Clinical Report: Evaluating Risk Prediction Models for PMASD in China
Overview
This systematic review evaluates 11 risk prediction models for Peristomal Moisture-Associated Skin Damage (PMASD) in China, revealing a significant risk of bias across all models. The strongest predictors identified include the history of radiotherapy, type of stoma, stoma opening height, and surgical wound in the plate area.
Background
Peristomal Moisture-Associated Skin Damage (PMASD) is a prevalent complication following enterostomy, significantly impacting patient quality of life and healthcare costs. The increasing incidence of colorectal cancer in China necessitates effective management strategies for PMASD to improve patient outcomes. Understanding and utilizing risk prediction models can aid healthcare providers in identifying at-risk patients and implementing timely interventions.
Data Highlights
Model
AUC
Incidence Rate
Model 1
0.812
22.8%
Model 2
0.914
59.1%
Key Findings
11 prediction models from 10 studies were included in the review.
Incidence rates of PMASD ranged from 22.8% to 59.1%.
AUC values for the models ranged from 0.812 to 0.914.
Key predictors included history of radiotherapy, type of stoma, stoma opening height, and surgical wound in the plate area.
All models exhibited a substantial risk of bias, limiting their clinical utility.
Future studies should focus on large-sample, multicenter, and high-quality prospective designs.
Clinical Implications
Healthcare providers should be aware of the limitations of current PMASD risk prediction models due to inherent biases. Implementing structured risk assessment protocols can enhance early identification and management of patients at risk for PMASD, ultimately improving care outcomes.
Conclusion
While existing risk prediction models for PMASD show promise, their clinical application is hindered by significant biases. Continued research is essential to refine these models for better predictive accuracy and utility in clinical practice.
Analyses by treatment timing, cumulative dose, and stunting supported the overall finding, though early-initiation and female subgroup data were limited.
