 ＞ 12.92 U/ml for the diagnosis of gallbladder adenomatous polyps were statistically significant respectively (P ＜ 0.05).Multivariate regression analysis showed regression equations 1.When P ＞ 0.256,the sensitivity of the diagnosis of adenomatous polyps was 72.4％,the specificity was 80.0％,and the whole correct ratio was 77.8％.The positive likelihood ratio of adenomatous polyps was 10.44,AUC was 0.828.As for neoplasm risk factors,the CT value ＞ 32.50 Hu for the diagnosis of malignancy was statistically significant (P ＜ 0.05).Multivariate regression analysis displayed regression equations 2.When P ＞ 0.496,the sensitivity for the diagnosis of adenomatous polyps was 86.0％,the specificity was 90.5％,and the whole correct ratio was 89.0％.The positive likelihood ratio for malignant tumor was 9.08,AUC was 0.938.Conclusion Preoperatively,patients' age,symptoms,ultrasonic parameters,can be used to develop a regression equation to predict accurately the nature of gallbladder polyps.