Chronic Nonhealing Wound Infection Prediction Model with Six Variables

Yang Jiang developed a random forest model to predict infection risk in chronic nonhealing wounds using routine electronic medical record data. The model was trained on a primary cohort of 500 patients from two tertiary hospitals in China and validated on an external cohort of 300 patients. It was last updated on May 19, 2026.

Use Cases

• Predicting infection risk in chronic wounds based on six clinical predictors identified in the study.
• Benchmarking machine learning algorithms for medical risk prediction using electronic medical record data.
• Evaluating model interpretability and clinical utility via SHAP analysis and decision curve analysis.
• Developing a deployable clinical decision support tool, as referenced by the web calculator.
• Studying the predictive importance of biomarkers like C-reactive protein and procalcitonin in wound outcomes.

Strengths

• Model was developed and validated on cohorts totaling 800 patients (500 primary, 300 external).
• Performance metrics are reported, including AUROC of 0.884 in testing and 0.855 in external validation.
• The model uses only six predictors, making it pragmatic and interpretable for clinical use.
• The study includes external validation, which suggests tested generalizability beyond the training sites.

Limitations

• The dataset size is 21.3 KB, indicating a very limited scope, likely containing only model documentation or summary statistics.
• Row count and column-level documentation are unknown, which limits suitability assessment for direct data analysis.
• Data may reflect geographic bias inherent to the two tertiary hospitals in China.

Provenance

Source

Yang Jiang via figshare

Collection Method

Two-center retrospective study using electronic medical record data.

Freshness

Last updated 2026-05-19 05:44:10

Geography

China