[Analysis of influencing factors and construction of a prediction model for high-risk colorectal adenomas].

Objective: To analyze the influencing factors of high-risk colorectal adenomas and establish and validate a predictive model. Methods: The clinical data of 321 patients with colorectal adenomas who underwent colonoscopy and biopsy at Beijing Shijitan Hospital from January 2017 to January 2019 were retrospectively analyzed, including 201 males and 120 females, with the age (61.5±10.6) years. According to the colonoscopy pathology results, the patients were divided into the high-risk colorectal adenoma group and the non-high-risk colorectal adenoma group, and the differences in clinical indicators between the two groups were analyzed. The patients were randomly divided into a training set (n=225) and a validation set (n=96) at a ratio of 7∶3 using random sampling function. Multivariate logistic regression was used to screen the influencing factors of high-risk colorectal adenomas, and a nomogram of the predictive model was drawn. The predictive ability, accuracy and clinical applicability of the model were evaluated by the area under the receiver operating characteristic curve (AUC), calibration curve and decision curve analysis. Results: Among the 225 cases in the training set, there were 107 cases in the non-high-risk adenoma group and 118 cases in the high-risk adenoma group. In the training set, the proportion of patients with a history of coronary heart disease, the proportion of severe hepatic steatosis, body mass index, triglyceride levels, low-density lipoprotein cholesterol levels, and high-density lipoprotein cholesterol levels were higher in the high-risk colorectal adenoma group compared to the non-high-risk adenoma group. Conversely, hemoglobin levels and total bilirubin levels were lower in the high-risk group (all P<0.05). The results of the multivariate logistic regression model analysis revealed that a history of coronary heart disease (OR=9.172, 95%CI: 2.444-43.143), severe hepatic steatosis (OR=12.803, 95%CI: 4.423-36.987), and a high level of body mass index (OR=2.132, 95%CI: 1.659-2.874) were risk factors for high-risk rectal adenomas. Conversely, a high level of high-density lipoprotein cholesterol (OR=0.055, 95%CI: 0.005-0.518) and a high level of total bilirubin (OR=0.907, 95%CI: 0.833-0.982) were identified as protective factors against high-risk rectal adenomas. Construct a predictive model for high-risk colorectal adenomas based on the aforementioned influencing factors. The AUC of training set was 0.925 (95%CI: 0.893-0.958), with a sensitivity of 88% and a specificity of 82%, while the AUC of validation set was 0.886 (95%CI: 0.823-0.949), with a sensitivity of 72% and a specificity of 88%. The calibration curves of the training set and validation set show that the predictive model has good calibration (training set: P=0.130; validation set: P=0.910). The decision curves for both the training set and validation set demonstrate that the predictive model offers a clinical net benefit when the risk threshold exceeds 5%. Conclusions: A history of coronary heart disease, severe hepatic steatosis, a high level of body mass index, a high level of high-density lipoprotein cholesterol, and a high level of total bilirubin are influencing factors for high-risk colorectal adenomas. Incorporating these factors into a nomogram model allows for a more intuitive assessment of the risk of high-risk colorectal adenomas.