A search of the literature was performed with MEDLINE, HlthStar, Curr Cont, Cinahl, and PsychInfo databases using the keywords "suicide," "no-suicide contracts," "no-harm contracts," and "suicide prevention/control" with limits to primary care, psychology, and the English language. Thirty-two articles were identified. Of 11 articles that directly addressed the use of no-suicide contracts, only 2 were empirically based, and neither directly addressed the effectiveness of this intervention to prevent suicide. Rather, these 2 studies only evaluated the frequency of no-harm contracts by nurses and primary care physicians.
The role of the primary care physician in the prevention of suicide is unclear. Although it is considered mandatory to inquire about suicidal ideation in psychiatric consultations, it is rarely part of a routine medical assessment.1 Past studies indicate that as many as 2% to 3% of patients seeking services in primary care clinics report significant suicidal ideation.2 However, in only 1 study3 is a primary care intervention demonstrative of a reduction in suicide rate. In this study, primary care physicians on the island of Gotland, Sweden, provided suicidal patients with an educational intervention resulting in a 50% reduction of suicide (P<.05). While teaching about suicidology was a part of this intervention, the specific role of the no-harm contract was not explored.
Although the no-harm or no-suicide contract is widely advocated as one approach in the management of suicidal patients, few recent sources are available to address the efficacy of using such interventions to prevent suicide. The scant information that does exist on the use of no-suicide contracts focuses on suicide prediction, medicolegal aspects of treating suicidal patients, the therapeutic alliance between the patient and the clinician, and countertransference issues with suicidal patients,4 none of which are considered recent. Miller et al5 found both by literature review and by their survey of clinicians that the use of no-harm contracting seems to be based more on the belief of effectiveness than on objective data or formal training. Their survey of 112 psychiatrists and psychologists revealed that most had never received formal training in the implementation of no-suicide contracts for patients at risk of suicide.4
The recognition of acute and chronic suicidal vulnerability during office visits is crucial to the evaluation and treatment of patients at risk of suicide. In a study of 134 patients who committed suicide, more than half consulted a physician within a month or less of their death.6 A high proportion of these patients not only had a psychiatric disorder but also had a medical disorder. Prevalence studies of medical disorders in patients who commit suicide have shown that approximately one third of these patients are physically sick. Moreover, in certain populations with chronic disease (ie, dialysis patients), the risk of suicide may be up to 100 times greater than the risk in the general population (P<.001).7
Recognizing the risk of suicide requires attention to subtle clues from the history as well as the mental status examination. A history of depression, alcoholism, drug abuse, schizophrenia, persistent pain, or chronic disease may signal an increased risk of suicide.8 Patterns of consultation, diagnosis, and treatment of mental illness by general practitioners were analyzed in a case-control study of patients who committed suicide.9 Independent risk factors for suicide in the study population included a history of attempted suicide, untreated serious mental illness (odds ratio >20), marriage in the previous 12 months, separation or divorce, alcohol abuse, frequent consultations with a primary care physician, and previous mental illness. However, findings revealed that general reporting of life events in patients' medical records was inconsistent, which may have led to underidentification of suicidal risk factors.
As with history taking, data from a mental status examination provides additional observations of suicidal risk.10 Mental content such as suicidal preoccupationespecially planning for suicideas well as fantasies concerning death add to general risk factors. Cognitive risk factors include incapacity for reality testing and self-object differentiation, delusions, hallucinations, and the presence of formal thought disorders. Patient cues to suicide risk are observed in the presence and intensity of suicide-inviting affects such as self-hate, aloneness, murderous hate, fatigue, despair (hopelessness), and depersonalization. Intellectual functioning is also impaired by disorientation and delirium. In addition, the patient's attitude toward the examiner (health-accepting vs help-rejecting) provides clues to suicidal ideation.
Psychiatric consultation is indicated in patients who clearly exhibit signs of high risk for self-injury, such as suicide intent, an overt plan for death, or a suicidal gesture.11 In addition, hospitalization is usually recommended for socially isolated patients who present with overt suicidal ideation, which may be complicated by encephalopathy, substance abuse, or injurious self-harm. In patients who do not manifest high-risk signs, the most urgent decision faced by the primary care physician is whether hospitalization is required. When the need for inpatient care is uncertain, the no-suicide contract is seen as a helpful adjunct to treatment.5 Because suicidal patients are often reacting to crisis, most respond well to supportive crisis intervention counseling, which may include a no-harm contract.
Use of no-harm contracts has been advocated by many authors. Stanford et al12 encourage the use of no-harm contracts as an assessment tool to uncover the nature and severity of a patient's suicidality, to identify specific troubling issues that are precipitating suicidal thoughts, and to evaluate the patient's competency to enter such a contract. They add that the no-suicide contract affords an opportunity to initiate a therapeutic alliance, to establish the limits of a psychotherapeutic framework, and to reduce anxiety in both patient and clinician. However, as these authors noted, the contract is not legally binding and does not protect against claims of malpractice. Although Egan4 states that a contract is justified as a lower level of intervention, he discouraged this form of risk management when patients are inebriated, psychotic, or unable to enter into legal contracts.
Johnson and Maile13 suggested the use of a document that includes the patient's statements that, until the next scheduled appointment, he or she agrees to not harm himself or herself in any way or not attempt suicide. They also suggested including statements in which the patient agrees that the contract is worth signing and upholding, agrees to get rid of any harmful objects, and agrees to contact an identified social support system if he or she feels unable to maintain the contract. The contract is then signed by both the clinician and patient, and in some cases a witness to add further accountability to the agreement.
Despite the wide acceptance of no-suicide contracts for at-risk patients, concerns have been raised that in some cases they may be harmful. Simon14 stated that although asking a patient if he or she is suicidal and obtaining a written or oral contract against suicide can be useful, these actions by themselves are insufficient to prevent suicide. He emphasized that a no-harm contract should never be used as a replacement for a formal suicide assessment. Egan4 also acknowledged that there may be undue reliance on a patient's apparent willingness to sign a contract for safety, particularly if there is little opportunity to establish a therapeutic relationship based on the rapport and trust necessary for such a contract. Although Stanford et al12 concluded that no-suicide contracts can provide diagnostic information and therapeutic advantage, the contract can also short-circuit comprehensive suicidal assessment and disposition decisions.
No empirical evidence supports the effectiveness of no-harm contracts in preventing suicide. Such contracts should be used only as an adjunct to full assessment with patients who are at high risk of self-harm. However, despite this lack of evidence, use of such contracts may serve several useful purposes, including fostering a therapeutic alliance between the clinician and patient and assisting in suicide assessment. Because primary care physicians have frequent contact with patients who may be at risk for suicide, the physician's role in identifying these individuals and preventing self-harm is crucial.