Of the 267 respondents, 152 (57%) used no-suicide contracts. Within this group, 62 (41%) psychiatrists reported that they had patients who had committed suicide or made a serious attempt after entering into a no-suicide contract. There was an inverse relationship between the use of no-suicide contracts and years out of residency training. The reason more experienced psychiatrists are less likely to use no-suicide contracts are not entirely clear, but it may in part relate to the seasoned judgement that if a therapeutic relationship is already present, then a formal contract has no moral claim on a patient. It may also be that psychiatrists recently out of training are more accustomed to using formal assessment instruments and legalistic forms than are older practi- tioners. Although experienced psychiatrists used no-suicide contracts less frequently, a higher percentage of ex- perienced psychiatrists reported serious or completed suicide attempts in their practices. It is assumed that this finding reflects the greater number of years at risk for having suicidal patients in one's practice.
There were 2 goals to this study. The first was to inquire whether there was a standard of care among psychiatric practitioners about the use of no-suicide contracts. The answer was clearly no; slightly more than half of the respondents used it. The second goal was to obtain a rough measure of the effectiveness of no-suicide contracts. The limitations of the study are many, including lack of a control group, self-reported data, and sample bias in who responds to a postcard survey. Nevertheless, it is clear that no-suicide contracts are of little assistance, on average, in preventing serious suicidal behavior; although, there undoubtedly are cases in which a no-suicide contract as part of a richer therapeutic relationship between patient and clinician is effective.
There are situations in which rote adherence to a no-suicide contract policy interferes with effective treatment.9 The policy at a rural mental health program that operated a crisis hotline was that with suicide calls, the telephone counselor could not disengage until a no- suicide contract or promise was extracted from the caller. A patient with borderline personality disorder (BPD) came to appreciate the opportunity this policy afforded and began telephoning the crisis line 3 or 4 evenings per week, keeping the counselor on the line for an hour or two before reluctantly agreeing to a no-suicide pledge. This patient had never posed a serious threat for suicidal behavior, but the policy, until modified, did not allow for common-sense contingencies, such as limiting the telephone calls to 10 minutes per evening, or scheduling the patient to call in each evening for a brief status report.
If we consider the diagnostic spectrum, it appears that no-suicide contracts, in and of themselves, are untrustworthy in patients with schizophrenia whose suicidality may be psychotically driven, with alcoholic patients who are impulse-driven and cognitively and conatively incompetent when intoxicated, and with patients who have BPD for whom requests for a promise are overladen with too many control, manipulation, and relationship issues. It is also likely that a no-suicide contract would have different meaning and usefulness to adolescent, middle-aged, and elderly patients, but this too is an unexplored arena.10
Is it time to dispense with the no-suicide contract?Recently, writers on this topic have concluded that the no-suicide contract is a construct that has outlived its usefulness. It sets up the clinician for a nasty court battle in cases of completed suicide and suicide attempts that result in serious injury to the patient and provides little or no advantage overall. In its place, most investigators advocate careful suicide assessment and documentation and a suicide prevention plan that is one piece of a larger therapeutic contract, however construed. The decision to have a formal contract versus an informal therapeutic understanding appears to have as much to do with the clinician's personality and theoretical approach as it does to an evidence-based decision.
Rudd and colleagues,11 working from a cognitive-behavioral model, have written an interesting piece criticizing the conceptual basis and practical use of no-suicide contracts. In its place, they advocate an individualized commitment-to-treatment statement that is drafted and handwritten by both patient and therapist. This statement encompasses many areas of commitment to a broadly conceived therapeutic relationship, including such housekeeping topics as attendance at sessions, setting goals, completing homework assignments, and voicing opinions honestly. Included in the commitment-to-treatment statement is a crisis response plan that again is tailored to a specific patient and, by agreement, is time-limited. If advisable, family members' roles are incorporated into the steps of the crisis plan for when self-management fails.
The notion of a broad commitment-to-treatment statement that includes a crisis response component makes much sense; however, it cannot be applied to all patients in an outpatient practice, nor would it fit in with different styles of psychiatric and psychological care. Although its virtues cannot be questioned, it may be too rational and too formal and not fit some psychiatrists' style of interaction with patients. Perhaps it is best suited for a cognitive-behavioral psychotherapy practice.
An alternative approach is to openly discuss suicide risk and available responses to suicidality, including many of the elements mentioned by Rudd and colleagues,11 and to convey to patients concern about their safety and survival without extracting a no-harm pledge from a patient. As stressed by most articles in the psychiatric literature on suicide prevention, it is imperative to document assessment of risk and the response options discussed with the patient. Lee and Bartlett12 present an extended discussion of suicide prevention without the use of a no-suicide contract that includes thorough assessment, creation of an appropriate management plan, involvement of family and significant others, consultation with other professionals, and implementation of a plan that assumes professional responsibility.12
It is hazardous to develop absolute dictums in psychiatric practice beyond the general understandings against exploitation of patients. Assuming that the therapist's intentions toward a patient are professionally benevolent, questions still remain about what is best for a particular patient in specific circumstances. In the absence of evidence-based practice guidelines—and most daily psychotherapeutic interactions fall into this category—it is impossible to say that a no-suicide contract should never be used. This modesty of imperatives keeps the notion of no-suicide contracts as one of the viable options that the clinician may consider in working with patients at risk for self-harm. There is no good statistical evidence that no-suicide contracts do more benefit than harm, but the possibility that such negotiations may be helpful for some clinicians in some situations precludes any absolute recommendation against their use.
Recent psychiatric literature on the topic advises against routine use of no-suicide contracts as a suicide assessment or prevention tool or as a method of providing some legal protection in the advent of suicide completion and serious attempts.2,4 This is especially the case when a no-suicide contract is expediently substituted for a careful assessment of risk and a thoughtful and collaborative development of a suicide prevention plan.
