Multistage machine learning model for automated referral triage in pain medicine

TL;DR

A new multistage machine learning model improves referral triage in pain medicine, increasing accuracy and efficiency in patient care.

Contribution

A multistage machine learning framework that outperforms single-stage models in referral triage for pain medicine.

Findings

The multistage model improved true positive rate (TPR) by up to 34.8% compared to single-stage models.

The framework also increased area under the curve (AUC) by up to 23%.

The approach is adaptable and can improve clinical workflows beyond pain medicine.

Abstract

•The multistage machine learning method shows improved performance over traditional single-stage models, delivering higher accuracy in referral triage in pain medicine.•By enabling more accurate triage in pain medicine, the proposed method ultimately reduces unnecessary visits, enhances overall healthcare efficiency and improves patient care.•The proposed framework is adaptable beyond pain medicine and holds strong potential for improving clinical workflows across other medical departments.

The multistage machine learning method shows improved performance over traditional single-stage models, delivering higher accuracy in referral triage in pain medicine.

By enabling more accurate triage in pain medicine, the proposed method ultimately reduces unnecessary visits, enhances overall healthcare efficiency and improves patient care.

The proposed framework is adaptable beyond pain medicine and holds strong potential for improving clinical workflows across other medical departments.

Effective referral triage in pain medicine is essential to ensure that patients receive timely and appropriate care. This study presents a multistage machine learning framework to better identify patients who may benefit from one of five specialised pain procedures. Two years of clinic data were used for training and validation, consisting of 231 features for 3,552 patients. The proposed approach started with the baseline model selection, followed by stage iterations acquiring the elbow method for stage decision. The Easy Ensemble was selected among methods and applied for each stage. Results showed that the approach improves prediction accuracy in the true positive rate (TPR) and area under the curve (AUC). The final-stage models achieved the improvement over single-stage model by as much as 34.8% and 23% on TPR and AUC. This multistage framework can enhance triage accuracy and hold potential for broader application in other clinical settings.

Figures