Analisi multilivello della sorveglianza endoscopica post-polipectomia per lo screening del tumore colorettale in Piemonte

Background . Colorectal cancer (CRC) is a highly prevalent and deadly cancer. Screening Programs play a crucial role in early detection and treatment of pre-malignant lesions. After polypectomy, patients may be considered at higher risk, depending on the type of lesion found, and may enter a surveillance program for closer monitoring. This study examines the situation in the Piedmont region, investigating the association of the patients’, exams and operators’ characteristics with the detection of advanced adenomas during the follow-up period. Methods. A cohort study, based on current data collected from the CRC screening information system of the Piedmont region, was set up. Multivariate logistic models are fitted to understand the association of the detection of advanced adenomas with patients’ and index exams characteristics. Since the dataset of analysis shows a multilevel nature, a further analysis is conducted with the estimation of random effects. Multilevel regression models examine the association of follow-up advanced adenomas with the level of experience of the operators performing the index exams, whose proxy is represented by the adenoma detection rate (ADR). A random intercept model allows the follow-up advanced adenoma to be specific for each group with different level of experience considered. A random slope model, adding ADR, is estimated to correct the model for the experience a group is characterized by, removing the random intercept for collinearity issues if all levels are included. Results. The higher risk categories for the detection of an advanced adenoma during follow-up are men and oldest people. Moreover, the quality of vision in index colonoscopy has a significant role: the better the quality, the lower the risk of follow-up advanced adenoma detection. The outcome at the index colonoscopy can influence the outcome at follow-up, as people with high-risk adenoma at index examination are at higher risk when compared to those who had low risk adenoma or negative outcomes at the index exam. The random intercept model suggests that variance between ADR groups is significant. Moreover, although not significant, the effect of the endoscopist’s experience on follow-up advanced adenoma in the random slope model suggests that an index colonoscopy performed by an experienced endoscopist has a protective effect: those patients that had undergone an index colonoscopy performed by an endoscopist with intermediate and high ADR are less likely to have an advanced adenoma at follow-up, when compared to low ADR. Conclusions. This study highlights that the endoscopists’ training should be made a priority. Beyond the risk factors linked to the personal characteristics of the patients, on whom there is a limited possibility of action, improving the experience of the operators who perform the screening tests it is possible to reduce the detection of advanced adenomas during follow-up.