Detection of suicidal risk among hospitalized veterans: preliminary experience with a suicide prediction scale.

The development of prediction scales which aim to improve, or at least complement the clinician's assessment of suicide risk, has a relatively short history. Research in this area has generally been developed using three different approaches to the problem. The first involves assessment via lethality scales which are behavioral and demographic in origin; the second deals with traditional risk assessment evaluations through the use of psychological tests; and the third has developed out of information gleaned from intensive survey researches of the characteristics of suicidal individuals.

Early work on the development of suicide prediction scales has been confined to a rather small range of sociodemographic and clinical data and has shown that variables such as older age (Kreitman, 1976), male sex, broken marriage or widowhood, social isolation or disorganization, and psychiatric illness, notably depression, are linked with suicide (Brown & Sheran, 1972; Lester, 1972). By reviewing 378 studies, Devries (1968) proposed a three dimensional model for predicting suicidal behavior and saw the need for comprehensive investigations along three major variables: the person's psychological characteristics, his social background and relationships, and his physical determinants. Nevertheless, even some of the more promising studies have continued to rely on police records (Tuckman & Youngman, 1968), on records of hospitalized psychiatric patients (Farberow & MacKinnon, 1974) or on the records of suicide prevention centers (Lettieri, 1974) instead of on detailed investigations of the lives, problems and mental states of those most at risk for suicide. Pallis and colleagues (1982) attempted to incorporate this information in a discriminate function analysis designed to separate a sample of 75 suicides from a sample of 146 attempted suicides on which comprehensive clinical and social data were recorded on an identical schedule. Two sets of discriminating items (with 18 and 6 variables) correctly classified 91% and 83% of the two samples in their respective groups.

There exist significant problems of a methodological nature which make the accurate prediction of suicidal behavior difficult. To date there has been only one study to have used a sample on which comprehensive data were available and which was large enough to have employed advanced statistical procedures for developing a predictive scale (Present Depressive State and Resources Scale) (Motto & Heilbron, 1976). This study also assessed the scale's efficiency by subsequent follow-up. Although the predictive accuracy of this particular scale was discouragingly low, a subsequent scale for a selected high risk group has produced more encouraging results (Motto, 1978). More recent attempts to develop a valid measure of suicidal behavior include the Scale for Suicidal Ideation (SSI) (Beck, Kovacs & Weisman, 1979), the Suicide Probability Scale (SPS) (Cull & Gill, 1982), and the modified Scale for Suicidal Ideation (MSSI) (Miller, Norman, Bishop, et al., 1986).

Importance of Assessing Suicide Lethality

Among Rehabilitation Patients

Epidemiologic studies have related suicide to many emotional states: hostility, despair, shame, dependency, hopelessness, ennui. It is well documented that persons with physical disabilities typically experience such symptomatology during the adjustment process. However, while there exists a limited data base which identifies the prevalence/incidence of active suicide among persons with disabilities, health care professionals generally are of the opinion that suicide among this population by indirect means and through self-neglect is a significant issue which must be addressed.

Committing physiological suicide by ceasing to care for one's body has been examined by several investigators. For example, Sainsbury (1955) found a relationship between the diagnosis of cancer and suicide rates among persons with the disease. Based on her review of multiple studies, Trieschmann (1978) reports that it would appear that between 12% and 48% of deaths among persons with spinal cord injuries may be related to self-destructive behavior. Abrams, Moore and Westevelt (1971) discuss the problems patients on hemodialysis have preserving a satisfactory self-image and maintaining interpersonal relationships. These difficulties are generally related to the issue of dependence. Since dialysis is usually prescribed only as a last resort for people with permanent (and terminal) renal failure, most patients have experienced a period of severe, debilitating illness and have had to accept the status of a dependent invalid. Once they are selected for dialysis, patients are expected to resume independence and self-sufficiency outside the treatment settings, and to accept their total dependence on a life-giving machine and the staff who run it, a difficult balance for most people to find. Acknowledgment that their bodies can no longer function normally and must have mechanical assistance to survive requires major adjustment in patients' self-image and can diminish their self-esteem. Abrams examines the widespread use of denial to cope with these troublesome issues and the incidence of active and passive suicide among hemodialysis patients.

Finally, Rubini (1966) reports on the results of a questionnaire study he conducted which contains data on 3,478 dialysis patients in the United States. He found 29 patients withdrew from programs, 117 died from an inability or a refusal to follow the medical regimen, and 20 successfully committed suicide (another 17 attempted suicide but were not successful). There also were 9 "accidental" deaths from shunts falling apart, 37 "unexplained deaths," and 107 "accidents" (such as shunts falling apart) without death. Therefore, including "unexplained" and "accidental" deaths, 166 patients, or approximately one out of every twenty, ended their lives through active and passive suicidal behavior. Here are some examples of these forms of suicide:

Death through Dietary indiscretion 

Rubini notes: "... We have had one suicide. He was a chap who, as a child, didn't like his teachers or his parents. As an adult he didn't like his bosses; they all told him what to do. After he was dialyzed two years, he didn't like dialysis either, as we tried to tell him what to do also. He picked the Easter weekend. He went on a tremendous feed, almost a Roman orgy. He started off in the San Francisco docks, where he purchased a number of kinds of shell fish and a jug of chianti. He went home, put a suckling pig on the spit; he partook of all these goodies and more. When admonished by his visiting mother, he retorted he knew better than his doctors how to take care of himself He returned to the hospital with severe hyperkalemia, suffered a cardiac arrest and he died."

Withdrawal from Program

An answer from our questionnaire and a newspaper clipping: ". . . two people requested to be taken off dialysis. In both cases, the patients were in extremely poor physical health and were unable to be successfully rehabilitated; one because of eighth nerve damage secondary to Kanomycin which resulted in permanent vertigo and complete deafness and in the other case that of a physician with personality problems prior to dialysis- simply a failure to thrive on dialysis despite the best possible management. Both patients' requests to be withdrawn from dialysis were acceded to by the staff with some degree of relief on our part. "

It's Time to Die, Ailing Man Decides read the headlines of an Associated Press release about a thirty-three-year-old dialysis patient. It describes his "wasting body and his demoralized mind." The article quotes his saying, "I'm taking myself off the machine ... I'm ready to die," and comments that a day later he signed a waiver removing himself from further treatment. He remarked, "When I signed the waiver I knew what kind of symbol that was, like signing your own certificate . . ." (Rubini, page 33).

Some Methodological Considerations

There are specific methodological problems that effect the development of suicidal risk assessment procedures. Careful attention must be paid to the factors of the influence of feedback. Successful intervention of predicted suicidal behavior destroys the possibility of future validation of that behavior. Hospitalization factors must be taken into account, and the nature and composition of the control comparison groups are critical. If the methodological problems associated with valid assessment of suicidal risk can be overcome, then the occurrence of false negatives will be significantly diminished. The reduction of the number of false negatives (i.e., suicidal persons who are erroneously diagnosed as nonsuicidal) is the primary goal of self-destructive risk assessment techniques.

Much of the research into suicide by indirect means among physically compromised persons is methodologically flawed. Follow-up data that would help identify more carefully the characteristics of those who have hastened their own death include duration of disability, severity of disability, preinjury adjustment history, age, marital status and other sociodemographic variables.

Study Overview

A quality assurance investigation was recently undertaken at the Veterans Administration Medical Center in Hampton, Virginia, in order to determine the effectiveness of efforts being made to identify the relative lethality of newly admitted patients. The investigation was specifically focused on examining the usefulness of the Suicide Probability Scale (SPS) (12) in identifying persons with active suicidal ideation. This instrument is routinely administered as part of the Center's Central Test Laboratory's initial test battery. One function of the Laboratory is to provide clinicians with an affective and cognitive profile of recently admitted patients referred for screening. The SPS was included in this screening battery in order to provide a gross estimate of suicidality in patients who had not yet been thoroughly screened or whose presenting problem was such that contact with mental health professionals qualified to make this determination was not indicated.

General Description of the Suicide

Probability Scale

The SPS, originally developed in 1982, is a 36-item, self-report measure that is designed to assess suicide risk in adults and adolescents. Individuals are asked to rate the frequency of their subjective experience and past behaviors using a 4-point Likert scale ranging from "none or a little of the time" to "most or all of the time." The responses are then hand scored to evaluate both general and specific suicide risk along several key dimensions. An overall assessment of suicide risk is reflected in three summary scores: a total weighted score, a normalized T-score, and a Suicide Probability Score which can be adjusted to accommodate different a priori base rates for particular clinical populations. These global scores are easily and quickly calculated.

As an aid to more detailed clinical interpretation, the SPS also provides four clinical subscales: Hopelessness, Suicide Ideation, Negative Self-Evaluation, and Hostility. These subscales evolved from a review of various theories that have been proposed to explain or predict suicide, and have been extensively revised and refined using factor analysis. All subscales are scored in the direction of increasing suicide risk so that a high score on a particular scale indicates a high level of assessed risk within that specific clinical dimension. Clinical interpretation of the SPS is based on individual item responses, the clinical subscales, the summary scores, and the integration of the test data with information from other sources, especially clinical interviews.

Instrument Validation

The psychometric properties of the SPS have been established. Cull and Gill (1982) determined overall test reliability via several methods of internal consistency. Alpha, test-retest, and split-half reliabilities reached .93 for the total scale. The standard error of measurement from these overall reliability estimates was 2.99 (SD = 11.3).

Estimates of the content, criterion-related, and construct validity of the SPS have also been variously demonstrated. Item analysis, subscale and item intercorrelations, as well as the SPS's ability to discriminate between criterion groups, show evidence of instrument validation. Furthermore, comparisons to other scales, i.e., MMPI, Rotter Internal-External Locus of Control Scale (Rotter, 1966), Berger Self-Acceptance Scale (Berger, 1953), provide further evidence for the validity of the SPS.

Limitations of the SPS

Identification of suicide risk is a complex task requiring clinical sensitivity and a thorough knowledge of the clinical and research literature on suicide prediction. The SPS is intended solely as a screening instrument. It was not designed for use in isolation. Instead, other methods such as clinical interviews by trained psychiatric professionals are to be used to supplement, corroborate, and investigate the test results.

The scale also has a number of specific limitations. First, the intent of the scale is not particularly disguised. Thus, the scores are subject to conscious and unconscious distortions by individuals completing the scale. Second, the scale assesses an individual's reported feelings and behaviors at one point in time. Research has shown that suicide is often an impulsive act. Therefore, the scale may not accurately predict temporal changes in suicide risk. Third, the validity of the scale for predicting future completed suicides has not been established. Due to ethical considerations, suicide research often must focus on past events and identified suicide attemptors; however, this leaves unanswered the important question of how well the scale actually predicts future suicidal behavior. Finally, in interpreting test results, the user is cautioned to consider cultural differences in general attitudes toward suicide and toward self-disclosure to psychiatric personnel.

Due to the need to identify patients at risk of attempting suicide within the context of a gross psychological screen, without prior benefit of clinical interview or chart review, the SPS was selected for use to ascertain lethality because of its high degree of face validity. Recognizing that its use as an initial screen has limitations and does not obviate the need for a skilled clinical assessment, including a thorough psychiatric history and mental status examination, following the identification of persons scoring in a "severe" range, the patient's primary physician or caseworker was informed of this test finding for appropriate follow-up action. Thus the original intent of the instrument was preserved, that is as representing a method to supplement skilled clinical judgments, not replace them.

Purpose of investigation The purpose of the present investigation was to determine the accuracy of the SPS in identifying patients with active suicidal ideation and, by so doing, detect those at high risk.

Method

Test results of 1,397 patients who were administered the SPS between July 1986, and October 1987, were examined. One hundred forty-five patients (10%) obtained test scores in the "severe" range. In order to assess the efficacy of the SPS, suicidal frequency data (i.e., number of serious gestures, attempts, and successfully completed suicides) obtained from hospital records during the above stated period were compared to number of similar suicidal incidents obtained for the previous 15-month period during which the SPS was not employed. The clinical records of a random sample of 20% (n=29) of these patients were obtained in order to determine the (a) follow-up action taken by the patient's primary clinician; (b) incidence of false positive findings (i.e., patients who were identified as having active suicidal ideation who upon interview were determined to be subacute); (c) incidence of patients scoring in the "severe" range who were chronically suicidal without a documented history of attempts; and (d) incidence of patients scoring in the "severe" range who had documented histories of previous suicidal gesturing. In addition, the records of the 1,252 patients not scoring in the severe" range were examined to determine the incidence of false negative findings, that is, the number of patients who made unsuccessful or successful suicide attempts and were identified by test results as being subacute.

Subjects

Age

Each of the 29 patients included in the sample was male. They ranged from 24 to 53 in age with a mean age of 36.7 years.

Education

The educational level of subjects ranged from 7 to 14 years. The majority (56%) either had completed high school or received their GED (General Equivalence Diploma).

Marital Status

Fifty-five percent (55%) of patients (n=16) were single, 13% were married (n= 4), and 31% (n= 9) either were divorced or separated at the time of testing.

Pre-hospitalization Living Situation

The living situations of patients included in the sample were examined and revealed that 59% (n=17) lived alone in rented rooms or apartments. Interesting to note is that when queried many of these patients responded that they were "homeless" previous to hospitalization. Twenty-seven percent (27%) (n= 8) indicated that they resided with their families (wife, children and/or close blood relatives). The remaining 14% (n= 4) identified their pre-hospitalization living arrangement as sharing an apartment, room, or home with a roommate.

Employment

Nearly all of the patients studied indicated that they were unemployed. Only 13% (n= 4) were employed immediately prior to seeking admission and/or could return to a position once discharged. Of the 86% (n= 25) who indicated they were unemployed, nearly all had been so for extended periods (16 months or longer), or considered themselves or were considered to be unemployed because of their longitudinal sporadic work history.

Referral Site

Patients included in this study were referred for testing from throughout the Medical Center. However, the great majority (87%) were admitted directly either to the inpatient psychiatry unit or to the Center's Alcohol Rehabilitation Unit. In part this skewing of referral source results from various policies which directs certain units in the facility to routinely refer all new admissions to the Centralized Test Laboratory for initial screening.

Diagnoses

Because many of the patients included in the sample were admitted directly to inpatient psychiatry or the Alcohol Rehabilitation Unit, it is not surprising that nearly all carried psychiatric diagnoses and/or a diagnosis of alcoholism. Many of the patients carried dual diagnoses, the most prevalent being major affective disorders and alcoholism, continuous. Of the 29 patients reviewed, 59% (n=17) carried diagnoses of alcoholism, 68% (n=20) as having a major affective disorder (most usually depression), and 21% (n=6) were identified as having post-traumatic stress syndrome.

Results

Follow-up Action

First, to determine the overall effectiveness of the SPS in identifying high risk patients, a one-group chi square test was employed comparing base line suicidal frequency data with observed numbers of suicidal incidents. The results of this analysis were significant, [X.sup.2](1) = 6.82, p < .01. Additional chi square tests were then conducted to evaluate the discriminative ability of the SPS. These results showed that although the SPS correctly identified serious gestures and attempts at a significant level [[X.sup.2](1) = 5.45, p < .01], its utility to identify completed suicides fell below the desired level of confidence [[X.sup.2](1) = 2.0, p < .10]. Although cells for the latter test were relatively small, they nevertheless suggest a trend.

Subsequent to the identification of patients who had abnormally elevated levels of suicidal ideation as measured by the SPS, the test examiner immediately contacted the patient's primary health care clinician to report this finding. In 90% of cases, a brief interview was conducted by the person contacted in order to verify the test finding and to determine what, if any, additional intervention might be indicated. In the remaining 10% of the cases, no follow-up verification of lethality by interview was undertaken and/or the action taken was not documented in the patient's clinical record. The nature of subsequent actions undertaken, if any, was wholly dependent upon the clinical judgment rendered following this interview assessment. These actions assumed one of three forms: (a) initiation of close observation to continue until the patient's lethality diminished; (b) admission to inpatient psychiatry if not already a patient on that unit; or (c) initiation of a consultation to the Center's Mental Hygiene Clinic. These treatment/follow-up actions were rarely utilized. Less than 1% of patients (n=2) were placed under close observation as the direct result of being identified as having a high level of suicidal ideation. One patient was admitted to inpatient psychiatry; and two patients were referred to the Mental Hygiene Clinic. Documentation of other follow-up actions undertaken as an immediate consequence of this test finding could not be identified in the patient's clinical record.

Incidence of False Negatives

The records of 1,252 patients obtaining less than a "severe" rating on the SPS were examined to determine the incidence of suicide and suicidal attempts among this population. In addition, all Incident Reports which documented untoward events within this timeframe (July 1986 - October 1987) were reviewed. It was determined that 14 patients attempted suicide during the study period. Four patients (.32%) were undetected by testing to be at risk; 4 were accurately identified by the SPS as being at risk of suicide. The remaining 6 patients who successfully or unsuccessfully attempted suicide during this timeframe were not administered the SPS.

Incidence of False Positives

Forty-one percent (41%) of patients identified by the SPS as having active suicidal ideation (n=12) upon interview were determined not to be imminently at risk. However, a significant collateral finding is that all of the patients falling into this category were identified by interview as being chronically suicidal. Thus it was determined by chart review that 100% of patients studied who were identified by the SPS as having potentially lethal levels of suicidal ideation were subsequently identified as being imminently suicidal or manifesting nonlethal chronic suicidal ruminations.

Previous identification by Staff as Being Potentially or Chronically Suicidal

Fifty-two percent (52%) of patients testing in the severe range of suicidal ideation were documented by staff as being acutely or chronically suicidal prior to administration of this test. In each case these patients had been directly admitted to the inpatient psychiatric unit and, previous to testing, had at least one diagnostic intake interview with a mental health professional (i.e., psychiatrist, social worker, psychologist or psychiatric clinical nurse specialist). The remaining 48% of patients not identified prior to testing as having an abnormally high level of suicidal ideation had been directly admitted to units of a non-psychiatric nature, to units primarily staffed by technician-level (nonprofessional) personnel, or to treatment settings whose intake assessment was more administrative than clinical in nature.

Documented History of Previous Suicide Attempts

Fifty-two percent (52%) of patients identified by testing to be at significant risk had documented histories of previous suicidal gesturing and/or serious attempts. Forty-eight (48%) of patients whose SPS lethality index was considered to be abnormally high had no documented history of such behavior. However, of these, 64% were later identified by interview as exhibiting suicidal ideation chronically.

Discussion

On both idiopathic and nomothetic levels, the sociodemographic profile of patients identified by testing to be at risk conforms closely with research findings into the demographic epidemiology of suicide. These variables, including education, employment, social relationships, marital and living status, as well as the presence of underlying psychopathology and alcohol abuse, appear to be valid identifiers of suicidal potential and, as amply documented in the literature, to be critically important factors in assigning lethality.

Results of this investigation into the clinical utility of the SPS indicate that while this suicide detection instrument appears to accurately identify patients who have potentially dangerous levels of suicidal ideation, it does not adequately identify lethality of ideational activity. The primary goal of any such instrument is to predict that a person identified to be "at risk" is highly lethal. While the SPS appears useful in detecting such "at risk" persons, based on this study's findings, it is not useful in determining the imminence of such occurring or in differentiating chronicity of suicidal ideation without intent. Thus, while the SPS efficiently identifies suicidal ideation, its diagnostic value is compromised by its inability to determine the acuteness and lethality of thought and potential for acting out behavior.

Despite this major shortcoming, the SPS can be a valuable adjunctive tool if used in the proper setting. Patients admitted to treatment settings without benefit of a thorough intake interview assessment conducted by a qualified mental health practitioner can quickly be identified as having active suicidal ideation which potentially may lead to their acting out in some fashion. However, in order to verify this finding, an interview is required for a clinical determination of risk to be made.

A second useful feature of the SPS as an initial screen for patients admitted to health care settings without benefit of assessment by a mental health professional is that it alerts otherwise uninformed staff of the patient's suicidal thought processes with a high degree of accuracy.

A third rationale for administering the SPS is that it affords its user a quick and economical method of documenting an empirically derived determination of risk. For example, if a patient who attempts suicide sues a facility for negligence in not identifying his/her lethal condition, that facility's position vis-a-vis liability would appear to be mitigated if an assessment for such via testing empirically demonstrated the level of risk as being subacute at the time of admission. The litigious patient scoring in the severe range would have been interviewed to assess the accuracy of this finding and to identify the chronicity of this ideational activity. Thus, administration of this instrument provides documentation of suicidal ideation and should compel staff to take some follow-up action as indicated through interview assessment.

In summary, while the SPS appears to be a useful tool in accurately identifying suicidal ideation, it does not differentiate degrees of lethality or imminence of acting out behavior. This information typically is obtained by clinicians. As such, its use in adequately staffed psychiatric settings may be redundant and unnecessary. However, its utility as an initial screen in nonmental health care environments may be highly desirable in that it appears to afford the untrained staff person an opportunity to accurately identify patients with high levels of suicidal ideation.

As demonstrated by this and other research, the prediction of suicidal acting-out behavior remains exceptionally difficult to determine, in part due to the chronicity factor, i.e., patients with chronic suicidal ideation who never attempt suicide. Litman and Wold (1976) identified a series of long range, chronic, relatively slow to change characteristics which contribute to high suicidal risk, but this line of investigation has not been aggressively followed up during the intervening decade. Research into suicidal assessment and prevention has yet to develop a valid and reliable method of differentiating chronic suicidal ideation without intent from an imminent suicidal attempt.

Practitioners must examine the rehabilitation process to seek ways to improve the delivery of services to persons with disabilities and develop means of identifying persons at risk of attempting suicide. It would appear that we are not completely successful at teaching people to live with their disability in their own environment so that they have the opportunity to find rewards and satisfactions in their lives. Perhaps the task of the rehabilitation community must be not only to teach ADL and mobility techniques, but to teach a person to find a reason for getting out of bed in the morning. While we cannot give a person a reason to continue to live, perhaps we can teach individuals with disabilities how to go about finding one for themselves.

References

Abrams, H. S., Moore, G. L. & Westevelt, F. B. (1971). Suicidal behavior in chronic dialysis patients. American Journal of psychiatry, 127,1199-1204.

Beck, A., Kovacs, M. & Weissman, A. (1979). Assessment of suicidal intention: The scale for suicidal ideation. Journal of Consulting and Clinical Psychology, 47,343-352.

Berger. E. M. (1953). Relationships among expressed acceptance of self, expressed acceptance of others, and the MMPI. American Psychologist, 8, 320-321.

Brown, T. R., & Sheran, T. J. Suicide prediction: A review. Suicide and Life-Threatening Behavior, 2, 67-98.

Cull, J. G., & Gill, W. S. (1982). Suicide Probability Scale (SPS) Manual. Los Angeles, Western Psychological Services.

Devries, A. G. (1968). Model for the prediction of suicidal behavior. Psychological Reports, 22(3), 1285-1302.

Farberow, N. L., & MacKinnon, D. R. (1974). A suicide prediction schedule for neuropsychiatric hospital patients. Journal of Nervous and Mental Disease, 158, 408-419.

Kreitman, A. (1976). Age and parasuicide  "attempted suicide'). Psychological Medicine, 6, 113-121.

Lester, D. (1972). Why people kill themselves. Springfield, IL: Charles C. Thomas.

Lettieri, D. J. (1974) Suicidal death prediction scales. In A. T. Beck, L. P. Resnek, & D. J. Lettieri (Eds.), The Prediction of Suicide. Bowie. MD: The Charles Press.

Litman, R. E., & Wold, C. 1. (1976). Beyond crisis intervention. In E. S. Shneidman (Ed.), Suicidology: Contemporary developments. New York: Grune & Stratton.

Miller, I. W., Norman, W. H., & Bishop, S. (1986). The modified scale for suicidal ideation: Reliability and validity. Journal of Consulting and Clinical Psychology, 54(5), 724-725.

Motto, J. A. (1978). The psychopathology of direct self-destruction: In K. Achte, & J. Lonnquist (Eds.). Psychopathology of direct and indirect self-destruction. Psychiatria Fennica Supplementum, 47-57.

Motto, J. A., & Heilbron, D. J. (1976). Development and validation of scales for estimation of suicide risk. In E. S. Shneidman (Ed.). Suicidology: Contemporary developments. New York: Grune & Stratton.

Pallis, D. J., Barraclough, B. M., & Levey, A. B. (1982). Estimating suicide risk among attempted suicides: Development of new clinical scales. British Journal of Psychiatry,]41, 37-44.

Rotter, J. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80(t, Whole No. 609).

Rubini, M. I. (1966). Proceedings, the conference on dialysis as a practical workshop. New York: National Dialysis Committee.

Sainsbury, P. (1955). Suicide in London. London: Chapman & Hall.

Trieschmann, R. (1978). The psychological, social and vocational adjustment in spinal cord injury: A strategy for future research. Washington, DC: Rehabilitation Services Administration.

Tuckman, J., & Youngman, W. F. (1968). A scale for assessing suicide risk of attempted suicides. Journal of Clinical Psychology,24,17-19.