Related Experiment Video
Updated: Aug 11, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
The Deterioration Risk Index: Developing and Piloting a Machine Learning Algorithm to Reduce Pediatric Inpatient
Laura O H Rust1,2,3,4,5,6,7, Tyler J Gorham5, Sven Bambach5
1Division of Clinical Informatics, Department of Pediatrics, The Ohio State University College of Medicine, Nationwide Children's Hospital, Columbus, OH.
Insights
A new machine learning algorithm, the Deterioration Risk Index (DRI), significantly improves early detection of clinical deterioration in hospitalized children. This advanced system is more sensitive and generates fewer alarms than previous methods, enhancing patient safety.
Area of Science:
- Pediatric critical care medicine
- Machine learning in healthcare
- Clinical informatics
Background:
- Timely identification of clinical deterioration in pediatric inpatients is crucial for improving outcomes.
- Existing situational awareness programs may lack the sensitivity and specificity needed for optimal patient monitoring.
- Machine learning offers potential for developing more accurate predictive algorithms.
Purpose of the Study:
- To develop and deploy a disease cohort-based machine learning algorithm for early identification of hospitalized pediatric patients at risk for clinical deterioration.
- To create an algorithm that outperforms the existing situational awareness program in sensitivity and alarm frequency.
- To validate the algorithm's performance and assess its impact on clinical deterioration events.
Main Methods:
- A retrospective cohort study was conducted at a quaternary-care children's hospital.
- Lasso-regularized logistic regression was used to develop separate algorithms for cardiac, malignancy, and general patient cohorts.
- Predictors included vital signs, oxygen use, nursing assessments, early warning scores, diagnoses, and lab results.
- Model performance was compared to the existing situational awareness program using a novel retrospective validation approach.
- Simulations with frontline staff informed user experience and workflow refinement prior to pilot implementation.
Main Results:
- The Deterioration Risk Index (DRI) demonstrated 2.4 times greater sensitivity than the existing program (53% vs. 22%, p < 0.001).
- The DRI required 2.3 times fewer alarms per detected event (121 vs. 276).
- Significant sensitivity gains were observed in cardiac (73% vs. 18%) and malignancy (81% vs. 27%) cohorts.
- A pilot implementation showed a 77% reduction in deterioration events (3 observed vs. 13.1 expected, p = 0.001).
Conclusions:
- Machine learning algorithms, tailored to specific patient cohorts, can create more sensitive early warning systems for pediatric clinical deterioration.
- Acknowledging unique pathophysiology in cardiology and oncology patients is vital for accurate risk prediction.
- Successful bedside translation requires leveraging existing platforms and considering operational impacts of model implementation.
Objectives:
Develop and deploy a disease cohort-based machine learning algorithm for timely identification of hospitalized pediatric patients at risk for clinical deterioration that outperforms our existing situational awareness program.
Design:
Retrospective cohort study.
Setting:
Nationwide Children's Hospital, a freestanding, quaternary-care, academic children's hospital in Columbus, OH.
Patients:
All patients admitted to inpatient units participating in the preexisting situational awareness program from October 20, 2015, to December 31, 2019, excluding patients over 18 years old at admission and those with a neonatal ICU stay during their hospitalization.
Interventions:
We developed separate algorithms for cardiac, malignancy, and general cohorts via lasso-regularized logistic regression. Candidate model predictors included vital signs, supplemental oxygen, nursing assessments, early warning scores, diagnoses, lab results, and situational awareness criteria. Model performance was characterized in clinical terms and compared with our previous situational awareness program based on a novel retrospective validation approach. Simulations with frontline staff, prior to clinical implementation, informed user experience and refined interdisciplinary workflows. Model implementation was piloted on cardiology and hospital medicine units in early 2021.
Measurements And Main Results:
The Deterioration Risk Index (DRI) was 2.4 times as sensitive as our existing situational awareness program (sensitivities of 53% and 22%, respectively; p < 0.001) and required 2.3 times fewer alarms per detected event (121 DRI alarms per detected event vs 276 for existing program). Notable improvements were a four-fold sensitivity gain for the cardiac diagnostic cohort (73% vs 18%; p < 0.001) and a three-fold gain (81% vs 27%; p < 0.001) for the malignancy diagnostic cohort. Postimplementation pilot results over 18 months revealed a 77% reduction in deterioration events (three events observed vs 13.1 expected, p = 0.001).
Conclusions:
The etiology of pediatric inpatient deterioration requires acknowledgement of the unique pathophysiology among cardiology and oncology patients. Selection and weighting of diverse candidate risk factors via machine learning can produce a more sensitive early warning system for clinical deterioration. Leveraging preexisting situational awareness platforms and accounting for operational impacts of model implementation are key aspects to successful bedside translation.
