Tianjin Medical Journal ›› 2022, Vol. 50 ›› Issue (5): 528-532.doi: 10.11958/20212060

Previous Articles     Next Articles

Colorectal cancer screening results and the establishment of the prediction model of advanced colorectal adenoma in Xihu district, Nanchang city#br# #br#

ZOU Xiahui, LIU Dunju, WAN Dehui, JIANG Huiying   

  1. Department of Gastroenterology, Nanchang Third Hospital, Nanchang 330063, China
  • Received:2021-09-06 Revised:2021-12-31 Published:2022-05-15 Online:2022-07-04

Abstract: Objective To investigate data of colorectal cancer screening in Xihu district, Nanchang city, and to construct a line graph prediction model to evaluate its predictive value for advanced colorectal adenoma. Methods A total of 10 618 patients aged 40 to 74 who underwent colon cancer screening were selected. All subjects were identified by two fecal immunoassay fecal occult blood tests (FOBT) and questionnaire. The high-risk group underwent colonoscopy, and the detection of colon diseases during the 5-year period was counted. According to the screening results, the clinical characteristics of non-advanced colorectal adenoma and advanced colorectal adenoma were compared, and the risk factors of advanced colorectal adenoma were analyzed by Logistic regression. Based on the screened risk factors, a line graph prediction model was constructed and its predictive value was evaluated. Results Among the 10 168 patients who participated in the screening, 2 078 patients were identified as the high-risk group by preliminary screening, and 1 573 patients were completed colonoscopy, with an examination rate of 75.7%. A total of 584 cases (37.1%) of colon diseases were detected, including 18 cases (1.1%) of colorectal cancer, 79 cases (5.0%) of advanced adenoma, 224 cases (14.2%) of non-advanced adenoma, 254 cases (16.2%) of non-adenomatous polyp, and 9 cases (0.6%) of inflammatory bowel disease. Univariate and multivariate Logistic regression analysis showed that male, old age, obesity, smoking, alcohol consumption, family history of colon cancer, appetite for red meat, and gross shape were the main risk factors for advanced colorectal adenoma. The area under receiver (AUC) of characteristic operating curve (ROC) predicted by the inclusion of the above indicators was 0.798 (95%CI: 0.737-0.859), and the internal validation C index was 0.797. The calibration curve showed that the incidence rate predicted by the linear prognostic model was in good agreement with the observed incidence rate (χ2=5.169,P=0.739). Conclusion The detection rates of colon cancer and advanced colorectal adenoma in 40-74 year-olds in Xihu district of Nanchang city are low, and the detection rate of advanced colorectal adenoma is 5.02%. The risk of developing advanced colorectal adenoma is significantly increased in patients with advanced colorectal adenoma who are aged, male, obese, drinking alcohol, smoking, complicated with diabetes, and a family history of colorectal cancer. Therefore, the linear graph model constructed according to the above factors has good predictive value.

Key words: colorectal neoplasms, adenoma, colonoscopy, nomograms, risk factors, advanced colorectal adenoma

CLC Number: