By JT Santana | jtwb768

Suicide prevention has never been an abstract public-health topic for me. More than three decades have passed since John died by suicide, yet there is no point at which that loss becomes a historical footnote. Time changes grief, but it does not erase the person, the questions, the memories, or the commitment that can grow from surviving such a loss.

That commitment is one reason I keep returning to suicide prevention. I want us to talk about it where people are comfortable talking about it, and I want us to talk about it where society would often prefer not to look. Jails and prisons belong in that second category.

Suicide remains a major cause of preventable death in the United States. Final federal data show 48,824 people died by suicide in 2024, producing a rate of 14.4 deaths per 100,000 population. Provisional National Center for Health Statistics data place the 2025 total at 48,789 deaths. That tiny numerical decline does not represent anything close to victory: nearly 49,000 people still died in a single year (Centers for Disease Control and Prevention [CDC], 2026).

Inside correctional institutions, the picture is particularly disturbing. The latest comprehensive federal mortality dataset covering local jails and state prisons still ends in 2019. That limitation needs to be stated plainly at the beginning of any serious discussion of correctional suicide in 2026. We now possess much newer information about how many people are incarcerated, yet the nationally comparable cause-of-death statistics for state prisons and local jails remain years behind (Bureau of Justice Statistics [BJS], 2021a, 2021b).

That data delay should trouble anyone concerned with accountability. We cannot credibly claim to address deaths in custody in real time when much of the national mortality picture available to researchers, journalists, families, policymakers, and the public describes conditions from seven years earlier. The available evidence still tells us something unmistakable: suicide is a persistent correctional-health emergency, the danger does not appear at one uniform point in incarceration, and it does not necessarily end when the cell door opens.

Jails and Prisons Are Different Suicide Environments

People routinely use jail and prison as interchangeable words. From a suicide-prevention perspective, that is a serious mistake. Jails and prisons operate differently, serve different populations, and expose people to different patterns of risk.

Jails are primarily local institutions. People cycle through them at an extraordinary rate, many remain there for relatively short periods, and a substantial portion of the population has not been convicted of the charge responsible for the current detention. The BJS reported 7.9 million jail admissions from July 2023 through June 2024, with an average daily population of 667,700. At midyear 2024, 658,700 people were held in local jails, and 68% were unconvicted (BJS, 2026).

Prisons function differently. Their populations consist primarily of people serving sentences after conviction, and incarceration frequently lasts years or decades. The most recently published BJS prisoner tables available as of September 2026 cover 2023, when approximately 1.254 million people were under state or federal prison jurisdiction at year-end. BJS still lists its Prisoners in 2024 report among forthcoming publications (BJS, 2025a, 2026b).

Those institutional differences shape suicide risk. In jail, the most dangerous period often arrives almost immediately. A person can move from ordinary community life to arrest, booking, strip searches, withdrawal from drugs or alcohol, loss of medications, fear about employment, uncertainty about court, separation from family, public humiliation, and confinement within a matter of hours. Some people enter custody already experiencing a psychiatric emergency, and others encounter a sudden concentration of stressors that can produce severe psychological distress.

Prison suicide follows another pattern. The first days matter, yet long sentences create additional periods of vulnerability: sentencing outcomes, appeals, disciplinary actions, isolation, family deaths, relationship breakdowns, deteriorating health, victimization, changes in custody classification, loss of hope, and the accumulated psychological effects of years spent in confinement. A suicide-prevention strategy built for one environment cannot simply be copied into the other.

The Jail Numbers Are Grim

The BJS recorded 355 suicides in local jails in 2019. Suicide was the leading single cause of death in local jails that year, with a mortality rate of 49 deaths per 100,000 jail inmates. Suicide accounted for approximately 30% of all jail deaths reported in 2019 (BJS, 2021a, 2021c).

The timing of those deaths may be the most important finding in the federal data. During 2015–2019, 44% of jail suicides occurred during the first week of incarceration. Roughly 66% occurred during the first 30 days, and about 12% occurred within the first 24 hours (BJS, 2021a). Those percentages change how suicide prevention should be viewed, since intake is not simply paperwork preceding housing assignment. The first hours and days constitute a documented period of heightened risk.

The federal statistics provide another finding that deserves attention. Across 2000–2019, people who had not been convicted accounted for almost 77% of jail suicide deaths (BJS, 2021a). That statistic requires careful interpretation. It does not establish that being legally unconvicted independently caused those suicides, since unconvicted people make up a large share of jail populations in the first place.

It does establish that suicide prevention in jails is inseparable from the treatment of people awaiting adjudication, many of whom may eventually be acquitted, have charges dismissed, receive noncustodial sentences, or resolve their cases without a prison term. The modern jail population makes that point even harder to ignore. In 2024, more than two-thirds of people held at midyear had not been convicted on their current charge. A system housing hundreds of thousands of legally unconvicted people cannot treat mental-health care as an optional service reserved for those serving long sentences (BJS, 2026).

There is another reason to resist simple national comparisons. The 2019 jail suicide rate of 49 per 100,000 and the general U.S. suicide rate are calculated from populations with very different age, sex, health, and demographic profiles. Quoting a simple ratio between those rates can be useful descriptively, but it should not be mistaken for a fully adjusted estimate of an individual person’s relative risk. Precision requires acknowledging what a statistic can prove and what it cannot.

Prison Suicide Is a Long-Term Problem, Not Merely an Intake Problem

State prisons reveal a different pattern. The number of suicides in state prisons increased from 168 in 2001 to 311 in 2019, an increase of 85%. The average annual suicide rate rose from 15 per 100,000 prisoners during 2001–2004 to 21 per 100,000 during 2015–2019 (BJS, 2021a).

Federal prisons recorded 29 suicides in 2019, up from 18 in 2001. During 2015–2019, federal prisons had an average suicide rate of approximately 16 per 100,000 prisoners (BJS, 2021a). Unlike the concentration of jail suicides near admission, prison suicides frequently occur much later. Roughly 75% of state-prison suicides and 64% of federal-prison suicides during 2015–2019 occurred after the person had been incarcerated for more than one year (BJS, 2021a).

That finding dismantles the idea that a successful intake screening clears someone indefinitely. A person can truthfully report no suicidal thoughts during intake and experience a devastating crisis six months, five years, or fifteen years later. Human beings do not remain psychologically frozen at the moment they enter a facility. Risk changes with circumstances, which means prison systems need ongoing suicide surveillance rather than a one-time questionnaire buried inside an admission file.

The federal Bureau of Prisons offers a sobering case study. In 2024, the Department of Justice Office of Inspector General released an evaluation covering 344 inmate deaths from fiscal years 2014 through 2021 classified as suicide, homicide, accident, or unknown. Investigators analyzed 183 suicide psychological reconstruction reports and found recurring problems involving suicide-risk assessments, mental-health classifications, communication among departments, required staff rounds and counts, emergency response, and facility suicide drills (DOJ OIG, 2024).

More than half of the people who died by suicide in the OIG review had been housed alone. In more than one-third of the suicides examined, investigators found deficiencies involving required rounds or counts. The OIG found broader recordkeeping failures as well: for 149 of the 344 deaths reviewed, the BOP could not provide one or more documents required under its own policies (DOJ OIG, 2024).

Those findings do not mean every suicide was preventable through a single intervention. Suicide is rarely reducible to one cause, one decision, or one employee. They do show that systems can create layers of protection, and failures across several layers can leave a person in crisis with very little between suicidal behavior and death.

Mental Illness Is Common, but Suicide Cannot Be Reduced to Diagnosis

Correctional suicide discussions often begin and end with mental illness. Mental-health conditions are an important component of risk, yet that explanation can become too convenient. It can shift attention from institutional conditions to an individual’s diagnosis.

The BJS Survey of Prison Inmates found that 43% of state prisoners and 23% of federal prisoners reported a history of a mental-health problem in 2016. About 14% of state prisoners and 8% of federal prisoners met the threshold for serious psychological distress during the 30 days preceding their interview. Major depressive disorder was the most commonly reported diagnosis, reported by an estimated 27% of state prisoners and 14% of federal prisoners (BJS, 2021d).

Those figures are important, yet their age is revealing. The underlying national prisoner survey was conducted in 2016. A new Survey of Inmates in Local Jails was carried out during 2024–2025 and collected information on physical health, mental health, substance use, treatment, and experiences in jail, but national reports from that collection remain in development (BJS, 2026c).

A suicide-prevention system cannot assume everyone at risk will arrive with a documented psychiatric diagnosis. Acute grief, withdrawal, shame, fear, insomnia, legal shock, sexual victimization, threats from others, family separation, sudden sentencing news, medical deterioration, and disciplinary events can alter risk. The National Commission on Correctional Health Care’s current standards reflect that wider concept.

Its 2026 mental-health standards strengthen suicide-prevention requirements, including direct staff observations for people placed on safety precautions and follow-up by qualified mental-health professionals after those precautions are discontinued. Video monitoring may supplement direct observation; it is not treated as an adequate substitute for human checks (National Commission on Correctional Health Care [NCCHC], 2026). That distinction is fundamental: a camera can record a crisis, but human intervention can interrupt one.

Solitary Confinement Cannot Be Separated From the Suicide Discussion

Restrictive housing deserves particular scrutiny. Research has repeatedly associated solitary confinement with adverse mental-health outcomes, and the connection does not disappear once someone leaves prison.

A large North Carolina cohort study followed 229,274 people released from state prison between 2000 and 2015. People who had experienced restrictive housing had a 24% greater hazard of death during the first year after release compared with people who had not experienced it. Their hazard of death by suicide during that first year was 78% higher after adjustment for measured characteristics (Brinkley-Rubinstein et al., 2019).

This was an observational study, so the association should not be described as proof that restrictive housing directly caused every excess death. People placed in segregation can differ from other incarcerated populations in ways that are difficult to measure completely. The study remains significant for its size, adjustment for several known risk factors, and the strength of the association.

Newer evidence has kept the issue alive. A study examining North Carolina prisons from 2021 through 2023 found that nearly one-quarter of 41,525 newly incarcerated people experienced solitary confinement. The observed all-cause mortality rate was 4.23 per 100,000 person-weeks among those exposed to solitary confinement, compared with 1.96 among those never exposed, though the study recorded only 43 in-custody deaths and cannot establish causation from those figures alone (Cloud et al., 2025).

International human-rights standards draw an important distinction concerning terminology. The United Nations Nelson Mandela Rules define solitary confinement as confinement for 22 hours or more per day without meaningful human contact. The rules define confinement exceeding 15 consecutive days as prolonged solitary confinement and prohibit prolonged or indefinite solitary confinement. United Nations experts have stated that prolonged solitary confinement can, depending on the circumstances, amount to torture or cruel, inhuman, or degrading treatment (United Nations, 2015).

It is more accurate to say that prolonged solitary confinement is prohibited under the Mandela Rules and may meet international standards for torture or ill-treatment than to declare every instance of isolation automatically equivalent to torture. That precision strengthens the argument rather than weakening it. Suicide prevention does not need exaggerated claims, since the documented evidence is disturbing enough.

Release Is Not the End of the Risk Period

Opening the prison gate does not automatically restore someone’s psychological safety. People leaving incarceration can encounter an abrupt collision of pressures: housing insecurity, unemployment, family conflict, parole requirements, untreated psychiatric conditions, interrupted medication, substance-use risk, debt, stigma, transportation problems, health problems, and the basic difficulty of rebuilding ordinary life after institutionalization. A release date can represent freedom and still produce fear.

A 2023 study examining people released from North Carolina prisons between 2000 and 2020 found that formerly incarcerated people experienced approximately twice the suicide mortality of the general state population during the first three years after release. The highest suicide mortality occurred during the first two weeks after release (Janca et al., 2023).

Washington State research reached a similar finding using a cohort of 140,281 people released from prison. Among them, 484 later died by suicide. After adjustment for age, sex, and race, previously incarcerated people had a 62% greater suicide risk than the comparison population (Morgan et al., 2022).

An international individual-participant meta-analysis published in 2024 assembled data on more than 1.47 million people released from incarceration across eight countries. Mortality was highest during the first week after release. Suicide mortality during that first week was estimated at 135 deaths per 100,000 person-years, though the confidence interval was wide, reflecting uncertainty around the estimate (Kinner et al., 2024).

These findings point toward a correction in how society thinks about reentry. Release should not be treated as a clean administrative handoff in which the correctional system’s responsibility simply disappears at midnight. If someone has been receiving psychiatric medication on Monday, releasing that person Tuesday without medication continuity, an appointment, transportation, insurance access, or a realistic plan for where they will sleep can turn a bureaucratic transition into a medical hazard.

NCCHC has urged correctional systems to improve insurance continuity before and after release, citing disruptions in mental-health and substance-use treatment as contributors to poor health outcomes and death. Access to health coverage is one part of reentry; actual access to clinicians, prescriptions, housing, transportation, and community support determines whether coverage translates into care (NCCHC, 2024).

The First Week Should Be Treated Like the High-Risk Period It Is

The jail statistics point to an obvious operational lesson. When 44% of jail suicides in the most recent comprehensive federal study occurred within seven days of admission, prevention cannot wait for a routine mental-health appointment scheduled several weeks later (BJS, 2021a). Screening at intake remains necessary, but screening must be treated as the beginning of assessment rather than its endpoint.

People may conceal suicidal thoughts out of fear that disclosure will produce isolation, loss of clothing, humiliation, or placement in an unpleasant observation cell. Others may not be suicidal during booking and become suicidal after a court hearing, family call, withdrawal episode, assault, medical diagnosis, or devastating personal news. A correctional suicide-prevention system needs repeated reassessment at predictable transition points.

The relevant moments include admission, movement into restrictive housing, return from court, sentencing, receipt of bad legal news, major disciplinary events, serious medical diagnoses, bereavement, changes in psychiatric medication, reported victimization, release from suicide precautions, and preparation for community reentry. Communication between custody and health staff is just as important.

The person who notices someone giving away possessions may be an officer. The person who learns about a recent family death may be a case manager. The person who recognizes severe depression may be a nurse. The person who hears a suicidal statement may be another incarcerated person, and a functioning system gets that information to someone capable of acting before fragments of warning disappear into separate files, shifts, departments, or computer systems.

Suicide Watch Cannot Become Punishment

There is a profound contradiction in some correctional responses to suicide risk. A person discloses suicidal thoughts and is then stripped of ordinary possessions, deprived of normal human contact, moved into an austere cell, watched through glass, and left with little therapeutic interaction. Some safety restrictions can be clinically necessary during an acute crisis, but the danger appears when suicide precautions become so punitive or degrading that people learn to hide suicidal thoughts to avoid them.

That creates exactly the wrong incentive. A person should not have to choose between disclosing suicidal thoughts and preserving basic dignity. Safety procedures should be clinically driven, regularly reassessed, and paired with actual treatment rather than functioning solely as containment.

The most recent NCCHC guidance for youth illustrates the broader principle. Its 2025 position statement calls for standardized screening, continuing reassessment, direct observation according to level of risk, clear communication across staff, clinical treatment, and formal review after a death. It states that suicide monitoring should never be used as punishment or retaliation (NCCHC, 2025).

Adult correctional systems should be judged by the same basic question: does suicide watch exist to keep a human being alive long enough to receive care, or does it merely protect the institution from immediate liability? Those are not the same objective, and conflating them can produce procedures that look protective on paper yet feel punitive to the person experiencing them.

Better Data Are Part of Prevention

One of the most troubling findings of this research is not a death count. It is the calendar. In September 2026, researchers can tell us how many people local jails held in 2024, how many admissions occurred, how many people were unconvicted, and how jail populations changed over the preceding decade. Yet the most recent comprehensive national BJS data describing who died by suicide in local jails and state prisons end in 2019.

Seven years is an unacceptable distance between a preventable death and national public accountability. Conditions changed dramatically after 2019. The COVID-19 pandemic altered correctional populations, staffing, health services, isolation practices, court operations, visitation, drug markets, and community reentry. Opioid and polysubstance mortality changed, mental-health systems changed, and jail populations declined and then shifted again.

A 2019 mortality dataset remains useful for identifying longstanding patterns. It cannot tell us with confidence what happened nationally in jails last year. Federal reporting needs to become timelier, more complete, and easier for the public to analyze.

Every custodial death should generate standardized information on location, legal status, time in custody, housing status, known mental-health history, recent clinical contacts, suicide precautions, recent disciplinary events, restrictive housing exposure, relevant medications, prior self-harm, emergency response, and independent review findings, subject to privacy protections that prevent identification where required. Numbers do not bring anyone back, but they can reveal failures that institutions would otherwise repeat.

Prevention Requires More Than Telling People to Ask for Help

Public suicide campaigns frequently tell people in crisis to reach out. That message has value in the community, but incarceration changes the equation. An incarcerated person cannot simply drive to an emergency department, change therapists, leave an unsafe housing unit, call a friend whenever needed, choose another pharmacy, go home, or remove himself from a threatening environment.

Access to almost every protective resource is controlled by someone else. That shifts greater responsibility onto the institution. Correctional suicide prevention needs adequately staffed mental-health services, prompt psychiatric evaluation, medication continuity, safe housing, direct observation when clinically indicated, access to therapeutic contact, staff training, effective communication, post-watch follow-up, reduced dependence on isolation, meaningful human contact, family communication where appropriate, competent emergency response, and rigorous mortality review.

Reentry needs similar seriousness. A person leaving incarceration with known suicide risk should not receive a folder of phone numbers and be expected to build a treatment system alone. Appointments should be arranged before release, medication continuity should be settled before release, insurance enrollment or reactivation should begin before release, and housing and transportation barriers should be addressed as part of health planning rather than treated as unrelated social inconveniences.

The evidence gives us no justification for treating the prison gate as the end of the risk period. The transition back into the community is itself one of the periods in which careful intervention may have the greatest capacity to prevent another death.

John Is Why I Refuse to Treat These Deaths as Statistics

When I read that 355 people died by suicide in American jails in one year, I know what the number means statistically. I know that epidemiology requires counts, denominators, confidence intervals, comparison populations, and careful language. I know something else, too: every number eventually becomes somebody’s empty chair.

John has been gone for more than thirty years. People might assume that three decades should make that sentence easier to write. It does not work that way. Grief can stop being acute without becoming irrelevant.

A person can build a life, laugh, work, advocate, love other people, survive terrible things, and still carry a death from decades earlier into the present. Suicide leaves questions that have no expiration date. That history is one reason I resist language that makes incarcerated people disappear into categories such as offender, inmate, detainee, or number.

Legal status can describe someone’s relationship to a court or correctional system. It does not describe the entire human being sitting in a cell at three in the morning wondering whether tomorrow is worth seeing. Some people reading this will respond that incarcerated people made choices that led them there.

Sometimes they did. Some have caused tremendous harm. Accountability for criminal conduct and protection from preventable suicide are not competing propositions. A sentence imposed by a court is the sentence, and suicide should never become an unofficial extra punishment.

What We Owe People in Custody

The evidence does not support a fatalistic view that suicide simply happens in correctional facilities and nothing substantial can be done. We know that jail risk is concentrated heavily in the first days and weeks. We know prison risk persists deep into long sentences. We know restrictive housing is associated with adverse mental-health and mortality outcomes.

We know many incarcerated people enter custody with substantial mental-health histories. We know risk rises sharply again after release. We know communication failures, missed rounds, inadequate assessments, insufficient follow-up, and poor emergency preparation have appeared in federal death reviews. None of that gives us the identity of the next person who will die, but it gives us multiple places to intervene.

That is what suicide prevention really looks like. It is rarely one dramatic rescue. It is a chain of ordinary decisions made correctly: the booking officer who takes a statement seriously, the nurse who verifies medication, the mental-health clinician who reassesses after bad news from court, the officer who notices a behavioral change, the administrator who removes a dangerous environmental feature, the reentry worker who schedules an appointment, the agency that reviews a death without protecting itself from uncomfortable findings, and the public that insists incarcerated lives count.

September may give us a designated month to talk about suicide prevention. People in crisis do not organize their suffering around awareness calendars, and neither should we. John’s death taught me long ago that suicide leaves a circumference far larger than the moment of death itself.

Families live with it. Friends live with it. Communities live with it. Correctional officers and incarcerated witnesses can live with it, and the people who survive an attempt live with what happened long after everyone else assumes the crisis has passed.

More than thirty years later, I still believe prevention is worth fighting for. That belief does not stop at a jail door, a prison wall, or a release gate. It cannot, if we are serious about treating suicide as preventable rather than inevitable.

If You or Someone You Know Is in Crisis

In the United States, the 988 Suicide & Crisis Lifeline can be reached by calling or texting 988, with chat available through the 988 Lifeline service. People in correctional custody may have restricted telephone or internet access, which makes facility suicide-prevention procedures, staff reporting, clinical access, and communication from family members particularly significant.

If someone in custody appears to face an immediate suicide risk, family members or advocates can contact the facility directly and ask that the concern be sent to medical or mental-health staff rather than relying solely on ordinary correspondence. When immediate danger exists outside a correctional setting, 988 or emergency services can connect a person with crisis support.

References

Brinkley-Rubinstein, L., Sivaraman, J., Rosen, D. L., Cloud, D. H., Junker, G., Proescholdbell, S., Shanahan, M. E., & Ranapurwala, S. I. (2019). Association of restrictive housing during incarceration with mortality after release. JAMA Network Open, 2(10), e1912516.

Bureau of Justice Statistics. (2021a). Suicide in local jails and state and federal prisons, 2000–2019: Statistical tables. U.S. Department of Justice.

Bureau of Justice Statistics. (2021b). Mortality in state and federal prisons, 2001–2019: Statistical tables. U.S. Department of Justice.

Bureau of Justice Statistics. (2021c). Mortality in local jails, 2000–2019: Statistical tables. U.S. Department of Justice.

Bureau of Justice Statistics. (2021d). Indicators of mental health problems reported by prisoners: Survey of Prison Inmates, 2016. U.S. Department of Justice.

Bureau of Justice Statistics. (2025). Prisoners in 2023: Statistical tables. U.S. Department of Justice.

Bureau of Justice Statistics. (2026). Jail inmates in 2024: Statistical tables. U.S. Department of Justice.

Centers for Disease Control and Prevention. (2026). Suicide data and statistics. National Center for Injury Prevention and Control.

Department of Justice, Office of the Inspector General. (2024). Evaluation of issues surrounding inmate deaths in Federal Bureau of Prisons institutions.

Janca, E., et al. (2023). Suicide mortality among formerly incarcerated people compared with the general population in North Carolina, 2000–2020. American Journal of Epidemiology.

Kinner, S. A., et al. (2024). Rates and causes of death after release from incarceration among 1,471,526 people in eight high-income and middle-income countries: An individual participant data meta-analysis. The Lancet.

Morgan, E. R., Rivara, F. P., Ta, M., Grossman, D. C., Jones, K., & Rowhani-Rahbar, A. (2022). Incarceration and subsequent risk of suicide: A statewide cohort study. Suicide and Life-Threatening Behavior, 52(3), 467–477.

National Commission on Correctional Health Care. (2024). Optimizing insurance coverage for individuals pre- and postrelease.

National Commission on Correctional Health Care. (2025). Suicide prevention and management of youth in custody.

National Commission on Correctional Health Care. (2026). Mental health standards.

United Nations. (2015). United Nations Standard Minimum Rules for the Treatment of Prisoners (the Nelson Mandela Rules).