目的：分析与胆囊胆固醇息肉（gallbladder cholesterol polyp, GCP）和胆囊腺瘤（gallbladder adenoma, GA）相关的因素,建立预测模型,术前区分两者。方法：研究2015年1月至2019年12月间,在首都医科大学宣武医院和北京大学第一医院两个中心,超声检查胆囊息肉（gallbladder polyp, GBP）直径≥10 mm、行胆囊切除术的病人。依据病理诊断结果分为GCP组和GA组。通过单因素和多因素分析研究GCP与GA相关因素,用Logistic回归方程,建立预测模型,评估诊断效果。结果：共384例GBP病人,男女比例1∶1.49,年龄（50.32±14.15）岁。GCP组250例（65.1%）,GA组134例（34.9%）。单因素分析显示,计量资料中年龄、血糖、低密度脂蛋白、白蛋白和总胆汁酸在两组差异具有统计学意义（P<0.05）,计数资料中年龄≥50岁、低密度脂蛋白≥3.1 mmol/L、总胆汁酸≥10.0 μmol/L差异具有统计学意义（P<0.05）。多因素分析可得出年龄、低密度脂蛋白以及总胆汁酸具有统计学意义。建立预测回归方程,以P≥0.601诊断GCP,灵敏度74.8%。回归方程经过ROC曲线拟合,曲线下面积为0.662。结论：用年龄、低密度脂蛋白和总胆汁酸因素建立的方程,可较好地术前预测GBP病人GCP与GA。

Objective To analyze the factors related to gallbladder cholesterol polyp (GCP) and gallbladder adenoma (GA) and establish predictive model for distinction between GCP and GA before surgery. Methods Patients with gallbladder polyps diameter≥10 mm diagnosed by ultrasonography in Xuanwu Hospital Capital Medical University and the First Hospital of Peking University from January 2015 to December 2019 were studied. There were GCP group and GA group based pathological diagnosis. Univariate and multivariate analyses were used to study the factors related to GCP and GA. Logistic linear regression was used to establish a predictive model and to evaluate the diagnostic effect. Results A total of 384 patients with gallbladder polyp were in this study with male to female ratio of 1:1.49 and age of (50.32±14.15) years. There were 250 patients (65.1%) in GCP group and 134 patients (34.9%) in GA group. Univariate analysis showed that five factors in quantitative data including age, blood glucose, low density lipoprotein, albumin, and total bile acid and three factors in count data including age 50 years or older, low density lipoprotein ≥3.1 mmol/L and total bile acid ≥10.0 μmol/L had significant difference statistically between two groups (P<0.05). Three factors were gotten as age, low density lipoprotein and total bile acid with significant difference statistically in multivariate analysis. The sensitivity of predictive regression equation were 74.8%, when P≥0.601 was used to diagnose cholesterol polyps. The regression equation was fitted by ROC curve and the area under curve was 0.662. Conclusions The equation on the factors of age, low density lipoprotein and total bile acid could be used to predict cholesterol polyps and adenomas preoperatively for patients with gallbladder polyps.