Suicide Risk Prediction Models Prove Cost-Effective in Healthcare

Statistical suicide risk prediction models could be implemented cost-effectively in healthcare organizations and may help save many lives each year, according to a study published in JAMA Psychiatry.

Suicide is the tenth leading cause of death in the US, researchers stated. While there are several effective interventions to reduce the risk of suicide, challenges in identifying people at risk of suicide and concerns about the potentially high costs of suicide-prevention strategies have hindered their widespread use.

Researchers from Massachusetts General Hospital (MGH) developed a mathematical model projecting suicide-related health economic outcomes over a lifetime for US adults treated by primary care physicians.

The model evaluated the practicality of predicting individuals’ risk of suicide and then offering either of two possible interventions for people at high risk. The first was active contact and follow-up, which consists of the patient at risk receiving an initial intensive evaluation and being contacted frequently thereafter by phone or mail.

The second intervention was cognitive behavioral therapy, a form of psychotherapy in which the therapist helps the patient identify and change self-destructive or disturbing thought patterns.

Using standard health economic measures, researchers found that both interventions could be cost-effective as long as the models used to predict suicide risk have a high degree of accuracy. When they examined risk prediction models developed by previous researchers, the team found that several of these models would be accurate enough to be practical and cost-effective.

The findings show that existing suicide risk prediction models are sufficiently accurate at identifying at-risk individuals, and could benefit the healthcare system.

"There are complex statistical models that researchers have come up with to predict who is at the highest risk of suicide or suicide attempts, and our analysis suggests that those models are now accurate enough that we could be implementing them in the real world," said Eric L. Ross, MD, a resident in the Department of Psychiatry at MGH.

"And if we do implement them, our analysis suggests they would be cost effective. That doesn't suggest that it would save the healthcare system money, but it does mean it would be a good investment of resources in order to improve people's quality of life and improve people's longevity.”

Implementing suicide risk prediction models in healthcare organizations could help providers detect high-risk patients.

“These findings suggest that with sufficient accuracy, statistical suicide risk prediction models can provide good health economic value in the US. Several existing suicide risk prediction models exceed the accuracy thresholds identified in this analysis and thus may warrant pilot implementation in US healthcare systems,” researchers said.

Suicide risk prediction models offer a new method of identifying and mitigating suicide risk in patient populations, researchers concluded.

"Suicide rates have increased substantially over the past 20 years, so it is clear that we need new tools for addressing this national problem,” said senior author Jordan W. Smoller, MD, ScD, of the Department of Psychiatry at MGH.

“Most individuals who die by suicide are seen by healthcare providers in the months before their death, so healthcare settings have a crucial opportunity to prevent this tragedy. Our results suggest that the tools exist to enable cost-effective interventions. And that, I think, is cause for hope."

Improving suicide risk prediction has emerged as a significant concern among healthcare researchers. Recently, a team from Vanderbilt University Medical Center (VUMC) developed a machine learning tool that can analyze EHR data and calculate suicide attempt risk. The algorithm could help providers know which patients to screen in nonpsychiatric clinical settings.

“Today across the Medical Center, we cannot screen every patient for suicide risk in every encounter—nor should we,” said Colin Walsh, MD, MA, assistant professor of Biomedical Informatics, Medicine and Psychiatry.

“But we know some individuals are never screened despite factors that might put them at higher risk. This risk model is a first pass at that screening and might suggest which patients to screen further in settings where suicidality is not often discussed.”