Improving patient screening for housing instability with electronic health records

Using electronic health record (EHR) data to measure how frequently someone changes their residential address may be an effective tool to screen for patients who may be experiencing homelessness, according to a Northwestern Medicine study published in PLOS ONE.

Abel Kho, MD, professor of Medicine in the Division of General Internal Medicine, of Preventive Medicine in the Division of Biostatistics and Informatics, and director of the Center for Health Information Partnerships in the Institute for Public Health and Medicine (IPHAM), was senior author of the study.

Approximately 770,000 people in the U.S. experienced homelessness in 2024, an 18% increase from the previous year, according to a recent report from the National Low Income Housing Coalition.

Housing stability can impact health outcomes and can help decrease health care costs. However, there remains a need across health care institutions to better screen patients for housing instability and to more effectively integrate existing EHR data to identify and ultimately prevent homelessness.

In the current study, investigators analyzed EHR patient data collected from 2018 to 2024 at eleven health care sites that was compiled in the Northwestern Medicine Enterprise Data Warehouse.

Data from approximately one million patients aged 18 years or older with at least three hospital encounters over three different years were included in the analysis. A patient’s “address churn” was determined as the number of address changes recorded in the EHR.

Using different statistical tests, the investigators compared demographic and clinical characteristics of patients with different address churn, evaluated the distribution of ages of patients for different levels of address churn, and evaluated the association between address churn and clinical diagnoses.

Overall, the investigators identified 756,222 patients with no address changes, 156,911 patients with one address change, 137,491 patients with two address changes, 9,558 patients with three address changes, and 8,129 patients with four or more address changes.

They also found that patients with no record of homelessness had an average of 0.6 address changes, while patients with a record of homelessness had an average of 1.8 address changes. Patients with four or more address changes also demonstrated an increased prevalence of psychiatric diagnoses (65.2%) compared to patients with no address changes (27.7%).

Alternatively, patients with fewer address churns demonstrated increased prevalence of high cholesterol and high blood pressure, according to the authors.

“As a patient had a higher address churn over the set period of time, we saw that they actually had more of the psychiatric diagnostic codes in their EHR, such as for anxiety or depression,” said Janet Song, ’24 MD, a former medical student and internal medicine resident at Kaiser Permanente Santa Clara Medical Center, who was lead author of the study.

“I thought that had a lot of implications both in terms of the stressor that moving itself has on a person, as well as the implications of, as we see people move more and more frequently, maybe they’re heading towards a more stressful life circumstance such as homelessness, which is ultimately what we’re trying to get ahead of and screen for before it actually happens,” Song said.

The findings underscore the importance of measuring residential address changes when screening patients for housing instability and the association between housing instability and psychiatric disorders.

“Oftentimes, physicians don’t end up talking about housing instability or patients don’t feel comfortable talking about it. So, from a screening standpoint, having some kind of tool like this would be really great,” Song said. “Our findings can also help public health policymakers in targeting vulnerable populations at risk of homelessness with multiple health comorbidities for housing interventions.”

Going forward, the investigators hope to implement their tool with other health care systems in the region and nationwide, as well as with health care institutions in the Chicagoland area who are part of the CAPriCORN patient-centered outcomes and research network.

Using electronic health record (EHR) data to measure how frequently someone changes their residential address may be an effective tool to screen for patients who may be experiencing homelessness, according to a Northwestern Medicine study published in PLOS ONE.

Abel Kho, MD, professor of Medicine in the Division of General Internal Medicine, of Preventive Medicine in the Division of Biostatistics and Informatics, and director of the Center for Health Information Partnerships in the Institute for Public Health and Medicine (IPHAM), was senior author of the study.

Approximately 770,000 people in the U.S. experienced homelessness in 2024, an 18% increase from the previous year, according to a recent report from the National Low Income Housing Coalition.

Housing stability can impact health outcomes and can help decrease health care costs. However, there remains a need across health care institutions to better screen patients for housing instability and to more effectively integrate existing EHR data to identify and ultimately prevent homelessness.

In the current study, investigators analyzed EHR patient data collected from 2018 to 2024 at eleven health care sites that was compiled in the Northwestern Medicine Enterprise Data Warehouse.

Data from approximately one million patients aged 18 years or older with at least three hospital encounters over three different years were included in the analysis. A patient’s “address churn” was determined as the number of address changes recorded in the EHR.

Using different statistical tests, the investigators compared demographic and clinical characteristics of patients with different address churn, evaluated the distribution of ages of patients for different levels of address churn, and evaluated the association between address churn and clinical diagnoses.

Overall, the investigators identified 756,222 patients with no address changes, 156,911 patients with one address change, 137,491 patients with two address changes, 9,558 patients with three address changes, and 8,129 patients with four or more address changes.

They also found that patients with no record of homelessness had an average of 0.6 address changes, while patients with a record of homelessness had an average of 1.8 address changes. Patients with four or more address changes also demonstrated an increased prevalence of psychiatric diagnoses (65.2%) compared to patients with no address changes (27.7%).

Alternatively, patients with fewer address churns demonstrated increased prevalence of high cholesterol and high blood pressure, according to the authors.

“As a patient had a higher address churn over the set period of time, we saw that they actually had more of the psychiatric diagnostic codes in their EHR, such as for anxiety or depression,” said Janet Song, ’24 MD, a former medical student and internal medicine resident at Kaiser Permanente Santa Clara Medical Center, who was lead author of the study.

“I thought that had a lot of implications both in terms of the stressor that moving itself has on a person, as well as the implications of, as we see people move more and more frequently, maybe they’re heading towards a more stressful life circumstance such as homelessness, which is ultimately what we’re trying to get ahead of and screen for before it actually happens,” Song said.

The findings underscore the importance of measuring residential address changes when screening patients for housing instability and the association between housing instability and psychiatric disorders.

“Oftentimes, physicians don’t end up talking about housing instability or patients don’t feel comfortable talking about it. So, from a screening standpoint, having some kind of tool like this would be really great,” Song said. “Our findings can also help public health policymakers in targeting vulnerable populations at risk of homelessness with multiple health comorbidities for housing interventions.”

Going forward, the investigators hope to implement their tool with other health care systems in the region and nationwide, as well as with health care institutions in the Chicagoland area who are part of the CAPriCORN patient-centered outcomes and research network.