Attitudes toward end-of-life decisions other than assisted death amongst doctors in Northern Portugal

Background Doctors often deal with end-of-life issues other than assisted death, such as incompetent patients and treatment withdrawal, including food and fluids. Methods A link to a questionnaire was sent by email three times, at one-week intervals, to the doctors registered in the Northern Section of the Portuguese Medical Association. Results The questionnaire was returned by 1148 (9%) physicians. This study shows that only a minority of Portuguese doctors were willing to administer drugs in lethal doses to cognitively incompetent patients at the request of a family member or other close person, and even less would do it on their initiative. Most doctors favored the withdrawal of life support measures in advanced and progressive diseases at the patient’s request. Still, much fewer doctors agreed with the suspension of supportive life measures at the request of a family member, another close person, or by their own unilateral decision. However, fewer agreed with that action concerning the rest of the food and fluids. Portuguese doctors favor the administration of drugs for suffering control, even foreseeing they could shorten life. Conclusion Most doctors in this study respect patients’ autonomy but disagree with measures decided by others that have an impact on patients’ survival. They also agree with the administration of drugs for suffering control, even considering the possibility of shortening life.


Introduction
The social and economic development and the huge evolution of medicine, mainly from the middle of the 20th century, signi cantly prolonged human life expectancy. Modern medicine has a wide range of treatments and devices to support life, to cure or prolong life. However, these methods are often used in patients with far advanced and irreversible diseases where it is not advisable to employ them, because they may be useless or even detrimental. Withholding or withdrawing treatments often imply di cult decisions because, besides clinical considerations, they involve ethical, philosophical, religious, and even legal issues.
Health care professionals deal sometimes with euthanasia and assisted suicide requests, mainly in countries where those practices were legalized, but sometimes also in countries where they were not. Those issues are perhaps the most visible ones and, therefore, the most publicly discussed, but doctors and health care teams may have to deal with many other di cult decisions. Whenever possible, patients and families should also have an important role in those decisions, directly or through advanced directives. After all, they are the most interested persons in those processes and their involvement may be what divides a legitimate from an illegitimate decision. However, some unilateral decisions taken by health professionals are also legitimate.
Other decisions about end-of-life issues include withholding or withdrawing arti cial ventilation, hemodialysis, transfusions, or antibiotics. Even after those decisions are made, one question often remains: should nutrition and hydration also be withheld or withdrawn?
Withholding and withdrawing some treatments are sometimes deemed euthanasia, but those actions are of a different nature.
When appropriately taken, those actions respect patients' autonomy or avoid prolonging the process of death. In fact, there is an important distinction between prolong life and prolong the process of dying, when patients are naturally close to death. When competent patients request treatment withdrawal, doctors should respect their wishes. However, how to decide when patients are incompetent? Is it acceptable that a family member decides for them, or can the doctor or the health team decide based only on their own judgment?
A study carried out in six European countries, which did not include Portugal, concluded that medical decisions frequently precede dying in all participating countries although with variations among them and that patients and relatives are generally involved in decision-making in countries in which the frequency of making these decisions is high [1].
Ten years ago, I published a study about oncologists' attitudes on end-of-life issues, excluding assisted death, trying to nd out their opinion on those questions [2]. Now the scope of this study is broader, including doctors from all specialties concerning the same issues.

Methods
This study was carried out in collaboration with the Portuguese Medical Association. In this study the questionnaire used was the same used in the previous study mentioned above. A link to the questionnaire was sent to doctors through the list of emails of the Northern Section of the Portuguese Medical Association, which was sent three times, at one-week intervals, in February 2019. The link included an explanation of the reasons and the importance of the study assuring the anonymity for those who participate. The author was interviewed by one of the main Portuguese television channels and the interview was broadcasted on news programs, including at prime time, as an additional mean of encouraging doctors to reply.
The development of the questionnaire, which was the same as the one used in the previous study was described in that paper [2]. The questionnaire included questions on various end-of-life issues and those concerning demographic data, on assisted death and palliative care and other attitudes on end-of-life decisions. The study presented here focus on the attitudes other than assisted death.
In the statistical analysis to evaluate the existence or not of associations between categorical variables the chi-square test was used. The variables with a signi cant association were further analyzed using a multinomial logistic regression for the multivariable analyses. The level of signi cance was deemed to be 0.05. The IBM SPSS Statistics version 25.0 software was used to analyze the data. The factors included in the analyses were the demographic data presented in Table 1.

Results
From the 13 192 questionnaires sent, 1148 (9%) were received. The demographic data of the doctors are displayed in Table 1.
Concerning gender and age, there are signi cant differences (p < 0.001) between the entire population of physicians and those who replied to the questionnaires. Therefore, it is not representative of the population studied. The questionnaires were completed carefully and consistently, as most questions had a 100% rate of replies, and the highest rate for any missing answers was never higher than 0.7%.

Cognitively incompetent patients
To the question "Would you give one or more drugs in lethal doses to someone unable to make decisions because of impaired consciousness and suffering from an incurable, advanced and progressive disease that would inexorably lead to death, at the request of a family member or other close person?", 117 (10%) answered "yes" ( Table 2). In the univariable analysis, gender, religion, religious practice, specialty, and number of patients seen in the last year, were associated with the doctors' opinion, whereas age, marital status, and workplace were not. In the multivariate analysis, catholic doctors, religious practicing and to have clinical specialty were less often in favor of those practice (Table 3). Fourteen (1%) doctors said they received such requests frequently, and 146 (13%) received them rarely; 10(1%) said they had already performed such an act.  The reference category is: No Q1. Would you give one or more drugs in lethal doses to someone unable to make decisions because of impaired consciousness and suffering from an incurable, advanced, and progressive disease that would inexorably lead to death, at the request of a family member or other close person?
Q2. Do you think that those acts should be allowed by the legislation? (referring to the question above) Q3. If you had an incurable, advanced, and progressive disease that would inexorably lead to death and you were unable to make decisions because of impaired consciousness, would you like that a doctor gives you one or more drugs in lethal doses, at the request of a family member or other close person?
Q4. Would you give one or more drugs in lethal doses to someone unable to make decisions because of impaired consciousness and suffering from an incurable, advanced, and progressive disease that would inexorably lead to death, by your own initiative (not on the request of anybody)?
Q5. Do you think that those acts should be allowed by the legislation? (referring to the question above) Q6. If you had an incurable, advanced, and progressive disease that would inexorably lead to death and you were unable to make decisions because of impaired consciousness, would you like that a doctor gives you one or more drugs in lethal doses, based only on his or her judgment?
Referring to the question above, 230 (20%) patients thought that these acts should be allowed by the legislation (Table 2). In the univariable analysis, age, religion, religious practice, specialty, and the number of patients seen, were associated with that opinion whereas gender, marital status, and workplace were not. In the univariable analysis, age, religion, religious practice, specialty, and number of patients seen, were associated with that opinion whereas gender, marital status, and workplace were not. In a multivariable analysis, younger doctors, religion, religious practicing, and the higher number of patients with advanced diseases seen, were less likely to agree with the legalization of these practices (Table 3).
To the question "If you had an incurable, advanced, and progressive disease that would inexorably lead to death and you were unable to make decisions because of impaired consciousness, would you like that a doctor gives you one or more drugs in lethal doses, at the request of a family member or other close person?", 321 (28%) answered "yes ( Table 2). In the univariable analysis age, religion, religious practicing, specialty, and the number of patients seen were related to that opinion, whereas gender, marital status, and workplace were not. In the multivariable analysis, only religion and religious practices were more likely to be against that practice.
The question "Would you give one or more drugs in lethal doses to someone unable to make decisions because of impaired consciousness and suffering from an incurable, advanced, and progressive disease that would inexorably lead to death, by your own initiative (not on the request of anybody)?", had a positive answer from 57 (5%) doctors (Table 2), and 15 (1%) said they had already done it; 181 (16%) thought that those acts should be allowed by the law. In the univariable analysis, religion and religious practicing were associated with that opinion, and in the multivariable analysis, only religion practicing doctors were less likely to support that practice.
To the question "If you had an incurable, advanced, and progressive disease that would inexorably lead to death and you were unable to make decisions because of impaired consciousness, would you like that a doctor gives you one or more drugs in lethal doses, based only on his or her judgment?", 161 (14%) answered "yes" ( Table 2). In the univariable analysis, age, religion, religious practice, and the number of patients seen, were associated with that opinion. in the multivariable analysis, only age and religious practice were associated with that opinion, being less likely to support it.

Treatment Withdrawal
To the question "Do you think that in relation to someone with an incurable, advanced, and progressive disease that would inexorably lead to death, it is legitimate to withdraw support of life measures at his or her explicit, repeated, informed, and wellpondered request?" 890 (78%) doctors replied "yes" (Table 4). In a univariable analysis age, gender, marital status, religion, religious practicing, workplace, and the number of patients with advanced diseases seen in the last year had a signi cant association with that opinion. In the multivariable analysis, the younger doctors and those who see more patients are more likely to answer yes, whereas religious non-Catholic doctors and religious practicing were more likely to answer differently (Table 5).
Concerning the question "Would you withdraw measures such as feeding and hydration? (referring to the question above)" Only 220 (19%) replied "yes" (Table 4). In the univariable analysis, age, religion, religion practicing, workplace, and the number of patients with advanced diseases seen were associated with that opinion. In the multivariable analysis, feeding and hydration withdrawal were favored by doctors who saw more patients and unfavored by younger and catholic doctors.  Q2. Would you withdraw measures such as feeding and hydration? (referring to the question above) Q3. Do you think that in relation to someone with an incurable, advanced, and progressive disease that would inexorably lead to death, who is unable to make decisions because of impaired consciousness, it is legitimate to withdraw support of life measures at the request of a family member or other close person?
Q4. Would you withdraw measures such as feeding and hydration? (referring to the question above) Q5. Do you think that in relation to someone with an incurable, advanced, and progressive disease that would inexorably lead to death, who is unable to make decisions because of impaired consciousness, it is legitimate to withdraw support of life measures by the unilateral decision of the doctor or the healthcare team?
Q6. Would you withdraw measures such as feeding and hydration? (referring to the question above) To the question "Do you think that in relation to someone with an incurable, advanced, and progressive disease that would inexorably lead to death, who is unable to make decisions because of impaired consciousness, it is legitimate to withdraw support of life measures at the request of a family member or other close person?", 315 (27%) doctors responded "yes (Table 4).
In the univariable analysis, religion, religious practice and specialty had a statistic association with that answer. In the multivariable analysis, only non-Catholic religion and religious practice were more likely to disapprove that opinion (Table 5). To the question "Would you withdraw measures such as feeding and hydration? (referring to the question above), only 136 (12%) answered "yes". In the multivariable analysis, doctors who see more patients were more likely to favor that position and religious doctors were more likely to disapprove it.
To the question "Do you think that in relation to someone with an incurable, advanced, and progressive disease that would inexorably lead to death, who is unable to make decisions because of impaired consciousness, it is legitimate to withdraw support of life measures by the unilateral decision of the doctor or the healthcare team?" 325 (28%) patients responded "yes". In a univariable analysis age, marital status, religion, religious practicing, specialty, workplace, and the number of patients seen were statistically associated with that answer. In the multivariable analysis, doctors working in hospitals were more likely to favor that practice, as well as doctors who see more patients; on the other hand, doctors with non-Catholic religions were more likely to be against that opinion. To the question "Would you withdraw measures such as feeding and hydration? (referring to the question above)" only 158 (14%) replied "yes". In a multivariable analysis, doctors in the age range between 31 and 40 years, doctors working in cancer institutes, and doctors who see more patients were more likely to favor that practice, whereas Catholic doctors were more likely to disapprove that practice (Table 5).

Suffering Control
To the question of administering drugs (ex., morphine) for suffering control, even foreseeing they could shorten life, 969 (84%) answered: whenever it is necessary. And on the hypothesis of the doctor being in that suffering situation, 1103 (96%) would like for someone to administer such a drug that would relieve their suffering even if it could shorten their lives.

Cognitively incompetent patients
The practice, which is called non-voluntary euthanasia, meaning killing patients incapable of given consent, was rst reported in The Nederland's [3]. The term euthanasia applied to those acts is misleading, as really patients are not asking to die, it is a decision of someone else. Even those in favor of euthanasia may disagree with those acts. These acts may be called murder [4].
In fact, euthanasia is usually de ned as a response to a voluntary request from a patient [4]. Therefore, the expression voluntary euthanasia is incongruent as it implies that euthanasia can be nonvoluntary. This study shows that only a minority of Portuguese doctors were willing to give one or more drugs in lethal doses to cognitively incompetent patients with an incurable, advanced, and progressive disease at the request of a family member or other close person, and even less would do it on their own initiative.
More doctors, but even so a minority, answered that if they were the patient, they want that the drug or drugs were given to them at the request of a family member or other close person. A smaller number would accept that the decision was taken by a doctor's initiative. This was also observed when doctors were asked about euthanasia [5] and in our study published in 2010 on the same topic [2]. Those data mean that some doctors want for themselves what they are not prepared to do for their patients.
They feel perhaps that those decisions involve a responsibility they cannot take. A small percentage of doctors, although a little higher, think that those acts should be allowed by the legislation. Therefore, most doctors were against those practices.
The most important factors consistently associated with doctors´ opinions were religion and religious practice. Those doctors were even more against those practices. It is not surprising that religious doctors were more likely to be against those practices, as several studies showed that religious doctors oppose euthanasia and assisted suicide more often than non-religious doctors [5][6][7][8][9][10][11][12][13]. Nevertheless, not all studies reach that conclusion [14][15][16].
Younger doctors were also more against the legalization of given lethal doses to cognitively incompetent patients in the mentioned conditions. This is a curious observation because the same younger doctors favored more assisted death practices [5]. They were even against those practices if they were the patient.
Against the legalization of those acts were also doctors who follow more patients with advanced diseases. The in uence of contacting with more patients with the approval of those practices was consistent with their opinion on legalizing of assisted death [5]. The observation that physicians who more frequently deal with patients with far advanced diseases were more likely to oppose assisted death than others was also made in other studies [7,9,17], but not in all of them [18]. Emmanuel explains this fact as the different perspectives of those doctors who are more distant from terminally ill patients and see the problem of assisted death in the abstract, as a philosophical question. But doctors who deal directly with the situations may see them as more real and personal [9]. An analogy with interventions that cannot be called assisted death for reasons already discussed seems reasonable.

Treatment withdrawal
Only 6% of doctors answered "no" to the withdrawal of support of life measures in advanced and progressive diseases at the patient's request. In fact, 78% answered "yes" and more 17% answered "in some circumstances". Much fewer doctors agreed with the suspension of support of life measures at the request of a family member (or another close person) or by the unilateral decision of the doctor or the health team. However, the answers "In some circumstances" were raised.
The Portuguese law, as occurs in many other countries [19], grants patients without cognitive impairment the right to refuse therapies, even those that sustain life. Data of this study seems to indicate that most doctors act according to Portuguese law and respect the ethical principle of patients' autonomy. However, they did not recognize the right of family members or even themselves to decide on this issue, although many could admit that, in some circumstances, that decision may be justi ed.
Under Portuguese law, family members are not legal representatives of mentally incompetent patients. Therefore, the wishes of family members are not determinant for decision-making in those situations. The role that families may have in those situations, according to the law, is different among countries. For example, in the United Kingdom, the law is like the Portuguese one [20].
On the contrary, in the United States, patients' relatives are legally appointed decision-making surrogates [19]. Doctors are entitled to decide on treatment withdrawal in cognitively incompetent patients acting in the patients' best interest. Nevertheless, most doctors answered that it is not legitimate that they unilaterally withdraw life support measures. Those data seem to indicate that doctors feel that a consensus among the health care team and family members would be the preferred way to reach a fair decision.
Concerning the suspension of feeding and hydration, referring to the questions indicated above, the percentage of answers "In some circumstances" was roughly one-third, and the percentage of "No" varied from 46 to 57%. Therefore, most physicians would not suspend feeding and hydration, but about one-third admitted it could be justi ed in some circumstances. Maybe most consider feeding and hydration a basic care and not a treatment even when it is delivered through medical devises. When patients at the end of life are able and willing to eat and drink, it is straightforward that food and uids should be provided; without a doubt, this is basic care. However, if patients are unable to do so, the provision of nutrition by arti cial means is controversial; if this is basic care or medical treatment is debatable. It would be stressed that what is asked in this study is about patients with advanced diseases and in this case, hydration seems not to prolong life, although it may have an impact on quality of life [21]. There are different situations of patients without an advanced disease who could have their life prolonged if feeding were maintained, such as in cases of permanent vegetative state [22]. In those cases, the life of patients is certainly shortened without nutrition and the question here is if patient bene t from nutrition and nor merely if they have their heart beating and breath spontaneously. However, these situations are beyond the scope of this study as well as cases of advanced directives. A study carried out in six European countries, which did not include Portugal, showed that decisions to forgo feeding and hydration varied among countries and that withholding was more frequent than withdrawal [23].
Although the numbers are not coincident, they go roughly in a similar direction compared with the previous Portuguese study [2].
In this study, decisions on treatment withdrawal were in uenced mainly by age, religion, and the number of patients with advanced diseases that doctors followed in the last year. Younger doctors tend to favor withdrawal at the patients 'request, but most did not favor withdrawal of feeding and hydration. Doctors who see more patients mostly favor treatment withdrawal in general and even feeding and hydration. In the contrary sense are religious doctors who do not favor the withdrawal of treatment, feeding and hydration even at the patients´request. The comparison with other studies is di cult because most of them did not detail the physicians' characteristics as they did on studies concerning assisted death.

Suffering control
Portuguese doctors seem to accept the administration of drugs for suffering control, even foreseeing they could shorten life, and more often if they were the patient. This attitude is often justi ed by the doctrine of double effect. However, Sykes and Thorns conclude that this doctrine is not essential for the justi cation of using these drugs, because there is no evidence that the use of opioids and sedatives is associated with hastening death and may act as a deterrent to the provision of good symptom control [24].
The main weakness of this study is the relatively small percentage of questionnaires sent back and the lack of representativeness of the sample of physicians. This may cast doubts on the generalizability of the data. Much effort was made to call the attention of doctors to this study, and obviously, they cannot be compelled to reply. Nevertheless, the absolute number of doctors who replied is not low, and this is, so far, the best approximation to the opinions of Portuguese doctors on end-of-life issues, as there are no other data on this topic in Portugal.

Conclusion
Most doctors who participated in this study do not agree with the administration of drugs to incompetent patients with incurable, advanced and progressive diseases. Nevertheless, they would respect the patient's wishes to withdraw life-supporting measures, although not so much when the request came from another person. Most would disagree with the withdrawal of feeding and hydration. Physicians, nonetheless, accept to administer drugs for suffering control, even foreseeing it could hasten death.

Declarations
Funding: The author received no nancial support for the research, authorship, and/or publication of this article.
Con icts of interest/Competing interests: The author declared no potential con icts of interest Availability of data and material: data is available on request from the author