A New Mortality Prediction Tool Could Transform End-of-Life Care in California Prisons

Researchers at the University of California, San Francisco have developed a new mortality prediction tool designed to identify incarcerated individuals who are most likely to die within two years. The goal is not punishment or surveillance, but something far more humane: improving palliative care, encouraging advance care planning, and making compassionate early release a realistic option for people who are nearing the end of life while still behind bars.

This research focuses on a population that is often overlooked in healthcare conversations. People in prison tend to experience serious illness and aging much earlier than the general population, yet systems to identify when someone is approaching the end of life are often slow, inconsistent, or nonexistent. The new tool aims to change that.

Why End-of-Life Care in Prison Is a Growing Concern

The prison population in the United States is aging rapidly, and California is no exception. Decades-long sentences, combined with limited access to consistent preventive healthcare before incarceration, mean many incarcerated individuals develop chronic illnesses earlier than people outside prison walls.

Research has shown that people in prison often die years younger than their non-incarcerated counterparts. A notable 2012 study of New York state parolees found that each additional year spent in prison was associated with a two-year reduction in life expectancy. This stark reality highlights why proactive medical planning is especially important in correctional settings.

Despite this, end-of-life care in prisons is frequently delayed until someone is visibly and undeniably dying. By that point, opportunities for meaningful medical decision-making or compassionate release may already be lost.

How the Mortality Prediction Model Works

The newly developed mortality prediction model was created by UCSF researchers in close collaboration with physicians working inside California’s prison system. This partnership ensured the tool was grounded in real-world clinical experience, not just theory.

The model draws on data from approximately 90,000 incarcerated adults, making it one of the most comprehensive efforts of its kind. Using this large dataset, the researchers identified factors that reliably predict the likelihood of death within a two-year window.

Key factors included in the model are:

Age
Mobility and functional limitations
Serious medical conditions, such as cancer, end-stage liver disease, and advanced kidney disease
Whether a person is already living in a specialized prison healthcare facility, such as a medical or geriatric unit

Importantly, the model does not rely on a single diagnosis. While conditions like metastatic cancer clearly signal high risk, many incarcerated people live with multiple chronic but individually non-terminal illnesses that together place them at serious risk of death. The tool accounts for this cumulative burden.

The researchers found the model to be highly reliable in predicting two-year mortality, offering a practical way to identify people who might otherwise fall through the cracks.

What This Means for Palliative Care

One of the most immediate benefits of the tool is its potential to support earlier palliative care involvement. Palliative care focuses on comfort, symptom management, and quality of life rather than cure. In prison settings, it is often introduced far too late.

By identifying individuals at higher risk earlier, healthcare providers and prison staff can begin conversations about:

Appointing a healthcare power of attorney
Documenting preferences for life-sustaining treatments
Establishing do-not-resuscitate (DNR) orders where appropriate
Aligning medical care with a person’s values and wishes

These discussions are routine in many hospitals and nursing homes, but they are still relatively rare in correctional environments. The model provides a structured, evidence-based reason to start them sooner.

The Role of Frailty and Accelerated Aging

One of the key insights behind the research is that incarcerated individuals often experience geriatric conditions much earlier than expected. Frailty, cognitive impairment, and functional decline can appear years ahead of what clinicians typically see in the general population.

This accelerated aging is influenced by multiple factors, including chronic stress, past trauma, substance use, limited healthcare access before incarceration, and the physical and psychological demands of prison life. As a result, traditional age-based criteria for end-of-life planning often fail to capture true medical risk in prison populations.

The mortality prediction tool helps bridge that gap by focusing on function and disease burden, not just chronological age.

Compassionate Release and Why Timing Matters

California, like many states, has policies that allow for compassionate or medical release of incarcerated individuals who are terminally ill. In practice, however, these programs are dramatically underused.

One major reason is timing. By the time it becomes obvious that someone qualifies, the review process involving prison staff, medical teams, and the courts can take many months or even more than a year. Tragically, some people die before a decision is ever made.

The researchers hope the mortality prediction model will help identify eligible individuals earlier, giving the system enough time to act. Earlier identification could mean more people spending their final months with family, in hospice care, or in less restrictive medical settings.

Ethical and Policy Implications

Beyond individual care, this research raises broader questions about how society treats incarcerated people who are seriously ill. The model does not make decisions on its own, but it offers data that can inform fairer, more humane policies.

From a public health perspective, better end-of-life care in prisons can also lead to:

More appropriate use of limited healthcare resources
Reduced emergency hospital transfers
Improved coordination between correctional and community healthcare systems

It also challenges the assumption that prison healthcare must always be reactive rather than proactive.

Looking Ahead

The researchers emphasize that the tool is not meant to replace clinical judgment. Instead, it serves as a supportive decision-making aid, helping clinicians and administrators recognize risk earlier and respond more thoughtfully.

If adopted widely, this approach could mark a meaningful shift in how end-of-life care is handled in correctional facilities—not just in California, but potentially across the country.

At its core, the study highlights a simple idea: being incarcerated should not mean being excluded from compassionate, dignified medical care, especially at the end of life.

Research Paper Reference:
https://link.springer.com/article/10.1007/s11606-025-10103-w