Personalized Medicine Algorithm Helps Veterans Administration Doctors Predict Risk of Homelessness in Veterans

Psychosocial care links vulnerable veterans to support services before they leave service for civilian life

Katherine Koh, MD, psychiatrist, at Mass General, leads research into predicting homelessness and psychosocial care
“We’ve long been limited in our ability to predict and prevent homelessness because most approaches have been focused on helping people after they’ve become homeless, rather than taking action before it ever occurs,” explained psychiatrist Katherine Koh, MD (above), in an MGH press release. “Our prediction model is highly actionable and we’re now designing an intervention that links the most vulnerable soldiers to support services before their active duty ends, then follows them over time,” she added. (Photo: Massachusetts General Hospital)

According to the American Medical Association (AMA), the United States faces a tremendous ethical problem in chronic homelessness. It correlates highly with health problems and is a huge consideration in the psychosocial care that healthcare providers can offer to certain populations.

Though precision medicine is typically thought of as using genomic data to tailor medical treatment, researchers from Harvard Medical School and Massachusetts General Hospital (MGH) have applied personalized medicine principles to predict and prevent homelessness, according to an MGH press release.

The approach is timely because of the prevalence of homelessness among America’s military veterans. To help recognize the risk of homelessness in veterans and allow for early interventions, the researchers used a machine learning algorithm to evaluate predictor variables.

The research team evaluated a database of nearly 17,000 soldiers for about 2,000 different predictors that could be associated with homelessness. Using artificial intelligence (AI), the researchers narrowed the thousands of potential variables down to three key predictors that strongly correlated with homelessness among U.S. Army soldiers who had returned to civilian life. These predictors are:

  • A self-reported lifetime history of depression.
  • The trauma of experiencing having a loved one murdered.
  • A diagnosis of post-traumatic stress disorder (PTSD).

The researchers highlighted how a small number of predictors were strongly correlated with the risk of homelessness for veterans, suggesting that a questionnaire could adequately cover these variables and help predict the risk of homelessness eventually occurring.

By recognizing which variables are most associated with homelessness, clinicians may be able to better tailor care to individuals, a key element of precision medicine. The researchers predicted that their findings would empower clinicians to be proactive, not reactive.

“For the first time we’re applying to homelessness a ‘personalized medicine approach’ that leverages differences in an individual’s biology, lifestyle, and environment to determine who is at greatest risk with a higher degree of precision than ever before,” said Ronald Kessler, PhD, in the MGH press release.

Kessler is the McNeil Family Professor of Health Care Policy at Harvard Medical School and senior author of the study.

The Harvard/MGH researchers published their findings in the American Journal of Preventive Medicine (AJPM), titled, “Predicting Homelessness Among U.S. Army Soldiers No Longer on Active Duty.”

Preventing Homelessness Among Other High-Risk Populations

While the Harvard/MGH research primarily focused on homelessness for veterans, the study’s lead author Katherine Koh, MD, believes that its results could be expanded to other populations as well. Koh is a psychiatrist with the Boston Health Care for the Homeless Program and at MGH.

“The presence of veterans among the homeless population in this country is still regarded by many as a matter of public shame, and for decades wasn’t given the attention it deserves,” she explained. “Our collaborative work is directly addressing that problem, and we believe utilizing prediction models such as the one we’ve developed could play a role in preventing homelessness not only among veterans, but also other high-risk populations.”

AI is not only enabling breakthroughs like facilitating clinicians’ ability to better predict homelessness, but it is also improving case management workflows and efficacy. One tool recently highlighted in GovCIO automates routine case management administrative tasks and facilitates communications between case workers and their patients.

The tool, called e-VA, is “an artificial intelligence-powered virtual assistant which allows you to schedule and reschedule appointments, text and email your Vocational Rehabilitation and Employment counselor, and submit documentation, all privately and securely from your smartphone, tablet, or computer,” according to the US Department of Veterans Affairs (VA).

The e-VA AI assistant has been adopted by the Veteran Readiness and Employment division under the Veterans Benefits Administration and is intended to help veteran counselors support veterans as they integrate into the civilian workforce after serving in the military. The e-VA tool uses AI to help veterans access services autonomously, reducing strain on VA case workers and healthcare providers.

“This is a big communications shift. It allows email communications back and forth that allows veterans to do several services that they previously couldn’t do unless they reached out to their counselor via the telephone,” Nick Pamperin, Deputy Director at the VA, told GovCIO.

Homelessness Affects America’s Healthcare Systems Overall

The AMA notes that those without shelter experience “cumulative health detriments from exposure, violence, poor nutrition, and limited access to hygiene opportunities and health services.” Additionally, homelessness has negative effects on the healthcare system as a whole and can create unfavorable situations for healthcare workers.

The AMA dedicated its entire November 2021 issue of AMA Journal of Ethics to homelessness and its effects on the medical system.

AI case management tools like e-VA are being proven in a veteran-centric environment, but like homelessness predictive machine learning research, they also can help provide psychosocial support for general populations. Advances in psychosocial care are allowing better recognition of risks for potential problems like homelessness as well as providing better means of support for those at risk.

Healthcare leaders can leverage the benefits of personalized AI technologies that provide individualized care for those at risk of psychosocial problems. While different from traditional thinking about personalized medicine, the ability to personalize psychosocial care is advancing in tandem with traditional precision medicine.

—Caleb Williams

Related Information:

Predicting Homelessness Among U.S. Army Soldiers No Longer on Active Duty

Homeless Health Care, Ending Homelessness are Matters of Justice

A Unique Machine-Learning Model Predicts Homelessness among U.S. Soldiers before Their Transition to Civilian Life

Predicting Homelessness Among U.S. Army Soldiers No Longer on Active Duty

This Virtual Assistant Uses AI to Improve Case Management at VA

Browse All Briefings