Retrospective study of doctors' “end of life decisions” in caring for mentally handicapped people in institutions in the Netherlands

Questionnaire survey

We drew a sample at random. To reach the number of interviews envisaged we had to draw 142 names. The people selected had to be currently working in the care of mentally handicapped people. Twenty six failed to satisfy this criterion; 15 others were untraceable or unable to participate because they had been ill for a long time. Twelve of the 101 people selected who satisfied the selection criterion and could be traced refused to participate. The 89 doctors reported all deaths of patients for whom they were the attending physician that occurred from 1 March 1991 to 1 March 1996. When less than five years were covered the doctor indicated the actual period. Deaths in hospital and deaths occurring while the respondent was working as a locum were excluded. Eighty nine doctors completed a questionnaire, which comprised six questions for each case, including whether an end of life decision had been taken. Doctors were recommended to use the patients' files while filling in the questionnaire. To estimate annual absolute numbers, we used a weight of 1.79 on the basis of the proportion of doctors represented in the sample.

Interviews

Sixty eight doctors mentioned at least one case in which they had taken an end of life decision, and they were invited to be interviewed about the most recent such death. One refused. Finally, 67 interviews took place between March and June 1996. The interviews were conducted by eight trained interviewers. All of them had experience as a doctor in caring for mentally handicapped people. The interviews lasted between one hour and two and a half hours.

The decisions we studied were withholding and withdrawing life prolonging treatments (non-treatment decisions), relieving pain and symptoms with opiates in dosages that may have shortened life, and ending life by giving lethal drugs (euthanasia by lethal injection). The interview schedules contained many questions identical with those in the Remmelink questionnaire.1 2 Because most mentally handicapped patients are incompetent, a refinement was made about the patient's request. We asked in detail about patients' other actions such as non-autonomous requests or non-verbal communication. Examples of non-autonomous requests were patients saying that they wanted to go to heaven or that they wanted to be left alone. Non-verbal communication included actions of the patient that were interpreted by the doctor or others as a wish to die or stop treatment—for example, constantly removing a feeding tube or resisting all medical treatment.

Mental handicap and diseases were classed according to ICD-10 (international classification of diseases, 10th revision).

Reported deaths

No patient had died in the previous five years for 10 of the 89 respondents, while the remaining 79 doctors reported 859 deaths in an average period of 4.7 years. Of these 79 doctors, 11 had not taken an end of life decision and 68 had taken such a decision in 350 out of 859 cases (41%); 254 were decisions to withhold or withdraw treatment, 92 were to relieve pain and symptoms with opiates, and four were to end life with a lethal drug (table 1).

The number of reported deaths differed considerably from year to year. However, the proportions of the various types of decisions were similar. We estimated absolute numbers for 1995 on the basis of 222 reported deaths and the distribution of decisions to end life for that year. This resulted in an estimated total of 397 deaths. For these deaths we estimated that there would be 135 decisions to withhold treatment, 40 decisions to relieve pain and symptoms with opiates, and 222 deaths with no end of life decision. Although we found no decisions to end a patient's life with a lethal injection in 1995, we assumed that one or two of these decisions would have occurred in that year on the basis of the four cases in the whole sample.

Interviews

Sixty seven doctors described the most recent death in which they had taken an end of life decision (tables 2, 3 and 4). Most doctors were trained as general practitioners (49) and were experienced, having spent an average of 13 years working in this specialty. Thirty six doctors said that they considered themselves to be part of a religious community. Most of the people who had died were under 65 years old, and almost all were considered by the doctor to be incompetent (table 2). They lived in 51 institutions with an average number of 427 residents. Many had diseases of the nervous and digestive systems. Two patients explicitly asked to die. Twenty three doctors said that they had noticed non-autonomous requests and relevant non-verbal communication (table 3).

The most common reason for taking the decision was the pain or suffering of the patient (19 doctors; table 3). In 46 cases the doctor discussed the decision with a colleague. Nurses were almost always consulted (60). Relatives or representatives were less often brought in (50), and consultation with the patient was reported in two cases (table 4).

Three cases of euthanasia were reported. These are included in the third column of tables 2, 3 and 4. The patients in these cases had severe illnesses, including severe heart problems (two patients), tetraplegia, severe epilepsy, and recurrent pneumonia (two patients). In two cases the suffering of the patient had increased substantially before the decision was taken, and the doctors saw no more options to alleviate this suffering. In the other case the doctor mentioned that the terminal phase had started. The doctors estimated that the lives of the patients were shortened by one week at most in all cases. All of them were incompetent and did not express their wishes about the decision. One patient, however, had refused hospital care a year before the decision, when he could still communicate. The doctors consulted a colleague or the nursing staff, or both, in all cases, and two doctors discussed the decision with the patient's relative or representative. Deaths were reported as natural in all cases.