Related Experiment Videos
Risk stratification and determinant identification of high-need, high-cost ICU patients using machine learning: a
Yufei Hou1, Qichao Shi1, Cheng Wang1
1Department of Health Economics, General Hospital of Northern Theater Command, Shenyang, China.
Insights
Machine learning accurately identifies high-need, high-cost (HNHC) intensive care unit (ICU) patients. This approach aids in understanding cost drivers and optimizing critical care resource allocation.
Area of Science:
- Critical Care Medicine
- Health Economics
- Data Science in Healthcare
Background:
- Intensive care units (ICUs) represent a significant portion of hospital expenditures.
- Identifying high-need, high-cost (HNHC) patients is crucial for cost management and resource allocation.
- Machine learning (ML) offers a potential solution for retrospective identification of HNHC ICU patients.
Purpose of the Study:
- To develop and validate machine learning models for identifying high-need, high-cost (HNHC) patients in the intensive care unit (ICU).
- To analyze the key clinical and resource-use variables associated with high costs in the ICU.
- To assess the stability and utility of ML models for cost auditing and resource allocation.
Main Methods:
- Retrospective study of adult ICU admissions (≥24 hours) from a Chinese tertiary hospital (2018-2024).
- Definition of HNHC patients as the top 5% of annual ICU costs.
- Development and evaluation of six ML models (including Random Forest) using feature selection, class balancing, cross-validation, and Bayesian optimization, assessed by AUC and SHAP values.
Main Results:
- Out of 51,056 ICU patients, 2,556 (5.0%) were classified as HNHC, exhibiting longer stays and higher intensive therapy use.
- The Random Forest (RF) model demonstrated superior discriminative performance (AUC=0.942) in identifying HNHC patients.
- ICU length of stay and mechanical ventilation duration were identified as primary cost drivers by SHAP analysis, with stable performance across DRG reform periods.
Conclusions:
- High ICU costs are predominantly driven by intensive resource utilization, not solely demographics.
- A validated RF model effectively identifies HNHC patients and cost drivers, aiding retrospective risk stratification.
- The developed ML tool supports more efficient allocation of critical care resources and structured cost auditing.
Background:
Intensive care units (ICU) account for a disproportionate share of hospital costs. Retrospectively identifying high-need, high-cost (HNHC) ICU patients using machine learning (ML) may inform structured cost auditing and more efficient healthcare resource allocation.
Methods:
This retrospective study included adult patients with ICU admission (≥24 h) from multiple specialty ICUs in a Chinese tertiary hospital (2018-2024). HNHC patients were defined as the top 5% of annual ICU costs. Clinical, laboratory, and resource-use variables were extracted and preprocessed. Six ML models were developed using feature selection, class balancing, cross-validation, and Bayesian optimization, with performance evaluated by area under the receiver operating characteristic curve (AUC) and related metrics. Shapley Additive Explanations (SHAP) was applied for model interpretability.
Results:
Among 51,056 ICU patients, 2,556 (5.0%) were classified as HNHC. HNHC patients had longer ICU stays, greater disease complexity, and substantially higher use and duration of intensive therapies, resulting in markedly increased total and ICU-related costs. After correlation and Boruta feature selection, 19 variables were retained for model development. Among six ML models, the random forest (RF) achieved the highest discriminative performance in the independent test set, with an AUC of 0.942 (95% CI 0.931-0.952), followed by Extra Trees and LightGBM. The random forest model showed a favorable balance between sensitivity and F1 score in this highly imbalanced population. Decision curve analysis demonstrated stable net benefit across clinically relevant threshold probabilities. SHAP interpretation identified ICU length of stay and mechanical ventilation duration as the strongest contributors, revealing pronounced nonlinear effects, while diagnosis-related group (DRG) reform did not substantially alter the contribution patterns of key features. Stratified analyses confirmed that model performance and feature contribution patterns remained stable across the pre- and post-DRG reform periods.
Conclusion:
High ICU costs were primarily associated with intensive resource use rather than demographics alone. A RF model reliably classified HNHC patients and remained stable across DRG reform, supporting its use as a tool for retrospective risk stratification, identification of cost drivers, and more efficient allocation of critical care resources.