Clinicians are required by law to obtain patients' informed consent before initiating any medical treatment or procedure. The process of obtaining informed consent allows patients to make treatment decisions on the basis of the best information. To be adequately informed, patients should know the nature of the proposed treatment and its benefits and risks as well as the benefits and risks of reasonable alternatives. The benefits and risks of no treatment should also be discussed (1,2).

Neuroleptic medications are used in the long-term treatment of schizophrenia and other psychotic disorders (2,3,4,5). However, these drugs have a number of unwanted side effects. For example, some patients develop abnormal and involuntary movements known as tardive dyskinesia (3,5). Tardive dyskinesia may occur in as many as 25 percent of patients on long-term neuroleptic medication.

Neuroleptic malignant syndrome (NMS) is a rare but potentially fatal side effect of neuroleptic medication. NMS is characterized by hyperpyrexia; severe extrapyramidal symptoms, such as rigidity; signs of autonomic instability; and alteration in consciousness (6). According to the best estimates, .1 percent to 1 percent of patients treated with neuroleptic medications develop NMS (6,7,8,9). When NMS is diagnosed early, the mortality rate for patients is unclear. However, one study found that up to 20 percent of cases were fatal (6).

In a previous report on psychiatrists' documentation of informed consent for neuroleptic medications, we noted that in a sample of 30 psychiatrists, 20 (67 percent) always or almost always provided information about tardive dyskinesia (10). The study reported here was undertaken to compare the frequency with which psychiatrists informed patients taking neuroleptics about tardive dyskinesia, a nonfatal but relatively common side effect, and NMS, a potentially fatal but uncommon side effect.

In 1992 psychiatrists at 12 sites who treated a minimum of ten patients on neuroleptic medication were selected to participate in the study. A total of 37 psychiatrists agreed to participate, and 30 completed a mailed questionnaire about their attitudes toward informing patients about the risks of taking neuroleptic medication and whether they did so.

The mean±SD age of the 30 psychiatrists was 44.8±8.3 years. Twenty-four (80 percent) were men. The median year of graduation from medical school was 1973; graduation years ranged from 1956 to 1985.

The 30 psychiatrists were equally divided between teaching and nonteaching hospitals in Ontario, Canada. No more than six psychiatrists from each hospital participated. Face validity of the questionnaire was determined by asking nine experienced psychiatrists and ten experienced general practitioners to complete it.

Student's t test was used to compare means from the Likert-type scale used for responses on the mailed questionnaire.

All the psychiatrists agreed that patients should be informed of a serious material risk that affects one of every 100 treated patients. Blindness or paralysis was given as an example of a serious material risk. Psychiatrists were significantly less likely to inform patients of the risk of NMS than to inform them of the risk of tardive dyskinesia (t= 10.47, df=29, p<.001). Only one psychiatrist reported always or almost always informing patients of the risk of NMS.

Although only one psychiatrist informed patients about the risk of NMS, 18 (60 percent) agreed or strongly agreed that patients should be informed of a serious material risk that affects one of every 100,000 treated patients. Twenty-seven psychiatrists (90 percent) agreed and one (3 percent) disagreed that patients should be informed of a serious material risk that affects one of every 1,000 treated patients.

Psychiatrists were significantly more likely to agree that a serious material risk that affects one of every 100 treated patients should be explained than they were to agree that a risk that affects one of every 1,000 patients should be explained (t=4.01, df=29, p<.001). As noted, all psychiatrists agreed with that statement. Similarly, psychiatrists were more likely to agree that a risk to one in 1,000 patients is more important to explain than a risk to one in 10,000 (t=3.18, df=29, p=.003).

Twenty-eight of the psychiatrists (93 percent) felt that guidelines about informed consent should be available to physicians.

Only one psychiatrist in this sample of 30 hospital-based psychiatrists informed patients taking a neuroleptic medication about the risk of NMS. The results indicate that although psychiatrists may endorse the importance of obtaining informed consent, something changes when it comes to implementing the process. For example, 60 percent of the psychiatrists agreed or strongly agreed that a serious material risk with a frequency as low as one in 100,000 should be explained to patients. However, only one of the psychiatrists always or almost always informed patients about NMS, even though the risk of death from NMS is greater than one in 100,000.

Furthermore, although all psychiatrists agreed that the risks and benefits of neuroleptic medication should be explained to patients, they were significantly less likely to inform patients about the risk of NMS than about the risk of tardive dyskinesia, a relatively common but nonfatal side effect.

There are a number of possible explanations for psychiatrists' not informing patients about NMS. First, they may not be aware of this syndrome. American Psychiatric Association task force reports have discussed tardive dyskinesia but have never specifically addressed NMS (3,5). Psychiatrists may be unaware of the frequency of this side effect and may never have seen a case. They may believe that NMS is so rare that it does not warrant obtaining informed consent. The lower the risk of an adverse event, the less likely the psychiatrists in our study were to believe that patients should be informed of that risk.

Second, psychiatrists may be uncertain about what they should say to patients about NMS, and so they do not say anything.

According to the principle of nonmaleficence, physicians should not inflict evil or harm on patients (2). Psychiatrists may be hesitant to inform patients about NMS because they believe it to be a relatively rare side effect that should not affect the decision-making process given that most patients who have schizophrenia need to take a neuroleptic medication. They might believe that informing patients would increase their anxiety about taking the medication.

One limitation of this study is its small sample size. Another is that the psychiatrists in the sample were hospital based, and the results may not be generalizable to all psychiatrists. However, hospital-based psychiatrists could be expected to be more likely to obtain informed consent than community-based psychiatrists because of their greater exposure to continuing education, quality assurance programs, and audits.

This research was supported by grant 6606-5304-301 from the National Health Research and Development Program of Health Canada.

The authors are affiliated with the department of psychiatry at Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5. They are also with the Clarke Division of the Center for Addiction and Mental Health in Toronto. An earlier version of this paper was presented at the Ontario Psychiatric Association meeting held in January 1998 in Toronto and at the Canadian Bioethics Society meeting held October 1998 in Toronto.