Using a Screening Program to Improve Suicide Prevention

Brian Ward

Sentinel Event Alert Sheds Light on Common Problem

Suicides were the third most common sentinel event of 2015, with 95 reported cases in 2015’s Sentinel Event Statistics. The total number of patient suicides reported to The Joint Commission is now up to 1,184 since the start of the decade.

That said, only 2% of sentinel events are reported to The Joint Commission, and the cases reported are only the ones that occurred inside a healthcare facility or within 72 hours of discharge. Nationally, suicide is the 10th leading cause of death, with 9.3 million adults having suicidal thoughts, 1.3 million attempting suicide, and 41,149 deaths in 2013. In addition to the loss of life, suicides cost $51 billion in combined medical and work costs annually.

In February, The Joint Commission released Sentinel Event Alert 56 ( to highlight the detection and treatment of suicidal patients. The accreditor found that 21.4% of accredited behavioral health organizations and 5.14% of accredited hospitals are noncompliant with National Patient Safety Goal 15.01.01, which focuses on suicide prevention. The alert calls on healthcare facilities to improve suicide prevention compliance by establishing suicide screening programs (SSP) to identify at-risk patients.

Universal screening

Many hospitals fall short in screening for suicidal ideation, if they screen at all, says Julie Goldstein, PhD, director of health and behavioral health initiatives at the Suicide Prevention Resource Center (SPRC) for the past three years. She also leads the organization’s Zero Suicide Initiative.

Industry attitudes toward suicide screening have changed, Goldstein says.

“Up until recently, the U.S. Preventive Services Task Force [USPSTF] only recommended screening in the event that there were services available to support the screening results,” she says. “So they said you should screen, but you should only really screen when you have those available services. Because you can’t screen positive and then have nowhere to make the referral. That’s a real limit.”

The USPSTF has since changed its stance and now recommends universal screening, as does The Joint Commission. Goldstein says these endorsements will hopefully aid suicidal patients in finding help.

First screening program in the U.S.

In 2014, the Parkland Health and Hospital System in Dallas became the first in the nation to establish a universal SSP in all its departments. The health system was one of the few organizations acknowledged in Sentinel Event Alert 56 for making significant progress in suicide prevention.

Celeste Johnson, DNP, APRN, PMH CNS, director of nursing and psychiatric services at Parkland, helped create the SSP. “The Joint Commission said you really should be screening everybody and coming up with an evidence-based tool,” Johnson says. “So we started meeting with them and developed a tool for everyone at Parkland. We pulled together all the stakeholders: IT, nurses, social workers, and there was lots of communication with administration.”

The program had been proposed a few years prior to implementation by Kimberly Roaten, PhD, director of quality for safety, education, and implementation in the hospital’s Department of Psychiatry. Roaten has a background in suicide prevention training and was a big part of the SSP’s development and implementation.

“From my perspective, the biggest hurdle was making sure of what resources would be needed to respond if a patient screened positive,” Roaten says. “Part of that was getting people in place to respond, and the biggest part was building our own clinical response algorithm.”

The Joint Commission classifies patient suicides as ones that occur inside a hospital or within 72 hours of discharge. However, that metric omits those who commit suicide in the days, weeks, and months after that 72-hour window. It also doesn’t show the number of patients who have suicidal thoughts or intentions, making it difficult to discover the percentage of patients who are at risk.

After the program’s first year, Parkland created a data sheet revealing the percentage of its patients who experienced suicidal ideation. Of the 243,000 patients to visit the hospital’s ER and inpatient units in 2015, 4% were moderate suicide risks and 2% were high risk. Of the 259,000 screened in the hospital’s community-oriented primary care facilities and correctional health programs, 2% were moderate risk and 0.15% were high risk.

Simply put, out of the 502,000 patients that Parkland saw in one year, 14,900 were at moderate risk and 5,249 were at high risk for suicide. That’s equal to one out of every 25 patients experiencing suicidal thoughts, many of whom wouldn’t have had those thoughts acknowledged without screening in place. Parkland also found that within the year prior to their death, 77% of people who committed suicide had contact with a primary care physician, 40% had contact with an ER physician, and 24% had contact with a mental health provider.

Use a standardized screening tool on everyone

An SSP must be standardized and applied universally to all patients, Goldstein says. While patients may not show some of the risk factors associated with suicidal ideation, it’s dangerous to assume you have the complete picture of their mental health. “Many times when people screen, they aren’t using a standardized screening tool,” she says. “Sometimes screening to them is saying, ‘You’re not thinking about suicide, are you?’ or, ‘You’d never kill yourself. Right?’ Which is kind of the opposite of how we want people to act. What that says to an individual is, ‘I’m not sure what I’ll do with your answer,’ or, ‘I’m unprepared,’ or, ‘I’m really hoping you’ll say no.’ And as hard as it is to disclose suicide, if the other person on the receiving end of that conversation doesn’t want you to say ‘Yes,’ then people will often not say ‘Yes.’ ”

Goldstein also notes that many informal screening programs only ask a patient once about suicidal ideation. If you only screen once and something changes in the person’s life three months later, a physician may not realize the patient has become a suicide risk, she points out.

Roaten says that Parkland’s call to action came when an unscreened inpatient with a substance abuse problem attempted suicide. Afterwards, Parkland started using the Columbia Suicide Severity Rating Scale (C-SSRS) to screen adults and the Ask Suicide-Screening Questions (ASQ) tool for children aged 12–17. Both tools are freely available for facilities to download and use, as are dozens of others that can be found at the SPRC’s website (

Parkland and the SPRC say a key part of an effective SSP is how screening tools are embedded into the electronic health record (EHR). At Parkland, Johnson and Roaten worked with their IT department to create an EHR that would prompt a user to ask the SSP questions, display a safety “banner” to identify the patient’s risk level, and provide instructions on what to do next.

“A nurse is prompted to ask the [screening] questions immediately,” Roaten says. “And if they can’t ask at that moment, they can silence that ‘best practice’ alert and it will pop up again a short while later. We also built a flag in EHR for anyone who’s high risk: They list out warning signs, plus a mental health facility to call.”

Ideally, the EHR is programmed so that staff can’t skip over the screening questions, or so they will be continually prompted until the questions have been asked and documented. When Parkland first implemented the EHR, Johnson and Roaten would email unit managers a list of the names of people who hadn’t been screened that week. As staff grew better at following the prompts, those emails were replaced with the number of people successfully screened.

Have next steps planned

Don’t forget to look ahead.

“[It’s important] that everybody in the hospital system really understands what happens next,” says Goldstein, “because if you screen people and they are at risk for suicide, there should be ongoing protocols and activities that happen next that everybody’s aware of, including the patient.”

Educate patients as to why they are being screened and what will happen if they screen positive, Goldstein says. The conversation should be like the one you give about a recent cancer diagnosis or high blood pressure, explaining the expectations for what happens next and what the ongoing care will be. Parkland’s EHR categorizes patients as being at “no risk,” “moderate risk,” or “high risk” of suicide and gives on-the-spot recommendations on how to proceed. Moderate-risk patients meet with a psychiatric social worker for assessment and referral, while high-risk patients are immediately sent to a behavioral healthcare provider.


Brian Ward is associate editor of Briefings on Accreditation and Quality. This article was originally published in the May 2016 issue of Briefings on Accreditation and Quality.