The trusted source for
healthcare information and
The ED at Parkland Hospital & Health System in Dallas has implemented a universal screening program to identify patients at risk for suicide. A six-item suicide screen is administered at triage, with patients stratified into three risk groups based on their answers: no risk identified, moderate risk identified, and high risk identified. Patients receive specific interventions based on their risk group.
With the suicide rate on the increase over the past decade, suicide is now the 10th leading cause of death in the United States and very much on the radar of public health authorities. Further, while The Joint Commission (TJC) requires accredited hospitals to assess for suicide risk in patients with behavioral health issues, studies suggest that a relatively high number of patients seeking care for nonpsychiatric issues are at risk of suicide as well. Many of these patients present to EDs for care, creating the opportunity to intervene.
However, can a universal screening approach for suicide be implemented efficiently in a busy emergency setting where taking care of acute problems is the primary focus? Parkland Hospital & Health System (PHHS) in Dallas has demonstrated that it can, and administrators there believe their approach, which also extends to the inpatient and outpatient settings, could be adapted for use in other hospital systems.
The impetus for developing a universal screening program for suicide at PHHS stemmed from the clear clinical need for such an approach, explains Kimberly Roaten, PhD, CRC, the director of quality for safety, education, and implementation in the department of psychiatry at Parkland, and associate professor of psychiatry at the University of Texas Southwestern Medical Center. “We know from existing research that patients who die by suicide are much more likely to be seen by [primary care physicians] and emergency medicine providers in the months and years before they die, and that they are much less likely to be seen by psychiatrists and psychologists,” she explains.
In addition, during a routine accreditation survey by TJC at Parkland in 2014, surveyors found that the hospital had neglected to conduct a suicide risk assessment for a medical inpatient who had a history of substance use. “Because he had that existing psychiatric condition, according to a national patient safety rule, he should have been assessed for suicide risk, and he was not,” Roaten notes.
That incident, combined with the clinical need for screening, prompted administrators to think creatively about how they could both stop missing patients who were potentially at risk as well as pick up on occult suicide risk — cases in which patients only disclose suicidal thoughts or behaviors if they are specifically asked about them.
With more than 250,000 patient encounters per year, the ED at Parkland offers a rich opportunity for identifying suicide risk, but developing a way to implement a universal screen without adversely affecting patient flow or capacity took some time. Safety issues also were a big concern.
“We very intentionally decided to put the screening questions in the triage process in the ED for a couple of reasons. One, we didn’t want patients in the waiting room in the ED with suicide risk that we didn’t know about, so we wanted to ask them [these questions] as soon as possible, but we also wanted to pose these questions with a bit more privacy,” Roaten explains. “So, as the nurse is taking vital signs and asking other basic questions, that is when these [suicide risk screening] questions are asked, and more specifically, they are asked in the context of other psychosocial screening questions, such as questions about alcohol use, drug use, and domestic violence.”
Parkland uses the Columbia Suicide Severity Rating Scale (C-SSRS), a validated six-item screen for patients 18 years of age and older. “Everybody gets question one and two, but if they respond ‘yes’ to question two, then they get questions three, four, five, and six. If they respond ‘no’ to question two, then we skip down to question six,” Roaten notes. All the questions are prompted through the electronic medical record (EMR), and it takes about two minutes, on average, to conduct the screen.
Developing an approach for how to respond to the screening results required some additional work, Roaten says. “The piece that was missing when we first tried to start this program was a way to translate the standardized screening tool that already existed into the EMR in a meaningful way that was user-friendly for frontline staff,” she explains. “So we created what we call the Parkland Algorithm for Suicide Screening. It is a weighting system for the screening items built into the EMR.”
For example, when the ED nurse asks each screening question, which she reads verbatim from the computer screen, and then enters the responses of “yes” or “no,” the EMR will prompt the nurse to take appropriate clinical actions, based on three distinct risk categories: no risk identified, moderate risk identified, and high risk identified. “The vast majority of our patients, about 96% in the ED, are completely negative. They say ‘no’ to all the screening items, and they fall into that no risk category,” Roaten observes. Most other patients fall into the moderate risk group, which is a very important group in terms of efficiency and resource allocation, she says.
“In the past, we were sort of throwing the whole ball of wax at everybody, and, frankly, not just for suicide risk, but anything psychiatric,” Roaten observes. “We were putting everybody on one-to-one [monitoring] and taking away their belongings.”
However, under Parkland’s new protocol, patients categorized as moderate risk are evaluated by a social worker, connected to outpatient mental healthcare, and given printed mental health resources, including information about suicide warning signs and crisis hotlines.
Patients in the high-risk group include individuals who are reporting current suicidal ideation or a very recent suicide attempt such as within the past week. “These patients are put on one-to-one [monitoring], they have suicide precautions in place, and they are required to be evaluated by a behavioral health provider,” Roaten explains. “This could be a social worker with suicide risk assessment competency, a psychologist, a psychiatrist, or a behavioral health nurse practitioner or physician assistant.”
Patients at high risk also would receive the same connection to outpatient mental healthcare and the same printed list of mental healthcare resources as the moderate risk group. “They could, of course, also be hospitalized for their suicide risk, but most are not,” Roaten notes.
In a study of the approach post-implementation, investigators found that 6.3% of patients presenting to the ED screened positive for suicide risk at either the moderate- or high-risk level.1 Those findings have remained stable, with between 6% and 7% of all emergency patients consistently screening positive for suicide risk at either the moderate- or high-risk level, Roaten shares.
Notably, the universal screening program extends to Parkland’s inpatient and outpatient care settings, too; however, the study data show that the odds of a positive suicide screen are higher in the emergency patient population. This finding could translate to many other EDs, but not necessarily all of them, Roaten advises. “In general, patients who show up in our Parkland ED have more psychosocial stressors that are associated with suicide risk factors than our patients who are traditionally seen in our primary care settings,” she says.
For example, Roaten notes that a high number of emergency patients face financial and social stress. Many have limited social support, employment challenges, primary relationship issues, and psychiatric comorbidities. “There is just a higher concentration of people with these issues in the emergency setting than there is in other places, particularly at Parkland,” she says.
Although the universal screening program has been in place in the ED since 2015, administrators recall that in the early days of development, there were concerns about adding one more screening task to busy frontline providers. “The medical director of the ED at the time was concerned that it might slow them down,” advises Celeste Johnson, DNP, APRN, PMH, CNS, vice president of nursing for behavioral health at PHHS.
However, the data regarding the rising rate of suicides and the number of patients at risk who get missed when they visit healthcare providers proved convincing. “I think there was a basic belief that this was the right thing to do,” Johnson notes. “We did not believe we would have a large percentage of patients that would be of high or moderate risk of suicide, but we felt like it was worth those two minutes to screen.”
Johnson notes that Parkland is fortunate that it is a teaching facility and that it operates a psychiatric ED, but a key element to the screening program’s success is targeting resources in the community for patients who screen positive for suicide risk. “There is not a wealth of care providers in the community for mental health; it is always a little bit tougher to get into,” she says. “But we have really been fostering those relationships.”
With funding for indigent care a continuing challenge, the psychiatric ED has experienced longer boarding times of late for patients in need of psychiatric beds, observes Johnson, but she notes that for most patients who require handoffs to outpatient mental health providers, the effort to nurture relationships with other agencies has helped tremendously.
Another key pillar of the screening program’s implementation process was education for all staff about suicide risk and the program’s components. “We created two separate education modules — one went to the people who would actually be administering the suicide screen — the nurses and nursing assistants, and the other module went to the physicians, house staff, and everybody else,” Roaten explains. The people conducting the screening received basic education about suicide risk, including the statistics about the number of people at risk who are seen by primary care physicians or emergency providers vs. behavioral health specialists. Further, this group was required to view a 30-minute video provided by the group that created the screening tool instrument, and they received some scenario-based instruction, including tips on how to respond to patients in problematic circumstances, such as when patients refuse to answer screening questions or patients state that they don’t want to see a behavioral health provider.
The second education module was much simpler, Roaten notes. “We did not require this group to view the video,” she says. “Instead, it consisted of just the basic education about suicide risk, why we were screening for suicide risk, and basic information about how it would work in our EMR.”
One early issue that emerged was a small subgroup of patients categorized as high risk would want to leave the ED before they could be evaluated by a behavioral health provider. Some of these patients did not meet emergency detention criteria, so the ED could not hold them. To alleviate this problem, administrators finessed the workflow so that behavioral health staff could respond to these cases quicker.
“We prioritized those activations in our ED so that we are very aware when those patients are asking to leave and need to be seen as quickly as possible,” Roaten explains. “We also have a very responsive police department that will help us convince these patients to stay, and we have provided additional education for the nursing staff about how they might talk to a patient about staying for the evaluation even when they don’t want to.”
Further, there is now a report released every morning that identifies any patient who was high risk, did not meet emergency detention criteria, and insisted on leaving the ED. “Then we follow up with those patients by phone, so our social workers will call them and try to do an assessment and make sure they have some kind of safety plan if that is needed,” Johnson observes. “The social worker will find out what the patient’s follow-up plans are, and sometimes those patients are ready to come back and really do need to be seen.”
There was some early hesitancy from the hospital social workers about getting involved in suicide screening. “They were fairly resistant to the idea of independently doing some of the suicide risk assessments. They were nervous about it,” Roaten notes. “But once we talked about what their clinical training looked like, and what their true competencies were, we actually came to believe — and I think the social workers came to believe as well — that we were underutilizing their clinical skills.”
The social workers are fully involved in the screening program at this point, and they have embraced their role. “Now that we have gotten past that initial anxiety, it is very much a part of their clinical practice, and something they seem to be very proud of,” Roaten adds.
Now that there are ample data about the number of patients who screen positive for suicide risk and a formalized risk stratification process that dictates what the next steps are for these patients in terms of interventions, administrators hope to turn their attention to measuring results, a task that is not without complications.
“We are looking at a number of different types of outcomes. Obviously, one of the biggest limitations with suicide research data has been our limited ability to use suicide as a true outcome variable,” Roaten explains. “It is a rare enough event that it is very hard to get the power you need statistically, but with a program this size we can do that.”
Investigators hope to focus on what specific elements of the program lead to reductions in suicide — the screening itself or perhaps specific interventions.
“We are looking at suicide outcomes, but we are also looking at other things like return visits for suicide attempts or other types of self-inflicted violence,” Roaten notes. “We are looking at ED recidivism and connection to appropriate outpatient mental healthcare, so we are definitely taking that next step and trying to look at this prospectively.”
Roaten’s advice to other hospitals that are thinking about traveling a similar road regarding suicide screening is to first thoroughly investigate what the existing behavioral health resources are, and how such resources might be leveraged into more active engagement with the hospital. “If you don’t have a psychiatrist on site, is telehealth an option?” she asks.
The screening isn’t the issue so much as putting the appropriate response mechanisms in place.
“If you approach an emergency medicine director with the idea of universal suicide screening with no idea of how to actually respond to patients, you are dead in the water,” Roaten cautions.
Also, put specific people in charge of the program, and make them accessible and responsive to stakeholders, Roaten advises. She notes that this approach worked well at Parkland as she, Johnson, and another co-leader who primarily worked in IT were very invested in personalizing their roles in the suicide screening program.
“If the medical director of trauma services had a problem with the suicide screening or felt that something was not right with the workflow, I wanted him to call me directly to figure it out,” she explains. “The three of us tried from the very beginning to take a lot of responsibility for this so that it wasn’t one more thing that an emergency nurse or physician had to do without a reason and a face attached to it. We tried very hard to be consistent with that, and we continue to do that now.”
Financial Disclosure: Physician Editor Robert Bitterman, Author Dorothy Brooks, Editor Jonathan Springston, Executive Editor Shelly Morrow Mark, and Editorial Group Manager Terrey L. Hatcher report no consultant, stockholder, speaker’s bureau, research, or other financial relationships with companies having ties to this field of study.