Awhite man who appeared to be over 50 years of age was brought by emergency medical services to the emergency room of a New York City medical center after he was found in his skid row hotel room, lethargic, with a suicide note and no other identifying information. He initially stated that he had taken 40 "sleeping pills" and later also reported taking warfarin and acetaminophen with codeine. He spoke English with no apparent accent and refused to divulge his name, age, or any other personal identifying information. He admitted to planning his suicide attempt over the prior 2 weeks. He did not reveal the reasons but stated that he felt "useless." His wallet was empty; the police had brought his suicide note, which was not addressed to anyone. It began, "I am not alocholic [sic], a drug addict or have a criminal record, and have never had my fingerprints taken," and continued for 1.5 pages, carefully written on lined paper torn from a notebook. After apologizing to the unknown person whose name he had borrowed from a telephone directory when checking into the hotel, it continued:

The patient admitted to the emergency room doctors that he had made a suicide attempt a few years earlier and said, "back then I didn't have the right medications." He stated that he had been depressed for years and had seen a psychiatrist but not in the past year. He admitted to ongoing suicidal ideation.

He was hemodynamically stable in the emergency room but had elevated coagulation times, hematuria, and guaiac-positive stools. Laboratory findings indicated elevated hemoglobin and hematocrit, which decreased with hydration, and a persistently elevated mean corpuscular volume of 104 µm³/cell. Results of a urine drug screen were positive for opiates. Serum acetaminophen and salicylate levels were not elevated; ammonia level was slightly elevated (40 µmol/liter). He was noted to be lethargic but never in cardiac or respiratory distress. He underwent nasogastric lavage and was treated with activated charcoal. Naloxone was administered, with no response.

The usual procedure in psychiatry, as in the rest of medicine, is first to collect data, then to establish a diagnosis and formulation, and finally to construct a treatment plan. In emergencies this process is often modified, important data may not be available and the diagnosis unclear, but there is no time to wait. We do not know who the patient is, and he will not (or cannot) tell us. The emergency room staff should have questioned the police, had his hotel room and medicine cabinet searched, and checked the labels in his clothes for further identification, but they did not. We do know that the patient appears to be suicidal and to have planned and acted on this; he is suffering from potentially dangerous but treatable medical complications of those acts, and there are no friends or relatives to take care of him. Hospital admission for medical stabilization, protection from self-destructive acts, and further evaluation is indicated.

Of course, that does not mean that we do not start thinking diagnostically and in terms of a case formulation. The serious suicide attempt in a middle-aged man makes us first think of mood disorders, schizophrenia, and substance abuse. The coherence and organization of the note, as well as its content, tilt us toward a diagnosis of mood disorder rather than schizophrenia. The direct denial of substance use is relevant but does not resolve the issue. Finally, there are suggestions of narcissistic, possibly grandiose themes and characterologic (axis II) pathology in the content of the note and in his refusal to divulge his identity—the latter unusual in pure mood disorders.

We are also aware, in the 1990s, of another problem. Hospitalization and medical treatment cost money, and from the moment of admission hospital administrators try to determine who should pay the bill—private insurance, Medicaid, Medicare, or the patient and family. An anonymous, apparently destitute patient who does not want treatment and refuses to identify himself is an administrator's nightmare. We do not know his residence or citizenship, and we cannot even threaten him with discharge if he does not cooperate, since he wants to kill himself anyway. However, these are not the primary considerations in an emergency, when the system still allows doctors to do what should be done.

He was admitted to the medical intensive care unit and immediately given fresh frozen plasma and vitamin K to correct his coagulopathy. Here he reported a history of three deep vein thromboses, the last one occurring 6 years earlier. He had been taking warfarin since that time. He denied alcohol or drug abuse and reported that he had smoked one pack of cigarettes per day for many years. He related that he was heterosexual and volunteered that he had tested HIV negative a few years before. He stated that he had worked as a health policy analyst but had been unemployed for a long time. He refused to divulge any other personal information. A nurse noted that he called himself a "real bastard." He was seen by the psychiatry consultation-liaison team, to whom he continued to express suicidal ideation. They assessed him to be at high risk of suicide and suggested his transfer to a locked inpatient psychiatry unit when he was medically stable.

We know a little more, but not much, and have some further questions as well. Why did he take the HIV test? He alleges that he does not use drugs and does not have a history of homosexuality. He claims to have worked in the health care industry; is his familiarity with the system influencing his behavior? However, the overriding issue is that the staff believes that his suicidal preoccupations have persisted. We do not have an official diagnostic category for "real bastard," but his self-description does correlate with some of our axis II designations.

Three days after his initial presentation in the emergency room, the patient was transferred to a locked psychiatric unit as a voluntary patient. He still refused to reveal any personal or demographic details about his life and expressed concern that if he were to do so, his friends would be able to find him, and that would burden them. He related a history including hopelessness, helplessness, insomnia, decreased appetite with weight loss, and low self-esteem over the past 4.5 years, after being laid off from a job as a health policy statistician. He had "given up" and had come to New York with a plan to kill himself anonymously. His appearance was quite disheveled, and he did not bathe or change his clothes daily on the unit. His speech was organized and coherent, his mood was depressed, and he denied psychotic symptoms. The treatment team's initial DSM-IV diagnosis was as follows: axis I, major depressive disorder, severe without psychotic features; axis II, deferred ("deferred" indicates that they were not ready to designate a specific category, although they believed that an axis II disorder was present).

The treating psychiatrist discussed antidepressant medication with the patient, who initially believed that it would "just numb" him rather than help him but finally agreed to a trial. The patient was started on oral regimens of paroxetine, 20 mg/day, and trazodone, 50 mg at night.

He has become a psychiatric inpatient. He now describes a full-blown depressive syndrome of 4.5 years' duration, but that duration is unusual and raises more questions about the reliability of his history and his possible attempt to distort it in order to manipulate the hospital. He also continues to refuse to identify himself—highly unusual behavior not characteristic of depression. The staff suspects an axis II disorder. Cluster B (antisocial, narcissistic, or borderline) seems most likely. Other possibilities include malingering in order to flee punishment or some other danger. The antidepressant trial makes good sense; the selective serotonin reuptake inhibitor is attractive in that it minimizes the risk of overdose, particularly if he were to continue the drug after discharge. The trazodone should alleviate his insomnia. Discharge planning would normally start now, but normally we would know the patient's name by now.

Psychological testing was performed in order to help resolve the diagnostic uncertainty. The MMPI-2 suggested clinically significant levels of depression, tension, and ruminative thinking. His depression was felt to be chronic and superimposed on a socially isolated passive character style. His Beck Depression Inventory score was 25, indicating moderate to severe depressive symptoms. He showed no signs of cognitive impairment on the Mini-Mental State (his score was 30), and his raw scores on the Shipley were indicative of a superior-range IQ. He also had a computerized tomography scan of the head; there were no significant findings.

Three weeks after starting medication the patient showed some signs of improvement. His appetite, sleep, and personal hygiene had all improved. He seemed to be more comfortable in the hospital but repeatedly stated to various staff members that he would kill himself if he were discharged from the hospital and still planned to do so anonymously. He was often quite hostile. In one incident, staff members were suspicious that he had violated the unit's no-smoking policy. He initially denied this, but the staff found cigarettes in his room. He then became extremely angry, predominantly toward a black mental health technician. The patient provocatively invited him to do a "full" body search, stating, "You want to see my asshole, right, boy?" Following this incident, some of the black staff began to complain that the patient was being treated preferentially because he was white and that if he were black he already would have been discharged.

As the evidence accumulates, we have more confidence in the diagnosis of both a depressive disorder and an axis II disorder. Both the psychodiagnostic assessment and the imaging study reinforce our clinical impressions but add little to our thinking, raising questions of their appropriateness in terms of cost/benefit ratios. The patient is starting to look better, perhaps in response to medication, but still talks of suicide and is, if anything, more openly hostile. His character pathology is increasingly clear—he lies to staff, provokes disputes among the staff, and acts in a way that would lead anyone without professional training to call him obnoxious. (Those with training react similarly but use those reactions as clues to the diagnosis rather than as opportunities for retaliation.)

Approximately 6 weeks after the patient had started medication, his roommate, another patient on the unit, told the staff that the patient had not been swallowing his medication and was secretly saving it with a plan to overdose. His room was searched again, and several tablets of paroxetine and trazodone were found hidden under the lining of his eyeglass case. The patient was placed on constant observation, and all further medications were crushed. This incident, together with the patient's lack of significant clinical improvement as he approached his third month of hospitalization, prompted consultation by three of the department's senior psychiatrists. They found the patient to be paranoid and suspected that he was concealing bizarre ideation about death, referring back to his suicide note. They diagnosed probable paranoid schizophrenia and recommended shifting treatment to an antipsychotic drug regimen and planning the patient's transfer to a facility for chronic patients in a state hospital. However, the patient refused to take any medication. He was converted from voluntary to involuntary legal status, and steps were initiated to medicate him against his will. A difference of opinion developed between the treating psychiatrist and the more senior clinical administrators. The treating psychiatrist did not believe that the patient was psychotic but found him manipulative and was concerned about his potential suicidality. The senior psychiatrists believed that he had a chronic psychotic disorder and that his ultimate disposition, discharge or transfer to a state hospital, was the preeminent clinical problem. This tension colored the subsequent legal proceedings.

After consultation with his mental hygiene lawyer, the patient refused transfer to the state hospital and submitted a request for a court hearing. He refused to discuss his suicidality but did express ongoing hopelessness and commented, "I'll do whatever I have to do." He explained his refusal to reveal his identity by his mistrust of people in the hospital and his fear of how any personal identifying information "might be used." However, he did reveal that he had had no contact with his family for 22 years and that both of his parents had died before that time.

The mental hygiene lawyer arranged to have the patient evaluated by an independent psychiatrist. The independent psychiatrist found the patient's claim that he was ready for discharge unfounded and believed him to be at a high risk of suicide. He recommended that the patient remain hospitalized and receive further treatment. His diagnosis was as follows: axis I, major depressive disorder, recurrent without psychotic features; axis II, unspecified. The independent psychiatrist recommended treatment with an antidepressant or ECT; he did not think that the patient was psychotic but did not rule out low-dose antipsychotic medication for his paranoid trends.

Things are not going so well. The patient has been in the hospital for too long, and it is not clear when or how he will be able to leave. He started to get better but then regressed, and his suicidality now seems to the treating staff as much axis II as axis I. The senior consultants offer yet a third possibility—an occult psychotic process that might help to explain his refusal to identify himself. They also suggest an alternative reading of the suicide note—the references to body parts and "chop it up" could be seen as bizarre. There is a suggestion that they think the treating staff has overly identified with the patient and is being "taken in," while the treating staff fears that the administrative staff's concerns about the length of the hospitalization might distract them from his suicidality. The patient's tendency to generate "splits" among his caretakers is dramatic.

Everyone—the patient, the treating staff, the senior staff, and the independent psychiatrist—is at wit's end. The diagnoses range from mood disorders to schizophrenia to axis II, the recommended treatment ranges from antidepressants to antipsychotics to ECT, and the hospital bill continues to mount. The core disputes—over involuntary transfer and involuntary medication—dictate involvement of the court.

Three and one-half months after the patient's admission the court conducted a hearing on his challenge to his involuntary retention and on the hospital's request to initiate involuntary treatment with antipsychotic medication. The patient entered the courtroom carrying a New Yorker magazine. The judge listened to the hospital psychiatrist and the independent psychiatrist and also engaged the patient in conversation, discussing a New Yorker article that each had read. The judge mandated retention for up to 30 days, a trial of antidepressant but not antipsychotic medication, and discussion of ECT with the patient. After the hearing and approximately 4 months into his hospitalization, a trial of the tricyclic antidepressant nortriptyline was initiated. ECT was discussed with the patient, but he rejected the material he was given by the staff as "too simple" and "unscientific."

At this point, when discussing his refusal to reveal his identity, he implied that he feared negative consequences if he did so. He indicated that he would be "forced" to go back to his hometown. The staff began to suspect that he might be a white-collar criminal. Their fantasy was that he would be apprehended and that the story would be in the newspapers.

Approximately 5 weeks after starting nortriptyline, and 4 weeks after a therapeutic blood level was reached, the patient was reevaluated by the independent psychiatrist in preparation for returning to court. (Both the initial hearing and this rehearing were delayed by administrative procedures and various vacations.) The patient was angry that the independent psychiatrist still believed him to be suicidal. The patient stressed that he had only passive suicidal thoughts. He admitted to a slight improvement in his mood and told the judge of his desire to stay in the hospital for 30 more days in order to give the nortriptyline a further chance to help. The judge granted an additional 30-day order of retention but added that at that point, either ECT or the addition of neuroleptics would follow. The patient clearly indicated his unwillingness to accept either of these treatments.

The judge enters the process and functions, in effect, like a senior clinician. She resolves the conflict over treatment (and implicitly over diagnosis), agreeing with the doctors that treatment is necessary, agreeing with the patient that the transfer is premature, and siding with the patient and some doctors, but against others, that an antidepressant rather than an antipsychotic regimen seems best. The question of using antipsychotic medication without a diagnosis of psychosis is not addressed; there is a limit to the court's psychopharmacologic sophistication. The switch to tricyclics was influenced by the possibility that the patient had already received an adequate trial of selective serotonin reuptake inhibitors, although questions of compliance leave us uncertain. The possibility of evaluating future compliance by measuring blood levels makes nortriptyline an attractive choice, although the suicidal potential might make one wary of tricyclics in general. Perhaps most important, the judge empowers both the patient and the therapeutic mission of the hospital, warding off administrative concerns about length of stay and cost. Everybody calms down, treatment starts, and the patient gets better but continues to taunt the staff and begins to show some evidence that he both can and will change enough to be discharged.

During the sixth and final month of hospitalization the patient demonstrated continued improvement in his depressive symptoms. The hospital routine became more burdensome to him. A new roommate who was psychotic punched him, cutting him and breaking his glasses, and it took longer than he thought appropriate for the staff to have them repaired. He began to talk about leaving the hospital, revealing anxiety about how he would fare but denying suicidal ideation. A discharge date was set, but there was repeated discussion of how his refusal to reveal his identity compromised discharge planning. It seemed clear that his depressive symptoms had improved, but it was unclear how to plan continued treatment after discharge. Concerns about the financial cost of the patient's hospitalization had been growing. No formal pursuit of reimbursement was possible while he remained "unknown." For several weeks, staff members had sarcastically remarked that the doctors were interested in retaining the patient in the hope that he would reveal his identity, so that the hospital could pursue financial reimbursement. The day before his planned discharge he revealed his identity to his mental hygiene lawyer and requested the lawyer's assistance in returning to his hometown. The patient had expressed more trust in the lawyer than in any of the hospital staff throughout his stay. On the day of discharge he finally revealed his name to his treating psychiatrist, who had made it clear that otherwise, a prescription for medication was not possible. He was discharged with a prescription and planned to return to his hometown. Approximately 1 month after discharge, he sent a postcard to the staff telling them that he had gotten home safely and thanking them for their help. He signed his name but added that they knew him better as "Unknown."

As the patient's depression improves with treatment his character pathology becomes even more apparent, the burdens of living in a hospital begin to outweigh the advantages, and the paranoid theme is clear. We finally learn his name, but in the end it does not make much difference. He managed to extract more inpatient treatment from the system than the vast majority of patients with this level of pathology, reflecting in part how our legal safeguards against inappropriate coercion and our clinical safeguards against self-destructive behavior can be exploited.

How suicidal was he? What would have happened if he had been discharged earlier? Would more rapid medication have affected his clinical course? We do not know the answers. His final decision to reveal his name and the postdischarge postcard are interesting. They suggest that his anonymity served primarily as a source of power, that he had been able to form a relationship with the staff, and that with this his paranoid trends diminished—all positive prognostic features. In sum, we have the story of a suicidal man with affective and characterologic pathology who is rescued from death but insists on dictating the terms of that rescue. The total cost to society through its mental health and judicial systems was over $150,000.

Received Jan. 21, 1997; revision received April 28, 1997; accepted May 2, 1997. From Cornell University Medical College. Address reprint requests to Dr. Robert Michels, Cornell University Medical College, 418 East 71st St., New York, NY 10021.