Related Experiment Videos
Dynamic Pressure Ulcer Risk Predictions for Hospitalized Patients: Development and Validation of a Machine Learning
Steffen Greup1, Denise Spoon2, Annemarie de Vroed3
1Department of Data & Analytics, Erasmus MC, Erasmus University Medical Centre, Rotterdam, the Netherlands.
Aim:
To develop and validate a dynamic prediction model for daily pressure ulcer risk predictions together with an expert group of nurses.
Design:
Diagnostic study.
Methods:
All admissions to general wards in a single-center tertiary university hospital in the Netherlands were included. A retrospective dataset with candidate predictors was collected from the electronic health records and split into a training set (December 2021-November 2022, N = 19,931) and a validation set (January 2023-June 2023, N = 11,387). The pressure ulcer outcome was identified from both structured and free text registration. Predictor definition, selection, and modelling choices were discussed with an expert group of nurses. Separate models were developed for the first 72 h of admission and > 72 h. The final models and predictive performance on the validation set were compared to the Waterlow score.
Results:
The difference in performance across candidate models was small on the training set. Logistic regression with an L2 penalty and a spline transformation applied to a limited set of predictors was chosen as the final model, and the candidate predictor set was reduced to a final predictor set. Expected remaining length of stay, age and Activities of Daily Living score were the predictors with the strongest contribution. The AUROC for the final model was 0.790 (≤ 72 h) and 0.795 (> 72 h) on the validation dataset. The model clearly outperformed the Waterlow score (0.816 vs. 0.702 (≤ 72 h) and 0.800 vs. 0.677 (> 72 h)).
Conclusions:
We developed and validated a dynamic prediction model for daily pressure ulcer risk predictions. During the development process, special care was given to considerations for implementation and user acceptance. The model was named DRAAI (Decubitus Risk Alert based on AI), which means 'turn' in Dutch and is a common pressure ulcer preventive measure.
Implications For Nurses:
Facilitates targeted pressure ulcer prevention and reduces registration burden.
Reporting Method:
TRIPOD.
Patient Or Public Contribution:
No Patient or Public Contribution.