Risk Assessment and Predicting Homelessness and Police Interaction in Calgary Through Administrative Health Care Data

Abstract

Background: Mental illness can lead to adverse outcomes such as homelessness and police interaction and understanding of the events leading up to these adverse outcomes is important.. Predictive machine learning (ML) models may not outperform logistic regression (LR). Method: An administrative healthcare dataset was used, comprising of 240,219 individuals in Calgary, Alberta, Canada who were diagnosed with addiction or mental health (AMH) between April 1, 2013, and March 31, 2018. The cohort was followed for 2 years to identify factors associated with homelessness and police interactions. We used a univariable and a multivariable LR model to identify predictive factors of homelessness and police integration by estimating odds ratios (ORs) with a 95% confidence interval. Then LR and ML models, including random forests (RF), and extreme gradient boosting (XGBoost) were compared. Results: After excluding prior outcomes before March 31, 2018, the cohort size decreased. Among 237,602 individuals, 0.8% (1,800) experienced first homelessness, while 0.32% (759) reported initial police interaction among 237,141 individuals. Male sex (AORs: H=1.51, P=2.52), substance disorder (AORs: H=3.70, P=2.83), psychiatrist visits (AORs: H=1.44, P=1.49), and drug abuse (AORs: H=2.67, P=1.83) were associated with initial homelessness (H) and police interaction (P). LR model with multinomial features, normalized data, and balanced classes showed superior performance (sensitivity =85%, AUC =84% for initial homelessness, and sensitivity =88%, AUC=81% for initial police interaction). Conclusion: This study identified key features associated with initial homelessness and police interaction and demonstrated the superior performance of the LR models using multinomial features, normalized data, and balanced classes.