Automated Identification of Patients with Advanced Illness

Abstract

Early identification of advanced illness patients within an inpatient population is essential in order to establish the patient's goals of care. Having goals of care conversations enables hospital patients to dictate a plan for care in concordance with their values and wishes. These conversations allow a patient to maintain some control, rather than be subjected to a default care process that may not be desired and may not provide benefit. In this study the performance of two approaches which identify advanced illness patients within an inpatient population were evaluated: LACE (a rule-based approach that uses L - Length of stay, A- Acuity of Admission, C- Co-morbidities, E- Emergency room visits), and a novel approach: Hospital Impairment Score (HIS). The Hospital impairment score is derived by leveraging both rule-based insights and a novel machine learning algorithm. It was identified that HIS significantly outperformed the LACE score, the current model being used in production at Northwell Health. Furthermore, we describe how the HIS model was piloted at a single hospital, was launched into production, and is being successfully used by clinicians at that hospital.

Figure 1.

Diagram shows the architecture of the HIS model. Machine Learning Part A: An unsupervised machine learning model was used to identify a cluster with the highest prevalence of frailty associated ICD-10 categories. Machine Learning Part B: Then a supervised random forest model was used to determine how well the IDC-10 codes that are over-represented in the frailty cluster, discriminate between a patient being in the frailty cluster vs. other clusters. Then based on random forest feature rank, weights are calculated to compute the frailty risk score of all patients. Finally, frailty risk score is associated with outcomes.

Figure 2.

Results of Machine Learning Part B. Supervised Learning: Feature Rank Plot