Could a Sensor Solve Our Prison Suicide Problem?

Could a Sensor Solve Our Prison Suicide Problem?

By Melissa Pandika



Because this ultrasensitive, low-cost sensor system could help save lives in prisons.

By Melissa Pandika

Justin Helzer, 41, believed his brother was God, so he helped him kill five people in 2000. Diagnosed as schizophrenic and delusional, Justin first attempted suicide at San Quentin State Prison by jamming ballpoint pens through both eyes and into his brain. He tried again, three more times. The fourth attempt worked, and he was found in his cell after a staff member shift change, hanging by a bedsheet. 

Suicide is a leading cause of death in American jails, and is about three times higher in jails than in the general population, according to the National Institute of Corrections. It accounted for 5.5% of deaths in state-run facilities and 35% of deaths in local jails in the U.S. in 2011 — more deaths than those from drug and alcohol intoxication, homicide, and accidents combined.  

Suicide accounted for 5.5% of deaths in state-run facilities and 35% of deaths in local jails in 2011.

Today, prisons house those potentially at risk for self-harm in isolation cells, while additional personnel check on them every 15 minutes. Although this form of surveillance “tends to be highly effective, it’s very labor-intensive,” not to mention costly and invasive, says Jeffrey Ashe, a principal scientist at General Electric. And like Helzer, inmates can simply attempt suicide between check-ins and personnel shifts.

Ashe and other GE researchers have devised a potential solution: a wall-mounted sensor — about the size of a smoke detector — that allows correctional officers to constantly monitor inmates’ heart rates, breathing and movements, alerting them when it detects signs of self-harm. Ashe and his colleagues modified a standard home security system that uses Doppler radar (yes, the weather thing) to detect whether the subtle rise and fall of the chest during breathing — or even the delicate pulsing of the skin as the heart pumps blood through the veins — falls within the normal range. Convulsions and a slowed heart rate might indicate asphyxiation, a common method of suicide among inmates. Correctional officers would be responsible for monitoring the system’s data, which right now tests at about 86% accuracy, and responding when it alerts them to possible self-harm. 

An estimated 50% of the U.S. prison population suffers from mental illness, compared to about 25% of Americans overall.

GE’s system might be one of the first examples of technology to help buck the prison suicide trend. Also on deck? South Korea is testing a robotic prison guard that uses cameras and behavior-monitoring software to detect suicide and other trouble. It’s not all about high-tech, though: Some prisons in the U.K. have opened “first-night units” — bright, comfortable units with nurses and mental health workers, where incoming inmates stay for at least two nights to ease the sudden transition to prison life. Meanwhile, Australia’s Mount Gambier prison has a Prisoner Listener program, in which prisoners trained in recognizing suicidal behavior offer support and advice to new inmates. 

A grant from the U.S. Department of Justice has funded the research to develop and test the GE sensor, perhaps suggesting an increase in efforts among correctional facilities to tackle prison suicide. But these efforts might stem not from a concern for inmate welfare, but a fear of litigation from inmates’ families for not doing enough to prevent suicide. And while a sensor might prevent suicides, it won’t address the underlying causes, many of which arise from the prison system itself. (For example, suicidal inmates are typically placed in isolation — a major predictor of suicide, according to multiple studies.)

Many inmates already suffer from pre-existing mental illness or substance abuse problems, owing to a higher rate of negative life experiences — such as poverty or abuse — and limited access to mental health services. An estimated 50% of the U.S. prison population suffers from mental illness, compared to about 25% of Americans overall, according to a recent review article. What’s more, about 65% of the country’s inmates meet the criteria for substance abuse and addiction, but only 11% received treatment. 

Prisoners’ rights advocates view preventing inmate suicide as a matter of human rights. Suicide, they argue, denies inmates of the opportunity to reform their lives and contribute to society to the best of their abilities. And amid rising incarceration rates, “it’s not as uncommon today to hear, ‘Yeah, I know someone in prison,’” says Meredith Dye, an assistant professor of sociology at Middle Tennessee State University. 

It could also be used to track the vital signs of elderly individuals, premature infants or others who require close monitoring.

Earlier attempts to develop prison suicide sensors emerged in the 1980s. Lindsay Hayes, project director at the National Center on Institutions and Alternatives, recalls one researcher who failed to patent an earpiece to monitor pulse and oxygen levels, which would sound an alarm if they fell outside the normal range — or if the inmate simply removed the earpiece. Which goes to show how far the research has come. 

And it could still go further: GE is working to commercialize its sensor as part of its push to own the trillion-dollar burgeoning “Internet of Things” industry. Which means the sensor could eventually be used to track the vital signs of elderly individuals, premature infants in neonatal intensive care units or others who need close monitoring.

GE tested the system last year at the Western Correctional Institution in Cumberland, Maryland, on 10 prison staff members who spent about 90 minutes inside the facility’s cells — moving, breathing at different rates and holding their breath as if their breathing had stopped. It’s still in an early research stage, awaiting many more trials, and the researchers don’t yet know when the sensor system could appear in correctional facilites, or how much it would cost. Ashe says it “compares favorably” to home security systems, which can cost a little more than $1,000 per home.

Which means we might have a very solvable problem on our hands.