There are currently several disturbing phenomena in the field of suicidology:
o Many papers are describing risk assessment and suggesting the need for high-risk patients to be hospitalized.
o Emergency department (ED) staff are complaining about spending much of their time trying to find beds for patients.
o Programs are claiming “crisis intervention” when, in fact, they only provide triage.
Crisis Intervention (CI) is an alternative approach to the suicidal patient that maximizes benefit to the patient while promoting safety. It also minimizes the burden on the system, and increases job satisfaction among clinicians and trainees.
Here, I briefly discuss experience and data from 13 years in the Parkland Hospital Psychiatric Emergency Room in Dallas and show that CI can be used as a very different and more effective approach to the suicidal patient.
The Standard Approach
The usual tools for assessing suicide risk divide patients into high- or low-risk groups, with high sensitivity and low specificity. Most people who commit suicide are in high risk group and a few are from the low risk group. Most patients in the high risk group do not kill themselves. Unfortunately, the tools to assess risk do not really help determine which patient is going to commit suicide unless precautions are taken.
Hospitalization can be life-saving and helpful, but it always causes some damage. There is stigma and expense; the patient and the family learn that they cannot cope; beds are tied up that might be needed for others; problems go unaddressed and therefore often worsen while the patient and family gets a respite, etc.
CI provides a way to spend time with a patient not while assessing suicidal risk but rather in helping lower that risk. The clinician asks: Is this patient suicidal? Is he at high risk? What is the problem—and what can be done about it? What would it take to help this patient become non-suicidal?
What follows is an extreme and simplistic care vignette that serves as an example of the way in which CI works:
A resident in the ED told the attending that her patient probably needed be hospitalized. The patient was a young man who had been brought to the ED by his mother after he skipped his court hearing and began saying that he preferred death to jail and was going to kill himself.
This is not a rare occurrence, and such behavior is usually highly manipulative. However, the resident was convinced that this suicide threat was serious. The attending also interviewed the patient and his mother. Like the resident, the attending concurred that hospitalization was appropriate. The patient still seemed at high risk and there was no alternative secure plan: the CI was not working.
The attending called the district attorney, who advised that jail time was not being requested for the patient and that the missed hearing would be rescheduled. On hearing this news, the patient beamed widely and announced that he was ready to go home.
The lesson? A “suicidal patient” became non-suicidal, and a hospitalization was averted, because we found out what was needed to help him be non-suicidal.
The Parkland Approach
The Parkland program was started in 1981. Many patients came to the program on mental illness warrants, which asserted that the individuals were mentally ill and dangerous. Other patients were transferred from surgery or the medical ED after being “cleared”—perhaps having had their wrist sutured or their stomach pumped. Some patients were voluntary walk-ins. We applied the CI approach to all.
We started with 2 principles:
1. No one gets hospitalized.
2. Every patient has an undiagnosed medical problem that either caused or contributed to the psychiatric symptoms. (We carefully evaluated each patient, and often this was true. We particularly enjoyed finding the medical problem after the patient had been “cleared” by the medical ED.)
Our statistics show the effectiveness of the program in helping patients safely avoid hospitalization. In 1980, before the program began, 83% of warrant patients were hospitalized. We were able to lower that figure to 65% in 1981, and then eventually to 35%. We lacked 1980 data on total hospitalizations, but during our program’s first year, it was 27% of all patients seen. Eventually, that figure dropped to about 12%; it later increased to roughly 18% after the court began to allow us to hospitalize patients with drug or alcohol addiction.
Crisis Intervention: How it works
One of the benefits of CI is that it is easy to learn and easy to teach. A brief outline provides a sense of how it is done; elaboration on the approach can be found elsewhere.1,2 This model uses a family-oriented, action-based approach, with limited goals. After first establishing communication and rapport with a patient, we assess his or her resources and strengths and then try to define a problem that can be worked on. Problem solving produces a plan, and incorporates support and follow up. We do not try to solve our patients’ problems, but we help them to begin to take steps toward a solution.
Working on a problem helps to remoralize, reorganize, and mobilize the patient and his family. The effect may be to decrease stress levels and to get them out of crisis/suicidal mode. The plan should have small steps that can be taken with probable success. Support and safeguards need to be built in.
There are important reasons for using each of these steps and there are specific techniques to help achieve them. CI is only a temporary bandage—but the bandage prevents hemorrhaging long enough to get some movement and often to get patients into ongoing treatment.
In short, the focus of the CI session is on helping the patient with his problem—not on suicide. In approaching each of our patients, we begin by asking ourselves, “What will it take to begin to resolve this crisis so that the person is no longer suicidal?”
We avoid the topic of suicide until the end of the interaction. Suicide risk assessment is practiced only at the end of the interaction. This gives the patient who may have been suicidal the opportunity to regroup and now truthfully answer “no” when asked about intent. If a patient is asked early about suicidal intent and says “yes,” it has a huge effect on the rest of the interaction and makes it harder for the CI to succeed.
If a patient brings up suicide early on, we say “You must be feeling really bad,” and then just ignore it and go on with our CI process. We assume that everyone is at high risk and proceed from there.
We take all suicide risk very seriously; we never use the word “gesture.” If someone speaks about suicide or has done anything even mildly suicidal, then he is thinking about the act, which becomes a potential coping mechanism.
We do finally apply standard risk factor assessment. Most of these data just come out during the interview without our having to ask, however. We also apply other assessing factors:
o Have we collaboratively developed with the patient/family an effective and safe treatment plan, with a support system?
o Does the patent truly buy into the plan?
o Does the patient have an investment in the future?
A Happy Ending?
I was on the fence about discharging a young woman who had been suicidal; we seemed to have worked out a good treatment plan for her. Still, I wasn’t sure. I asked her, “If you go home now, could you come back and see us Monday?’
“No,” she replied.
My heart sank. “Why not?”
“Because we’re taking inventory on Monday.”
We set a different follow-up date and the patient went home.
Sometimes it is useful to ask a patient, “How do you see yourself 2 years from now?” It is a bad sign if he or she can’t come up with an answer.
Finally, if, after CI, a patient’s risk is still high, we will hospitalize.
Over 13 years, and after treating 700 very high-risk patients each month, 9 people committed suicide after being discharged. That seems like a good result. One law suit was filed. (You can be sued for hospitalizing as well as for discharging; liberal hospitalization does not eliminate suicides.) In all 9 of these suicides, we detected a red flag present in each—a hint that by the end of the CI the clinician and patient had not achieved a good rapport.
Objections to adopting CI
“That sounds really interesting, but you don’t understand our situation; that wouldn’t work here.”
I have heard this objection all over the country, and also heard it when I first went to Parkland to begin our program. It really means that “CI wouldn’t work here, unless we made some changes.” That is correct: instituting a CI program usually requires changes in philosophy, training, staffing patterns, and possibly physical layout. The alternative is business as usual—hospitalizing people who don’t really need to be hospitalized, spending time on the phone trying to find beds, and providing patients with triage instead of helping them with emergency care.
1. Puryear DA. Helping People in Crisis. Jossey-Bass, 1979.
2. Puryear DA. Psychological Treatment Principles. In: Hilliard JR, ed. Manual of Clinical Emergency Psychiatry; American Psychiatric Press, 1990.