Abstract:

Objective To explore the influencing factors of peristomal moisture-associated skin damage（PMASD） in patients with colorectal cancer and predict the probability of PMASD occurrence using an artificial intelligence health assistant consulting software based on these factors. Methods Clinical data from 300 patients with colostomy were analyzed to identify PMASD influencing factors. Data of 200 patients were randomly selected as the training set for pre-training the artificial intelligence health assistant consulting software，while the remaining 100 cases served as the testing set. The software predicted the probability of PMASD occurrence for the 100 test cases，and the accuracy of the results was evaluated through manual verification. Results The incidence of PMASD was 56.3%（169/300）. Logistic regression analysis revealed that comorbidities，stoma height，peristomal skin folds，stool consistency，enterostomy type，application of leak-proof products，and frequency of stoma bag replacement were significant influencing factors for PMASD. The artificial intelligence health assistant consulting software achieved a prediction accuracy rate of 92%. Conclusion Through identifying high-risk factors for PMASD and integrating predictions from the artificial intelligence health assistant consulting software，healthcare professionals can effectively anticipate PMASD incidence，therefore enabling the formulation of targeted interventions to reduce its occurrence.

Key words: Peristomal moisture-associated skin damage, Artificial intelligence health assistant consulting software, Colorectal cancer