Developing an operational definition of housing instability and homelessness in Veterans Health Administration’s medical records
Abstract: The main objective of this study was to examine how homelessness and housing instability is captured across data sources in the Veterans Health Administration (VHA). Data from 2021 were extracted from three data repositories, including the Corporate Data Warehouse (CDW), the Homeless Operations Management System (HOMES), and the Homeless Management Information System (HMIS). Using these three data sources, we identified the number of homeless and unstably housed veterans across a variety of indicators. The results showed that the use of diagnostic codes and clinic stop codes identified a large number of homeless and unstably housed veterans, but the use of HOMES and HMIS data identified additional homeless and unstably housed veterans to provide a complete count. A total of 290,431 unique veterans were identified as experiencing homelessness or housing instability in 2021 and there was regional variability in how homelessness and housing stability were captured across data sources, supporting the need for more uniform ways to operationalize these conditions. Together, these findings highlight the and encourage use of all available indicators and data sources to identify homelessness and housing instability in VHA. These methodologies applied to the largest healthcare system in the U.S. demonstrate their utility and possibilities for other healthcare systems. Transparent practices about data sources and indicators used to capture homelessness and housing instability should be shared to increase uniform use.
Abstract: Novel and automated means of opioid use and relapse risk detection are needed. Unstructured electronic medical record data, including written progress notes, can be mined for clinically relevant information, including the presence of substance use and relapse-critical markers of risk and recovery from opioid use disorder (OUD). In this study, we used natural language processing (NLP) to automate the extraction of opioid relapses, and the timing of these occurrences, from veteran patients' electronic medical record. We then demonstrated the utility of our NLP tool via analysis of pre-/post-COVID-19 opioid relapse trends among veterans with OUD. For this demonstration, we analyzed data from 107,606 veterans OUD enrolled in Veterans Health Administration, comparing a pandemic-exposed cohort (n = 53,803; January 2019-March 2021) to a matched prepandemic cohort (n = 53,803; October 2017-December 2019). The recall of our NLP tool was 75% and our precision was 94%, demonstrating moderate sensitivity and excellent specificity. Using the NLP tool, we found that the odds of opioid relapse postpandemic onset were proportionally higher compared to prepandemic trends, despite patients having fewer mental health encounters from which to derive instances of relapse postpandemic onset. In this research application of the tool, and as hypothesized, we found that opioid relapse risk was elevated postpandemic. The application of NLP Methods: to identify and monitor relapse risk holds promise for future surveillance, risk prevention, and clinical outcome research.