“Dying Comfortably” of Starvation and Dehydration: What Is the Evidence?

In 1993, Bernat, Gert, and Mogielnicki published an article in the Archives of Internal Medicine called “Patient Refusal of Hydration and Nutrition: An Alternative to Physician-Assisted Suicide or Voluntary Active Euthanasia.”

Selling Starvation

In this article, they offer arguments to alleviate the guilt feelings of physicians who help patients with deliberate starvation and dehydration. They emphasize the obligation to comply with the patient's refusal of food and fluids.

However, they do not address the issue of whether the patient's refusal is properly informed or not. A patient's choice is greatly influenced by the information he receives from his physician and other medical personnel. If the information he receives is flawed or biased in any way, and if he bases his decision of self-starvation upon this information, then whoever gave him this information would indeed be responsible for his death.

Patients are being told that starvation is comfortable. Bernat and his coauthors wrote, “The most pressing need is to dispel the myths about suffering caused by dehydration and to publicize as widely as possible to both physicians and their terminally ill patients the availability of PRHN [Patient Refusal of Hydration and Nutrition] as a means of shortening the dying process.”

Bernat and his coauthors call the idea that dehydration causes suffering a “myth,” and indicate that research has disproven this “myth.”

The typical study cited by proponents of Patient Refusal of Hydration and Nutrition is either a case report about how a patient starved and didn’t complain or a survey of hospice nurses' opinions about whether starving patients were comfortable. In this article I would like to introduce the reader to some quite ordinary standards of medical research and to evaluate this kind of research according to these standards.

Factors That Affect Perception of Pain

First, we must discuss the formidable philosophic difficulties of doing any kind of conclusive medical research on something so subjective as the sensations of a dying patient.

The psychological experiences of a human being are determined not just by his physical condition and his circumstances, but also by the ideological slant that he brings to these circumstances. The pain that a person reports, and indeed, even the pain that a person experiences, has a lot to do with his thinking process.

A person who is doing something that he deeply believes in and greatly desires to do does not complain about the incidental discomforts that occur in the process. In fact, he may not even perceive these potential discomforts. The classic example of this are reports that injured soldiers often do not feel pain at the moment of their injury, but only later.

Activists for a cause who undergo a hunger strike speak out about the rightness of their cause. They don't complain about their hunger pangs and their feelings of weakness. Onlookers take their ability to continue fasting as a measure of their devotion to the cause.

Inside the Passion of the ChristThe smiles of Ramon Sampedro, a euthanasia supporter who filmed his own suicide in 1998, are an example of how the ideological slant of a patient determines whether he complains or not. According to a recent account by Gregory Jordan in the New York Times, he “slurped poison through a straw, licked his lips and smiled. Heaving and lunging for the next 40 minutes, he still summoned mischievous smiles throughout his death throes.” Jordan wrote that the actor who played Sampedro in a recent movie thought that Sampedro “kept smiling for the camera as he died because he was amused at the idea of recording himself for the needy arbitrators on both sides of a national debate about death.”

This well-known fact of human psychology — that the reports of human beings about their own sensations are very much affected by their thinking process — is a good example of the kind of factor that has to be taken into account when interpreting the applicability of a medical study. To be widely applicable, medical research has to be conducted according to certain rules. If we do a study on a certain unique group of people, those who chose hospice or those who refuse nutrition and hydration, the conclusions may not pertain to any other group. If we want our conclusions to be relevant to other groups, then subjects cannot be picked according to some special behavior or belief. They have to be randomly selected.

Twisted Results from a Skewed Sample

Random selection is only the beginning. In a typical medical experiment, the subjects are blind to whether they are in the treatment group or the control group. If we want to determine if subjects who are deprived of nutrition and hydration die any more comfortably than those who are nourished and hydrated, the subjects would have to be unaware of whether they are receiving food and fluids or not.

Finally, those standing by and judging the quality of their deaths would have to be impartial judges, not proponents of Patient Refusal of Hydration and Nutrition. Moreover, they also would have to be unaware of whether the person whose comfort they were judging was receiving food and fluids or not.

The research studies that Bernat and his coauthors were seeking to publicize did not meet the criterion of random selection. On the contrary, neither the patients described in these case reports nor the doctors and nurses whose opinions about the comfort of the patient were solicited were randomly selected. The typical subject had refused nutrition and hydration. The typical doctor or nurse rating the patients' comfort worked in a setting where patients were being deprived of nutrition and hydration. There would be a tendency for medical personnel who felt that starvation caused patient discomfort not to practice in these settings.

Typically, the patients and the medical staff were aware that the patients were not being nourished and hydrated. Thus, if the medical personnel rating the patients' comfort had any bias towards believing that starvation is comfortable, this bias could have indeed influenced their ratings.

Later studies of this nature, such as the one done by Linda Ganzini and her coinvestigators, also share this flaw.

In the absence of these ordinary safeguards — random selection and keeping subjects and raters blind to the treatment received by the patient — safeguards which are standard in medical research, case reports and opinion surveys indicating that some patients appeared to starve comfortably do not imply that others would as well.

Also there is the question of interpreting the lack of patient complaints other than thirst. If a patient agreed to be starved to death, he knows that the nurse can give him ice chips to alleviate thirst, so he asks for them when he feels uncomfortably parched. But when he feels a feeling of complete exhaustion and weakness, why would he complain? What could the nurse offer him for this? Given that he has asked to die, he may welcome the weakness as a sign that his request to die is being fulfilled.

Bernat and his coauthors do not claim that starvation and dehydration are painless or euphoric, but, rather, see themselves as dispelling the “myth” that the suffering caused by thirst and hunger is “intractable” or “unbearable.”

To say that starvation, with the option of attentive nurses giving ice chips and sedatives, isn't unbearable isn't saying much. Most painful medical conditions can also be made more bearable by sedatives and good palliative nursing care.

Ira Byock, another proponent of helping patients to die by withholding food and fluids, admits that the clinical reports of patients dying peacefully or in a euphoric state are just anecdotal and have not been matched by research studies. He writes, “While currently available studies do not refute this clinical impression, neither are they sufficient to substantiate it. Thus, while the heightened probability of a gentle passing by PRNH [Patient Refusal of Nutrition and Hydration] is intriguing, at present, it remains speculative.”

Given the glaring flaws in the research about the sensations of starving patients and the ethical impossibility of designing research studies on this question that are any better, it is clearly an expression of bias to publicize to patients the “comfort” of self-starvation.

© Copyright 2005 Catholic Exchange

Dr. Olevitch is a clinical psychologist and author of Protecting Psychiatric Patients and Others from the Assisted-Suicide Movement: Insights and Strategies (Praeger, 2002).

Subscribe to CE
(It's free)

Go to Catholic Exchange homepage