November of the Soul

by George Howe Colt

  Psychologists at the University of Washington decided to focus on why people don’t kill themselves; their Reasons for Living scale scores forty-eight factors connected to a patient’s coping skills, fear of social disapproval, moral objections to the act, and concern for family. And while most scales are intended for those who may be thinking about suicide, the Risk-Rescue Rating, devised by Boston-area therapists Avery Weisman and William Worden, computes the lethality of an actual attempt by assigning points to five risk factors (which include actual damage inflicted, and method—pills get one point, jumping and shooting, three) and five rescue factors (which revolve around the chances of being found in time to survive). The rating is tabulated by dividing risk score by risk plus rescue scores and multiplying by one hundred. Thus, it is demonstrated that a thirty-eight-year-old unmarried waitress who ingested sedatives, then went to a movie theater, where she was found in a coma and subsequently died, received an eighty-three, the highest possible score.

  While these scales provide useful checklists for clinicians, playing “the numbers game,” says psychiatrist Douglas Jacobs, can be dangerous. “You have to be careful. On a percentage basis, young people are less likely to kill themselves, but if one of them does, for that person it’s one hundred percent.” Jacobs points out that the risk factors used in most scales may change. The dramatic increase in adolescent suicide has subsided; the rate for the elderly has gone down. (It has been suggested that different scales be developed for specific populations: for young females, for Native Americans, for middle-aged alcoholics, for patients in psychiatric hospitals.) Suicidal feelings fluctuate over time, and a rating that may be valid one day may be invalid the next; like the weather, points out psychiatrist Robert Simon, suicide risk must continuously be monitored. And no matter how specific the scale, no matter how frequently applied, there will be exceptions to the rule. University of Alabama researchers, declaring that most assessment tests are “too complex or cumbersome for practical and routine use,” devised the SAD PERSONS scale, an acronym for ten risk factors. Scoring one point for each, the patient’s probability of making an attempt is rated from one to ten. Suggested treatment is based on score, ranging from “send home with follow-up” (0–2) to “hospitalize or commit” (7–10). The researchers concede that some people may slip through the statistical net. “For example, a fourteen-year-old girl who attempted to hang herself ‘because the devil came and told me to’ might score only three points on the scale.”

  Some believe the patient may be the best judge. In their 1973 paper “Patient Monitoring of Suicidal Risk,” a group of California therapists suggested that clinicians simply ask the depressed patient how long and under what circumstances he will stay alive. The patient is then asked to make a pledge: “No matter what happens, I will not kill myself accidentally or on purpose at any time.” The authors wrote, “If the patient reports a feeling of confidence in this statement, with no direct or indirect qualifications and with no incongruous voice tones or body motions, the evaluator may dismiss suicide as a management problem.” Claiming that in five years none of the six hundred patients who had made “no suicide” decisions had broken their pledge, the authors declared that their technique was “suitable for use by inexperienced nonprofessionals as well as by experienced professionals.”

  “Patient Monitoring of Suicidal Risk” may smack of a certain inmates-running-the-asylum naïveté, but over the decades it has evolved into the widespread and controversial practice of “no-suicide contracts,” in which a patient promises, verbally or in writing, not to kill himself, and to keep his therapist informed of any self-destructive impulses. Although studies have shown that under the temporal and financial pressures of managed care, clinicians increasingly rely on no-suicide contracts as a form of risk management, no studies have shown whether they actually work. A survey at Harvard Medical School found that 86 percent of psychiatrists and 71 percent of psychologists worked in settings where no-suicide contracts were regularly invoked, yet fewer than 40 percent had been trained in their use. Many clinicians find them helpful; others point out that signing a contract may make the therapist feel safer but is no guarantee the patient won’t kill himself. “Indeed, the use of such contracts flies in the face of clinical common sense and may in fact increase danger by providing psychiatrists with a false sense of security, thus decreasing their clinical vigilance,” wrote Marcia Goin, president of the American Psychiatric Association, in 2003. “. . . We can make contracts with builders, insurers, and car dealers, but not with patients.” Arguing that no-suicide pacts cannot substitute for comprehensive risk assessment, psychiatrist Robert Simon concludes, “The contract against self-harm is only as good as the soundness of the therapeutic alliance.”

  What assessment scales and no-suicide contracts prove most convincingly, it seems, is that our ability to predict suicide is negligible. “Although we may reconstruct causal chains and motives after the fact, we do not possess the tools to predict particular suicides before the fact,” concluded psychiatrist Alex Pokorny after his scale for suicide risk proved unsuccessful in a prospective study of forty-eight hundred inpatients at a Texas VA hospital. Says Robert Litman, “Even for someone in a high-risk category, the chances of suicide within a year are much less than the chance that he will not have committed suicide within that time. In twenty-five years, I can remember perhaps three cases where I felt the chance of a certain person committing suicide within the next year was more than ten percent.”

  Even the most effective scales are intended to be a supplement to clinical judgment, not a substitute for it. “You can know all the statistics and scales and still not have any ability to assess a patient,” says one psychiatrist. But if the scales haven’t proved their worth, neither has clinical intuition. In one study a computer was shown to be more accurate than experienced clinicians in predicting suicide attempters. Adding insult to injury, half the patients preferred the computer to the therapist as interviewer.

  For some of these reasons, a number of mental health professionals scoff at the “clues” approach to suicide prevention. “We’ve reached the point of no return in defining vulnerable populations,” says psychiatrist Herbert Hendin. “It amounts to looking for the proverbial needle in a haystack.” Hendin knocks Shneidman’s proposed educational blitz. “I don’t follow the logic of putting millions into educating the lay public in something that psychiatrists haven’t proven they can identify. It makes more sense to do something for the people you do find. A lot of seriously suicidal people present themselves to us in ways nobody can miss—they jump from five-story buildings—and nobody does anything for them.” The highest predictor of suicide risk is a previous attempt; between 25 and 40 percent of completed suicides have tried before, and 2 percent of those who attempt will complete within one year, 10 percent within ten years. Yet most attempters are returned to the community after being stitched up or pumped out, without provision for further treatment. (In one study, half of all adolescents brought to emergency rooms after a suicide attempt did not receive follow-up care—surprising until one learns that 70 percent of emergency-department physician training programs offer no instruction in the management of psychiatric problems.) “If you could identify twenty percent of the seriously suicidal from those who make attempts and cure ten percent,” declares Hendin, “you could literally change the suicide rate.”

  Can clinicians “cure” suicidal people? The question is rarely asked. The bottom line at most prevention centers and in most prevention literature is to get the suicidal person to professional help. Although getting the person to that help can be difficult—reluctance by the person or his family to admit there is a problem, the stigma of being in treatment, and the high cost of quality care are a few of the obstacles—it is often assumed that once we do, the problem is solved. Yet professional help is no guarantee against suicide. Clinicians often point out with alarm that slightly more than half of people who kill themselves have never seen a mental health profes
sional; the flip side—that nearly half of people who kill themselves have seen a mental health professional—should be considered nearly as disturbing. Indeed, people who have made attempts and entered treatment have the highest suicide rate of any patient group. Yet the focus of suicide prevention has been on assessment and prediction of suicide risk; treatment has largely been ignored.

  How do clinicians treat suicidal people? A therapist’s first task, of course, is to address the crisis and decrease the risk of suicide, just as counselors are trained to do on the SPC phone lines. “The immediate goal of a therapist, counselor, or anyone else dealing with highly suicidal people should be to reduce the pain in every way possible,” writes Shneidman. “Help them by intervening with whoever or whatever is causing their distress—lovers, parents, college deans, employers, or social service agencies. I have found that if you reduce these pressures and lower the level of suffering, even just a little, suicidal people will choose to live.” In his book Definition of Suicide, Shneidman described a counseling session with a distraught college student. Pregnant, single, profoundly religious, and overwhelmed by shame and guilt, the girl had decided to kill herself. Shneidman’s initial task was to help her to realize that alternatives existed.

  I did several things. For one, I took out a single sheet of paper and began to “widen her blinders.” Our conversation went something on these general lines: “Now, let’s see: You could have an abortion here locally.” (“I couldn’t do that.”) . . . “You could go away and have an abortion.” (“I couldn’t do that.”) “You could bring the baby to term and keep the baby.” (“I couldn’t do that.”) “You could have the baby and adopt it out.” (“I couldn’t do that.”) “We could get in touch with the young man involved.” (“I couldn’t do that.”) “We could involve the help of your parents.” (“I couldn’t do that.”) “You can always commit suicide, but there’s obviously no need to do that today.” (No response.) “Now, let’s look at this list and rank them in order of your preference, keeping in mind that none of them is perfect.”

  The very making of this list, my non-hortatory and non-judgmental approach, had already had a calming influence on her. Within a few minutes her lethality had begun to de-escalate. She actually ranked the list, commenting negatively on each item. What was of critical importance was that suicide was now no longer first or second. We were then simply “haggling” about life—a perfectly viable solution.

  Once the immediate danger has passed, how does a therapist treat a suicidal patient? Ask almost any therapist and he or she is likely to answer, “Suicide is a symptom, not a diagnosis.” (Although “suicidality” is included as one of nine symptoms of a depressive episode in the Diagnostic and Statistical Manual of Mental Disorders, suicide itself is not listed as an illness.) Nor is suicide dependent on a specific disorder. “Suicidal behaviors may be generated in the presence of practically any diagnostic entity, and at times in the absence of pathological states,” says psychiatrist Jerome Motto. Because a clinician can’t treat suicide as directly as he might treat, for example, strep throat—there is no antibiotic for suicide—he must treat the patient’s closest diagnosable ailment, which is, more often than not, some form of depression. Many clinicians believe that if they successfully do so, they’ve treated the suicidal patient, as if suicidality were simply a nasty side effect of the underlying illness. Yet some suicidal patients, albeit a minority, have no diagnosable underlying illness, and patients often kill themselves shortly after coming out of a depression—or long after a depression has lifted. “Suicide proneness is primarily a psychodynamic matter; the formal elements of mental illness only secondarily intensify it, release it, or immobilize it,” psychiatrists Dan Buie and John Maltsberger have written. “The urge to suicide is largely independent of the observable mental state, and it can be intense despite the clearing of symptoms of mental illness.”

  How is the “underlying illness” of the suicidal patient treated? Although there is no pill for suicide, there are scores for depression and other psychiatric conditions. Over the past several decades, growing emphasis on the role of mental illness in suicide has combined with extraordinary advances in the development of psychotropic medications to effect a sea change in the treatment of suicidal people. Twenty or thirty years ago, depressed and possibly self-destructive people were likely to be treated with psychotherapy, supported, where indicated, by medication. By the new millennium, drugs were the treatment of choice—in most cases the only treatment—for depression as well as nearly every other psychiatric condition, with psychotherapy occasionally playing a supportive role. As journalist Daphne Merkin wrote in the New York Times, “In our age the triumph of the pharmaceutic has overtaken the triumph of the therapeutic; for all but a select few the cost-effective discussion of dosages has replaced the expensive discussions of dreams.”

  What is now referred to as the “drug revolution” had its roots in the late 1940s, when French naval surgeon Henri Laborit was looking for something to calm his patients before administering anesthesia. He found that chlorpromazine, a sedating antihistamine, induced a “euphoric quietude.” He recommended the drug to his psychiatrist colleagues, who tested it on a variety of mentally ill patients and found it effective in the treatment of manic depression and schizophrenia. In 1954 it was introduced to the United States. Thorazine, the brand name by which chlorpromazine would be known, achieved remarkable results. Patients who had been unruly and assaultive were suddenly docile. In some hospitals the use of straitjackets, wet packs, and seclusion was virtually abandoned. Many patients were able to return to the community. Described—admiringly—by some as a “chemical lobotomy,” Thorazine became the drug of choice in American mental hospitals. Although many therapists believed Thorazine would be a panacea for suicidal patients, the drug offered control, not cure. In 1954, the year tranquilizers were introduced at Metropolitan State Hospital in Norwalk, California, the inpatient suicide rate more than doubled. Investigators suspected that the staff might have relaxed their vigilance because of the drugs’ efficacy in controlling symptoms. (Indeed, Thorazine rendered patients “immobile” and “waxlike,” as described by the lead investigator for the pharmaceutical company that manufactured it. He meant this as high praise.)

  Over the following decades, Thorazine was followed by a succession of seemingly ever more miraculous medications: lithium, a naturally occurring salt, proved effective in moderating the roller-coaster mood swings of manic depression; clozapine helped quiet the nattering voices of schizophrenia; monoamine oxidase inhibitors and their successors, the tricyclic antidepressants, such as Tofranil, had a leavening effect on severe depression. Perhaps the biggest change came in 1988, when fluoxetine (better known by its brand name Prozac) was introduced in the United States, the first in a new class of antidepressants known as selective serotonin reuptake inhibitors (SSRIs), which, as their name suggests, act by blocking the removal of serotonin at the synapses, thus increasing the availability of serotonin in the brain while leaving other neurotransmitter levels unaffected. Prozac and its cousins—Zoloft, Paxil, and so on—were so successful in providing relief to depressed men and women, as well as in giving a lift to millions more mildly unhappy people, that by 2000, one in ten Americans was taking antidepressants. An increasing number of citizens of Prozac Nation, as it has been dubbed by the writer Elizabeth Wurtzel, are young; in 2002, nearly 11 million children and teenagers were prescribed antidepressants, accounting for 7 percent of all antidepressant prescriptions, more than triple the number ten years earlier.

  Despite their success in relieving the symptoms of psychiatric illness in millions of people (and, in the process, doubtless keeping many of them from killing themselves), these medications have had a more complicated relationship to suicide than might be expected. Among all the psychopharmacological treatments, the only one clinically proven to reduce suicide risk is lithium. In 2001, Harvard Medical School psychiatrists Leonardo Tondo and Ross Baldessarini analyzed thirty-three stu
dies conducted over the previous thirty years and found that patients with major depression or bipolar disorder who hadn’t taken lithium were thirteen times more likely to have completed or attempted suicide than those who had taken it. (The suicide rate of those who had taken lithium was, nevertheless, nearly three times higher than that of the general population.) In a German study, 378 psychiatric inpatients, half with bipolar disorder, half with major depression, were randomly assigned to lithium, an anticonvulsant, or an antidepressant on their release from the hospital. Over the following two and a half years, four killed themselves and five made serious attempts; none were in the group taking lithium. The study’s author suggested that lithium may have a direct effect on suicidal behavior, independent from its effect on depression, perhaps by reducing aggression and impulsivity. Unfortunately, not all patients respond well to lithium, either because their systems won’t tolerate it, or because of its possible side effects—blurry thinking, weight gain, tremors, lethargy. Others may balk because of the hassle: lithium treatment requires monitoring of blood levels every few months. Or, like Brian Hart, they may consider lithium stigmatizing and stop taking it, either periodically or permanently. (Nearly one-half of all patients with bipolar disorder fail to adhere to their medication regimens at some point.) If there is anything more dangerous than a bipolar patient not being on lithium, it is a bipolar patient going off lithium; Tondo and Baldessarini found that suicidal acts rose sixteenfold in the first year after discontinuing treatment.