Marquette Elder's AdvisorVolume 3Issue 2 Fall Article 7

Terminal Sedation and DehydrationCharles F. McKhannYale University School

Follow this and additional works at: http://scholarship.law.marquette.edu/eldersPart of the Elder Law Commons

This Featured Article is brought to you for free and open access by the Journals at Marquette Law Scholarly Commons. It has been accepted forinclusion in Marquette Elder's Advisor by an authorized administrator of Marquette Law Scholarly Commons. For more information, please contactmegan.obrien@marquette.edu.

Repository CitationMcKhann, Charles F. (2001) "Terminal Sedation and Dehydration," Marquette Elder's Advisor: Vol. 3: Iss. 2, Article 7.Available at: http://scholarship.law.marquette.edu/elders/vol3/iss2/7

Terminal Sedation andDehydration'

The range of medical intervention at

life's end can vary greatly, from

sedation to euthanasia.

By Charles F. McKhann

ichael was thirty-seven when hebegan to have abdominal pain.It was a vague pain in the upperpart of the abdomen, and wasthought to be due to an ulcer.

However, two months later, the pain had increasedand it was clear that he was losing weight in spite ofeating fairly well. An operation was performed thatrevealed that he had cancer of the pancreas. More-over, the tumor was extensive and could not beremoved surgically. Michael declined chemotherapyafter hearing that it was very unlikely to prolong hislife significantly. Although he had increasing discom-fort, it was controlled with medication and hecontinued to be fairly active as a dentist.

Michael was part of a large Catholic family, hav-ing three brothers and a sister in addition to his wifeand four small children. His parents were still alive.Understanding his grim prognosis, namely that hemight have only a few weeks to live, Michael dis-cussed his outlook with all of the members of hisfamily except the children, whom he thought were

Charles F. McKbann, M.D. is a professor in the Depart-ment of Surgery at Yale University School of Medicine,New Haven, Connecticut.

too young to understand. These discussions weremostly one on one, but occasionally in small groups.He was a strong-willed, pragmatic man who was notprone to sentiment. He explained, "This is the situ-ation, and I am sorry about it. I am sorry that mylife is going to be cut short and that I have to die soyoung, leaving Nancy and the children. If there issomething beyond the grave, I will enjoy it. Untilthen, I would like to have as much togetherness andas little sadness as possible with all of you."

Michael also came up with a new idea, that whenhe was no longer able to eat or drink by himself, hewould stop doing so altogether. He did not want tohave anyone helping him to eat, nor did he want tohave any tubes or intravenous feeding. He realizedthat this step would probably shorten his life by afew days, but also felt that the quality of his life, aswell as that of his family, would be totally destroyedif he waited for the disease to take its natural course.He wanted to have some control over when and howhis life ended. His family could see that he was go-ing downhill rapidly, and no one objected to hisproposal. His physician, a personal friend, promisedto provide him with good pain relief and came to hishome often. He wanted to spend as much time athome as possible, but wanted to die in the hospitalso that the home itself would not be associated withthe memory of his death.

Five weeks after his surgery, he felt the time hadcome to stop eating or drinking. Nobody tried totalk him out of it, and his physician supported himwith appropriate medication so that he felt very littlediscomfort. A brother from out of state came to spenda day with him, preferring to be with him when hewas awake and coherent rather than at the time ofhis death. The dying process took a week. Towardsthe end, he was comatose and was transferred to the

I Elder's Advisor

hospital, as he requested, where he spent his last twodays. The timing of his death was predictable, andmost of his family was present.

During the final week, and in the weeks andmonths following his death, there were no recrimi-nations. The entire family felt that Michael's choicewas the right choice under the circumstances, andmany said that they would do the same thing in asimilar situation. At no time did he seem depressed.Indeed, he provided meaningful moral support toother members of his family, particularly his wifeand children. His memory lives on with his familyas a strong-willed, "nuts and bolts" guy, who facedhis impending death directly and unflinchingly. Hisown acceptance of "what must be" provided a foun-dation of reality and acceptance for his family thatthey still remember.

I heard the history of Michael's death from hisbrother, shortly after it occurred. It was my first ex-posure to the concept of voluntary terminal sedationand dehydration, and I was deeply impressed withhow autonomous and brave the step seemed to be. Ihave subsequently seen it in a broader perspective.It was still an autonomous and brave decision, but itmay also be a sad but necessary compromise withreality. I do not know if Michael had the remotestthoughts of any form of assisted dying. At the presenttime, other end-of-life options, including assistedsuicide and euthanasia, are against the law, but ter-minal sedation is not. The components of terminalsedation that I will look at include dehydration, se-dation, double effect, intent, and some legalconsiderations. First, we must see where terminalsedation fits in the spectrum of possibilities for theperson who wants to die on his own terms.

Clinical DifferencesThe range of medical intervention at the end of lifeincludes none at all (where none is available orwanted), refusing or discontinuing treatment includ-ing hydration, various forms of sedation, assistedsuicide, and euthanasia. Morally, patient refusal oftreatment, terminal sedation, assisted suicide, andeuthanasia are very similar. The patient wants to diesooner rather than later and at some level, even if itis against the law, the physician may be willing tohelp. If one is acceptable, they probably all shouldbe. However, legislatures, courts, and many physi-cians and medical ethicists see significant differencesbetween these activities. The advantage of terminal

sedation over assisted suicide and euthanasia at thepresent time is that sedation is legal. This is an im-portant dividing line.

The right to refuse treatment is a basic elementof patient autonomy. In making such a decision, thepatient shoulders most of the responsibility, and thephysician is required by law to honor the decision,even if it appears to be contrary to the best interestsof the patient. It is now widely accepted that the bodyof a competent person cannot be violated againsthis will by any treatment, however beneficial it mayseem to be. Refusal can be for highly technical sup-port systems, such as renal dialysis or artificialventilation, or for more basic elements of support,such as artificial feeding, or any other treatment thatcould extend life. It can also include refusing nor-mal food and drink. At this point, it becomesvoluntary terminal dehydration. The addition of se-dation to control symptoms advances this to terminalsedation.

A person who refuses to eat or drink can only benourished artificially. Since administering unwantednourishment is against the law, the refusal must beaccepted. This point is important because it extendsthe concept of terminal dehydration from the pa-tient who is terminally unable to eat or drink, to onewho is pre-terminal and refuses to eat or drink. In-deed, theoretically, death by dehydration can beelected by anyone who wishes to take such a step toend his life, regardless of age or medical circum-stances. Most people who are physically well butsuffering from depression have other ways of killingthemselves that are instantaneous and carry little riskof outside intervention.

The use of drugs to ameliorate pain, anxiety, andgeneral distress in a dying patient, representing thesimplest application of terminal sedation, is recog-nized and used throughout the world where theappropriate drugs and medical care are available.This use of sedation is morally, medically, and le-gally acceptable, and has been for centuries. Sedationat the end of life is the natural extension of provid-ing narcotics to relieve the suffering of any patientwho has a chronic, painful illness. 2 There is no rea-son why sedation for the management of symptomscannot be extended to the patient whose terminaldehydration is elective, just as it would be to onewhose dehydration was an obligatory part of thenatural course of his disease. As will be seen, termi-nal dehydration and sedation can be treated as

ARTICLE I Terminal Sedation and Dehydration 45

separate entities, but they can also be blended to-gether in various ratios that give their combinationa special place in this spectrum.

Beyond refusal of treatment, terminal sedation,and dehydration are two much more controversialareas of intervention: physician-assisted suicide andeuthanasia. In assisted suicide, the patient is given aprescription for lethal drugs that he may take whenhe feels that the distress of the illness is overwhelm-ing. Many patients feel reassured that they are "incontrol" by having the drugs on hand, even thoughthey may never take them. In physician-assisted sui-cide, the doctor's role is a more active onepsychologically, morally, and legally, even though itis limited to filling out a prescription for a poten-tially lethal drug. Moral responsibility is shared withthe patient, who must still undertake the final act oftaking the drug. Many physicians would prefer this,since it does not require being present at the time ofdeath. Indeed, the physician may not even know ifthe patient died of his terminal disease or from tak-ing the drug. Physicians may deceive themselves, orbe deceived by their patients, into thinking that theprescription was actually requested for sleep. Lawspermitting assisted suicide have been looked at andrejected by many states, but were passed and arecurrently in effect in Oregon.

In voluntary euthanasia, the patient asks his phy-sician to personally administer a lethal dose of drugs,usually intravenously. Here the final step is clearlyin the hands of the physician. Death occurs within afew minutes of administration of the drugs, and thephysician's direct responsibility is obvious. For thisreason alone, many physicians who support the con-cept of assisted dying prefer that it be assisted suicideand not euthanasia. Legislators who support assisteddying view euthanasia as being more radical andvulnerable to abuse than assisted suicide. Conserva-tive religious groups view euthanasia as the "end ofthe line" and often cite assisted suicide as a danger-ous step in that direction.

DehydrationMany people die of dehydration as a natural out-come of their disease. This includes most people whohave cancer, particularly intra-abdominal cancers,severe stroke, or other serious deterioration of men-tal function, or any progressive illness that leadsto increasing weakness to the point of being unableto handle food or liquids. At that point, artificial

feeding, intravenously or by a gastric tube, can beinstituted if the patient's underlying illness is not ter-minal. Such steps are usually omitted if the patient isnot expected to live long. People who die at homerarely have access to artificial nutrition. It is gener-ally understood that when a person is admitted to ahospice facility, no artificial feeding of any kind willbe instituted, and when a person becomes unable toeat or drink, no attempt will be made to circumventthis terminal event. In non-fatal illnesses such as se-vere stroke, dementia, or brain death, the use ofsupportive treatment in the form of artificial nutri-tion may be seriously questioned by the family and/or physician, and eventually rejected or withdrawn.However, in many parts of the world that do notshare the widespread use of our medical advantages,patients with fatal chronic illnesses usually eventu-ally die from dehydration. There are strong reasonsto question whether our own medical technologyshould be utilized as frequently as it is.

Cessation of all fluid intake, particularly if nofood is taken, usually causes lethal dehydration inone or two weeks. There is no obvious systemic re-action in the first few days. The most commonsymptoms are thirst and dryness of the mouth. Mildsedation can control the thirst easily and oral dry-ness responds well to sips of water, sucking on icechips, and lubrication of the lips. Neurological signs,consisting of lethargy, weakness, confusion, and even-tually progressive coma, appear later, beginning infour or five days. Pain and discomfort are moder-ated by accumulation of ketones in the body as itutilizes fat rather than carbohydrates to meet its en-ergy needs, and by a form of autosedation, the causesof which are not known, seen in the dying process.Death is due to loss of circulating blood volume andto electrolyte imbalance, particularly the accumula-tion of potassium and calcium, which cause cardiacarrhythmias.

Many studies report that dehydration in termi-nal disease is associated with no significant suffering.3

"For individuals carrying an intolerable burden ofillness and disability, or those who have no hope ofever again enjoying meaningful human interaction,the withdrawal of food and fluid may be consideredwithout concern that it would add to the misery."4

Two hospice nurses wrote, "We have not seen evi-dence that dehydration occurring at the terminationof life results in any pain or distressing experiencesfor the patient. "I The symptoms of thirst or dryness

I Elder's Advisor

of the mouth can be easily relieved.6 When terminaldehydration is elected by a patient who is not reallyterminal, control of thirst is well within the range ofcommonly used sedation.

In spite of this information, terminal dehydra-tion is associated in the minds of most people withserious and unnecessary suffering. Families and phy-sicians alike may be reluctant to suggest dehydrationbecause of the universal symbolism associated withfeeding and nurturing as expressions of care andlove.7 Voluntary dehydration is a way to end life thatoften requires enough willpower and resolve on thepart of the patient to not only see the process throughhimself, but also enlist the support of reluctant fam-ily members and even physicians.

Dehydration is also thought to be closely linkedto starvation, and both have very negative connota-tions. Starvation is a time-honored form ofpunishment and, when self-imposed, is often usedto gain public sympathy for political purposes. Inboth of these situations, dehydration is not includedfor the simple reason that death would occur muchtoo soon. Death from dehydration in terminal dis-ease is usually preceded by significant starvationbecause adequate amounts of fluids are more easilyingested than sufficient food. However, in voluntarydehydration this may not be the case. It must thenbe understood that the critical step is dehydration.Starvation alone, with little or no restriction of fluidintake, can require many weeks or even months tobe lethal, depending on the nutritional status of theindividual at the beginning.'

Sedation and Palliative CareRelief of suffering is the oldest and still the most basicelement of medical care. When treatment is directedat symptoms alone and not at the cause of the ill-ness, it is considered palliative care, and is intendedto provide as much comfort as possible. This includesspiritual and personal support for patients and theirfamilies, better home care with adequate help, andattention to all causes of distress and suffering. Thelatter has focused on symptom control, looking atthe indications for medication, the most appropri-ate drugs to use, and the dosage required to relievedifferent types of suffering. There is no question thatmuch of the new interest in palliative care was stimu-lated by the "threat" of legalized assisted dying, forwhich there is strong support from the general pub-lic (about sixty-five percent) and moderate support

within the medical profession (as many as forty per-cent of physicians who care for dying patients).

The cornerstone of good palliative care is ap-propriate medication to relieve all physical andpsychological suffering. Symptoms and distress in-crease towards the end of life in chronic illnesses,and often reach a climax in the final weeks and days.It is not surprising that steps to control these symp-toms must also increase. In many illnesses, thiseventually requires the use of moderate to large dosesof narcotics. The increasing need for narcotics withadvancing symptoms has caused some misunder-standing among physicians. A few years ago theywere hesitant to give enough medication to controlsuffering out of fear of causing tolerance and addic-tion, even though short-term addiction in a persondying of cancer is hardly a threat to the patient or tosociety. More recently, concern about causing ad-diction has faded, only to be replaced by the fear ofkilling the patient by inadvertently giving too muchmedication. Again, many physicians are reluctant togive enough medication to provide even reasonablerelief of symptoms. In fact, the amount of sedationrequired to control symptoms is extremely variablefrom one person to another. Some cancer patientsrequire such enormous doses of narcotics just to con-trol their underlying pain that upper limits on howmuch morphine should be legally permissible aremeaningless.

Terminal Sedation and DehydrationWithin the spectrum of interventions, between re-fusal of treatment and euthanasia, electivedehydration and sedation assume greater importancewhen they are combined. The resulting terminal se-dation usually takes place at home or in a hospiceinstitution. Most acute care hospitals are too single-minded in their mission, too aggressive in theirtreatment, and too conservative in their thinking toaccept such a responsibility. The patient can electdehydration at any point in his illness up to the ter-minal phases, where it may be a function of hisdisease and no longer elective at all. Sedation, usu-ally in the form of narcotics, can be provided inamounts just sufficient to control both the symptomsof his underlying illness and any added discomfortcaused by the dehydration, primarily thirst. Indeed,the medications required to control the symptomsof the underlying disease may cover thirst too,with nothing more needed. The level of sedation is

47

obviously an important factor. The patient who hasvery little underlying discomfort may need only mildsedation to control thirst. For several days he maybe conscious and able to communicate with his fam-ily, helping them to adjust. He will have time toreconsider and even reverse his decision if he wantsto. Conversely, patients with serious underlying painand/or severe anxiety and fear of death may wantenough sedation to render them somnolent most orall of the time. They may accept impending deathand want nothing more than to shorten the dura-tion of their suffering.

It must be recognized that high levels of seda-tion eliminate the desire or even the ability to eat ordrink. This becomes an important issue when thepatient requests total sedation to the point of som-nolence in order to be unable to eat or drink. In otherwords, the process begins with heavy sedation, afterwhich fluids are withdrawn. This precludes any suf-fering from underlying discomfort or thirst. Death,however, results from dehydration, requiring severaldays or more for completion, during which time thedrugs are given in sufficient quantity to maintain thepatient in a state of "chemical oblivion." At this levelof sedation, he certainly is not able to recant his de-cision or reverse the process. Thus, terminal sedationcan utilize enough, and only enough, sedation to re-lieve physical symptoms, a practice that is unlikelyto be challenged. Higher levels of sedation, however,can render the patient semiconscious and unawareof dehydration and the dying process. As we willsee, this is a step with more significant moral andlegal implications.

Between the extremes of minimal and total se-dation there is clearly room for negotiation betweenpatient and physician. The timing of the beginningof sedation is a factor. Sedation could begin on thesame day, at the same time, that food and drink arefirst refused. A more punitive approach would be torequire that the patient undergo one, two, or eventhree days of dehydration before receiving symptom-atic relief. As we have seen, the agreed-upon level ofsedation is clearly a very important variable. In prac-tice, the entire agreement concerning terminaldehydration and sedation is essentially made betweenpatient, family, and physician, and is extremely dif-ficult to regulate or monitor from outside. Thevarious combinations of terminal sedation and de-hydration blend into each other so completely as todefy precise definitions. They are densely tangled with

the concepts of active and passive euthanasia, doubleeffect, and intent.

Active and Passive EuthanasiaA persistent area of confusion is the concept of ac-tive versus passive euthanasia.9 Euthanasia performedby a physician is the standard of active participationin a patient's death. Many people also see thephysician's role in assisted suicide as being an activeone, even though the activity may be limited to writ-ing a prescription without ever knowing whether ornot the prescription was filled or the drugs taken.The term passive euthanasia is applied to all situa-tions where the patient refuses to start or elects todiscontinue any form of treatment. If the treatmentis chemotherapy of marginal value in the face of aprogressing cancer, the issue may not be of greatmoral significance to either the patient or the physi-cian. However, when the request is to discontinue arespirator for a patient who is competent and con-scious, physicians usually see this step as being quiteactive on their part, far more so than simply writinga prescription for sleeping medication. Indeed, thelevel of activity is reflected in how the law wouldreact if the respirator were turned off under identi-cal circumstances by a family member, or anyoneother than a physician.

The basis for passive euthanasia is the concept thatthe patient will die of his underlying disease when thetreatment is stopped. For the cancer patient stoppingchemotherapy, more rapid progression of the diseasemay indeed take place. However, discontinuing artifi-cial respiration is usually expected to cause deathpromptly. Truly, the patient may have been on respira-tory assistance because of pulmonary failure, but hiscondition was well maintained and stable until the ven-tilator was stopped. The patient need not have diedand, indeed, would not have died had the respiratorbeen left in operation. It is stretching reality signifi-cantly to attribute the patient's death directly and onlyto his underlying disease. This is an important pointbecause when a patient whose death is not imminentelects terminal dehydration, his death will be due todehydration and not to his underlying disease. Theo-retically, in dehydration the physician's role can becompletely passive; the act can be carried out by thepatient without any physician involvement at all.

There are even various levels of activity within"passive euthanasia." Not feeding a person in theface of starvation or not giving antibiotics in the face

ARTICLE I Terminal Sedation and Dehydration

48 I Elder's Advisor

of infection, or not starting a respirator in the faceof respiratory failure are deliberate decisions not toact, and to many people seem very passive indeed.On the other hand, discontinuing treatment alreadyin progress, particularly removal of a respirator, isviewed by most as being much more active.

The ambiguity of active versus passive isstraddled by terminal dehydration, with or withoutsedation. Death by dehydration of a person who istoo ill to eat or drink is certainly passive. The deci-sion not to provide artificial hydration and nutritionis only slightly less so. On the other hand, refusal ofhydration by one who is able to eat and drink is anactive decision to end life. While the process of de-hydration to the point of death may seem passive,death is elected and would not occur if food anddrink were continued.

Double EffectMuch of medical treatment is the balancing of po-tential benefits against risks. A simple example ofthis is the risk that a patient may be encouraged totake in open-heart surgery or bone marrow trans-plantation. The physician intends to make the patientbetter, but recognizes that failure could be fatal. Theuse of narcotics to control pain and suffering carriesthe risk of depressing respiration and ending thepatient's life. The limits of this compromise arereached when it is foreseen that life will probably beshortened, so long as the primary intent is to relievesuffering. This time-honored acceptance of risk ac-knowledges the fact that narcotics needed to relievesuffering do have a side effect that may be fatal.

The Doctrine of Double Effect was introducedinto our thinking by St. Thomas Aquinas to ratio-nalize killing in self-defense and in war, twocircumstances where Christ's teaching to turn theother cheek was not politically acceptable. 10 The in-tended good effect of such killing is that I will live,while the unfortunate but foreseen bad effect is thatmy attacker must die. Translated into the typicalmedical situation, there are four requirements to thedouble effect:

1. The drug must be given primarily for its goodeffect, to relieve suffering

2. Death should not be intended, but may be fore-seen and accepted as a consequence of givingthe drug

3. The benefit to the patient (relief of suffering)should outweigh the risk of harm (earlier death)

4. The bad effect, death, must not be required inorder to relieve suffering

Much of the relief of suffering that physicians havebeen able to provide for their patients over the cen-turies has been due to the acceptance of the doubleeffect doctrine by physicians, the church, the public,and the law. Within this acceptance, compassionatephysicians have often pushed the unfortunate butforeseen secondary effect of narcotics when it wasclear that mercy required a prompt and painlessdeath.

The principle of double effect began as a compro-mise between Christian morality and political reality.When applied to human suffering, it provides a moral,religious, and legal loophole by which physicians canmeet the most urgent needs of their patients, as theyhave done for centuries. Extension of the double effectinto the broad area of terminal sedation is also com-monplace, particularly in the last days of life. Througha prior understanding with the patient, in response torequests from the family, or simply out of compassion,physicians often maintain high doses or even increasethe administration of narcotics to shorten the agonalphase of dying. Eating and drinking usually havestopped, and the patient may neither know nor carewhat is happening.

However, there is also the patient who does knowand care what is happening, and who does not wantto witness his own dying. Preliminary sedation canrender him oblivious to his predicament and suffer-ing, at the expense of being able to eat or drink. Wheninterpreted strictly, the principle of double effect doesnot apply to this form of terminal sedation verywell.11 The primary intent in giving the drug is torelieve the symptoms of the underlying disease andof thirst in a patient who will not eat or drink. How-ever, if sufficient sedation is given, the secondaryeffect is that the patient cannot eat or drink. Theamount of sedation needed to relieve most sufferingand thirst may be less than the amount required toproduce somnolence, and usually much less than theamount required to significantly shorten life by theuse of the drugs alone. Dehydration, then, is the pri-mary intent of the patient, and even the initial stepbegun by him, and sedation is given only to relievesymptoms. However, dehydration soon becomes thesecondary, unintended effect of the sedating drugs.Rather than being just possible and perhaps undes-ired, with adequate sedation and fluid restriction,death is inevitable. Thus, terminal sedation and the

ARTICLE I Terminal Sedation and Dehydration

double effect are related in the minds of many people.Although they are not identical, both are venerablepractices that are accepted as components of pallia-tive care and both are considered to be within legaland medical standards.12

IntentThe patient who wishes to shorten his life in orderto avoid unnecessary suffering may feel that he haslittle to lose. Dying seems the lesser of two evils, whencompared to continuing life as it is. Furthermore,the patient may see little difference between stop-ping treatment (including taking fluids) and assistedsuicide or euthanasia. His goal is the same, and onethat he would like to attain as painlessly and rapidlyas possible. Assisted suicide would take a few hoursand may be vastly preferable to dehydration, requir-ing as long as two weeks. Euthanasia, which requiresonly a few minutes, might be preferable to assistedsuicide. Indeed, in Holland, where assisted suicideand euthanasia are both legally accepted, far morepatients request euthanasia, usually at the hands oftheir family doctors, than request assisted suicide.This is clearly the easiest and least stressful approachfrom the patient's point of view.

The desires of the autonomous patient still re-ceive little consideration when pitted against thebarriers of our legislatures and our conservative re-ligious and medical organizations. The rapid growthof interest in biomedical ethics in the past few yearshas focused mostly on physicians and other healthcare professionals, with somewhat less attention tothe desires and needs of patients and their families. 3

Pressure for personal autonomy has gained somerespect from a reluctant and paternalistic medicalprofession, but this definitely does not yet extend toa complete range of choices in end-of-life issues.

It is commonly stated in psychiatry that anyonewho wishes to die is mentally ill, by definition. Cer-tainly, most suicides are due to depression and shouldbe prevented, if at all possible. However, these gen-eralizations are too broad and do not apply to peoplewho are already dying and who wish to shorten theduration of their suffering. My own interviews withterminally ill patients, some only a few days fromdeath, with physicians, and with two psychiatristswho deal frequently with dying patients, support this.Some perfectly sane and rational people would liketo die sooner rather than later, and for very under-standable reasons. Indeed, most physicians who takecare of significant numbers of people with slowly

fatal illnesses know that their patients are perfectlysane and would be amused or insulted to be consid-ered otherwise.

Some conservative religions teach that sufferingat the end of life is an important and desirable as-pect of dying, an opportunity for growth for thepatient who must endure it and for the family whomust witness it. Any thought or attempt to shortenone's suffering and life are the equivalent of suicideand therefore forbidden. The concept of "rationalsuicide," or wishing to die even at the extreme endof life, has no place in such religious teachings. It isaccepted that nothing need be done to prolong life,but stressed that nothing should be done to shortenit. This extreme view is certainly not shared by thegeneral public and is often rejected even by mem-bers of the same conservative religious groups whenthey are faced with their own personal decisions. Atthis point, the preferences of the individual may besubordinated to the dictates of the religion, enforcedby family and the religious institution. Most physi-cians feel that it would be unconscionable to say thatthe person who elects to shorten his life by dehydra-tion must forego all pain medication, regardless ofhis underlying needs. The precedent for sedation iswell established. When a conscious and competentpatient requests that artificial respiration be discon-tinued, the physician is expected to provideappropriate sedation in advance so that the terror ofsuffocating is eliminated.

Deeply embedded in this controversy is thephysician's intent. Here it must be understood thatintent really means willingness to personally acceptresponsibility for the death of the patient as the pos-sible, probable, or even inevitable outcome of theaction taken. In the absence of serious underlyingillness, physicians rarely, if ever, want to kill patientsor to simply help them die. The physician's intent isto try to meet the wishes of the patient in relievingunbearable suffering. The extent to which the physi-cian accepts direct responsibility for ending thepatient's life is a measure of this intent.

It is perfectly acceptable for a patient to intendto die. But it is not all right, under most circum-stances, for a physician to openly intend the deathof a patient. It is well known that some compassion-ate physicians help patients die under appropriatecircumstances, while fully intending to do so, butonly in the presence of severe suffering. This is donein private, with the understanding of the patient and/or family. While it may be publicly acknowledged in

50 I Elder's Advisor

anonymous polls, and in very general terms, specificinstances are rarely mentioned.

The gray zone of physicians' intent is underscoredin any discussion of terminal sedation. Many fac-tors contribute to this: Which comes first, theinitiation of sedation or the initiation of dehydra-tion? If the process begins with the patient's refusalto ingest fluids, when is sedation begun? How muchsedation is required or permissible? (Just enough toallay symptoms, including thirst? Enough to producesomnolence most of the time? Enough to provideoblivion to the dying process throughout its entireduration?) Should sedation be moderated towardsthe end when it may no longer be needed? Howclosely can or should the process be monitored bythe physician? Is the risk of overdosing and pharma-cologically shortening life acceptable?

A physician's intent is often unrecognized becauseit is easily denied or rationalized to meet the circum-stances. The law recognizes this and tries to hold usto our actions, rather than our intentions, intentionsbeing hard to prove or disprove. "The law typicallyholds people responsible for the foreseeable conse-quences of their acts, even if they had no intent tocause those consequences."4 Interestingly, however,the law has chosen not to look very closely at physi-cians' intentions in matters related to double effectand terminal sedation. Indeed, the physicians' ratio-nale is relatively easy: they are giving narcotics tocontrol the suffering of dying patients. The use ofsedation to prevent suffering at the end of life is tra-ditional and essential to good care.

Legal Aspects of Terminal SedationThe legality of terminal sedation is based on twoimportant precedents, and has been strongly en-dorsed by the U.S. Supreme Court. The right to refusemedical treatment is deeply ingrained in the Consti-tution, and relief of suffering is a basic componentof medical care. Justice Brandeis wrote in 1928 that"The makers of our constitution ... conferred, asagainst the government, the right to be let alone-the most comprehensive of rights and the right mostvalued by civilized man.""5 Relatively recently, theconcept was extended to include artificial feeding 6

and normal ingestion of food and drink.17 In sup-port of this, Chief Justice Rehnquist wrote, "We thinkthe distinction between assisting suicide and with-drawing life-sustaining treatment, a distinctionwidely recognized and endorsed in the medical

profession and in our legal traditions, is both impor-tant and logical; it is certainly rational. " 8 Moreimportant, as the U.S. Court of Appeals for the NinthCircuit pointed out, regardless of the patients' ex-pectation or intent, death resulting from refusal oftreatment is not legally considered to be suicide. "Webelieve that there is a strong argument that a deci-sion by a terminally ill patient to hasten by medicalmeans a death that is already in process, should notbe classified as suicide." 9 This is a critical point be-cause, by extension, it means that providing sedationfor such a patient is providing relief from suffering,but it is not assisting suicide.

In 1997, the U.S. Supreme Court unanimouslyrejected a constitutional right to physician-assistedsuicide, overturning the opinions of two CircuitCourts of Appeal.20 However, this rejection of as-sisted suicide was accompanied by very strongstatements concerning palliative care and sedationby Justices O'Connor, Ginsberg, Breyer, Souter, andStevens. Speaking most emphatically for this group,Justice O'Connor said, "[A] patient who is sufferingfrom a terminal illness and experiencing great painhas no legal barriers to obtaining medication, fromqualified physicians, to alleviate that suffering, evento the point of causing unconsciousness and hasten-ing death. ' 21 Justice Breyer supported this, saying,"[T]he laws of New York and Washington do notprohibit doctors from providing patients with drugssufficient to control pain despite the risk that thosedrugs themselves will kill." 22 Professor Robert Burtof Yale Law School notes that "The Supreme Courtmajority has thus provided an unexpected but strongand very welcome directive requiring states to re-move the barriers that their laws and policies imposeon the availability of palliative care. ' 23 By statingthat terminal sedation for symptomatic relief is notassisted suicide, the Court has endorsed an aggres-sive practice of palliation.

Professor David Orentlicher of Indiana Univer-sity School of Law goes beyond this, pointing outthat the extension of palliative care to include termi-nal sedation could create a situation that is morallyand legally close to euthanasia, and well beyond as-sisted suicide.24 Total sedation of an individual priorto stopping of food and fluids renders the patientcomatose and helpless, at the mercy of his physicianfor the remaining days of life, a form of slow eutha-nasia. In this extreme form of terminal sedation,physicians take almost as much responsibility for the

ARTICLE I Terminal Sedation and Dehydration 51

patient's death as they would by giving a lethal in-jection. Orentlicher's purpose is not to condemnterminal sedation, but to point out that the Courtseems to be endorsing a possible form of euthana-sia, while rejecting physician-assisted suicide, whenthe latter may be more desirable for the patient andcarry less risk of abuse.

In the spectrum of end-of-life interventions, ter-minal sedation is an extension of the autonomy ofrefusal of treatment, to which is added sedation toprovide physical and psychological comfort. Likepersonal refusal of treatment, terminal sedation islegal in all states. It is not subject to oversight orregulation. The final decision and all details areworked out by patient, family, and physician, withno requirement for any other input or notification.Assisted suicide, on the other hand, is legal only inone state, Oregon, where it is strictly regulated.

For the competent, conscious patient, a line cur-rently separates terminal sedation, which is legal,from physician-assisted suicide, which is not. Theunconscious or incompetent patient presents a morecomplicated problem. The laws of most states nowpermit a family member or some other responsiblesurrogate to speak for the unconscious patient inrequesting withdrawal of treatment, including arti-ficial feeding. All states respect this if there is strongevidence (a living will, etc.) as to what the patient'spreferences were when well, and many states do notrequire any formal evidence at all. There is no needor place for legal surrogacy in physician-assisted sui-cide, since taking one's own life is a conscious act bydefinition. Similarly, a surrogate wishing to end thelife of an unconscious patient need not consider ter-minal sedation since simple withdrawal of thenecessary artificial feeding will suffice and is legal.

The most complicated issue is the person who istotally incompetent, but not unconscious, as inAlzheimer's disease, the most widely feared of all end-of-life scenarios in this country. A person who isfarsighted enough may certainly request, in a livingwill, that his life be ended by terminal sedation uponreaching some arbitrary stage in the illness (unableto recognize family members, etc.) Leaving the finaldecision to a surrogate, even with a detailed livingwill, pushes such a death into the realm of non-vol-untary euthanasia. The mechanism for inducing thedeath may be some blend of dehydration and seda-tion, since they are "legal." However, the step mustbe taken by other people, be based on the original

living will, and be administered to a person who isnot currently competent to consent to anything andis not even dying. This is already a subject of moraldebate, and will become a legal issue in years to come,but only after we have reached some level of accep-tance of assisted suicide and even of voluntaryeuthanasia. Pressure for this will come from demandsto respect the patient's original autonomy, generalhorror of living for years totally demented and de-pendent, and the economics of caring for largenumbers of elderly people who are so hopelessly ill.Today, about 4.6 million Americans are over the ageof eighty-five and about half of them, 2.4 million,have severe dementia. By the year 2040, both of thesenumbers are expected to double to 12.3 million overthe age of eighty-five, and 5.1 million with severedementia.

Terminal Sedation-Today & TomorrowRecently there has been strong interest in providingbetter care at the end of life. It is recognized that ourdominant medical goal of curing disease distracts usfrom the needs of those who cannot be cured. Manygroups have taken steps to study and provide betterpalliative care, while others search for ways to al-low more dignified and peaceful deaths, free fromunnecessary suffering. Many ethical and moral as-pects of end-of-life care are being examined, includingactive versus passive, double effect, terminal seda-tion and dehydration, and physician-assisted suicide.Society, the law, and the medical profession are try-ing to decide what limits to set on how people maydie, recognizing that some do not want to live thefinal weeks that are necessitated by the natural courseof their illnesses. They do not have many options.As a society, we are not yet ready to legalize assistedsuicide in most states, but even the U.S. SupremeCourt acknowledges that, with more experience, thatday may come. Although the Court does not sup-port individual autonomy to the extent of givingphysician-assisted suicide constitutional protection,it does feel that people who are terminally ill andsuffering significantly should have some meaningfulrecourse to medical help for relief, even if it meansshortening life. The Court clearly leaves the dooropen for terminal sedation.

Some undoubtedly see terminal sedation as astepping stone to physician-assisted suicide and eveneuthanasia, a moral and legal loophole that shouldbe closed. Conservative lawmakers could attack

52 I Elder's Advisor

terminal sedation on the basis of physician intent orimproper use of drugs, if only to curry favor witheven more conservative constituents. The argumentswould be the same as those being used to counterthe Oregon laws permitting physician-assisted sui-cide. Others view terminal sedation as an end in itself,an approach to shortening life that will be accept-able to some patients who are dying and to thegeneral public, a line that can be held against pres-sure for physician-assisted suicide.

I think terminal sedation is indeed a stepping stonetowards physician-assisted suicide, a compromise forthe time being. It seems cruel to ask a patient and afamily to accept a route to death that takes a week ortwo to travel, however painless the journey may be. Ifand when assisted suicide becomes legal, terminal se-dation will not be needed or wanted by very manypeople. However, until our laws become more accom-modating, interest in terminal sedation will grow, aspressure for patient autonomy continues to increase.Blurry as terminal sedation may be in concept, it isaccepted and practiced. Therefore, terminal sedationwill be hard for lawmakers to back away from, andharder still for government agencies to control, moni-tor, or bury under oppressive restrictions.

At the present time, in the mind of the public,

voluntary dehydration and sedation is not an attrac-tive way to die. As such, it appeals to a relativelyfew, unusually strong-willed and well-informedpeople. The advantage to elective terminal sedationis that it is legal and can be done painlessly. The le-gality does not derive from close analysis of the actsunder all circumstances, but from recognition thatmany, if not most, people die naturally of dehydra-tion, and that sedation to control suffering at theend of life is an expected, accepted, and traditionalrole of medicine.

As concern about end-of-life issues increases andassisted suicide remains legally off limits, more andmore emphasis will be placed on terminal sedation.Indeed, legislative rejection of assisted suicide mayprovide a "dam" effect that forces more people toturn to terminal sedation as their only acceptableoption. It can, should, and will be discussed morepublicly, and should certainly enter into conversa-tions between physicians and patients who want toend their suffering sooner. As we continue to exam-ine the needs and desires of dying patients, ourwillingness to provide appropriate medical interven-tion will increase. The needs of people at the end oflife are multiple and varied, and our responses tothem should be, too.

Endnotes1. In 1998 & 1999, 1 had the pleasure of meeting with

several people to discuss various aspects of terminalsedation and dehydration, intending to prepare amanuscript on the subject. Although the manuscriptwas never completed, the discussions were valuableand some of the ideas considered are included here.The other participants were: Sidney H. Wanzer,M.D., Professor Emeritus, Harvard Medical School;Prof. Charles H. Baron, Boston College LawSchool; Garrick E Cole, Esq., Smith, Duggon &Johnson, Boston, Massachusetts; William E Comer,M.D., Coram Healthcare, Portland, Oregon; andProfessor James Vorenberg, Harvard Law School.

2. G.P. Smith, Terminal Sedation as Palliative Care:Revalidating a Right to a Good Death, 7 CAM-BRIDGE Q. HEALTHcAR ETHICS 382-87 (1998).

3. R.J. Sullivan, Accepting Death Without ArtificialNutrition and Hydration, 8 J. GEN. INTERNAL MED.220-24 (1993); P. Schmitz & M. O'Brien, Observa-tions on Nutrition and Hydration in Dying CancerPatients, in By No EXTRAoRDiNARY MEANS: THE

CHOICE TO FORGO LIFE-SUSTAINING FOOD AND WATER

29-38 (J. Lynn ed., 1989); J.C. Ahronheim & M.R.Gasner, The Sloganism of Starvation, 335 LANCET

278-79 (1990).

4. Sullivan, supra note 3, at 220-24.

5. Schmitz & O'Brien, supra note 3, at 29-38.

6. J.A. Billings, Comfort Measures for the TerminallyIll: Is Dehydration Painful?, 33 J. AM. GERIATRIC

Soc'Y 808-10 (1998); J.O. Maillet & D. King,Nutritional Care of the Terminally Ill Adult, 9HOSPICE J. 37-54 (1993).

7. J. Lynn & J.E Childress, Must Patients Always beGiven Food and Water?, 13 HASTINGS CENTER REP.17-21 (1983).

8. MODERN NutrRmON IN HEALTH AND DISEASE (M. Shilset al. eds., 8th ed. 1994).

9. CHARLES E MCKHANN, A TIME TO DE: THE PLACEFOR PHYSICIAN ASSISTANCE 100-02 (1999); see also B.Gert et al., An Alternative to Physician-Assisted

ARTICLE I Terminal Sedation and Dehydration 53

Suicide, in PHYSICIAN ASSISTED SUICIDE: EXPANDING

THE DEBATE ch. 12 (M.P. Battin et al. eds., 1998).

10. Religious and ethical aspects of the rules of doubleeffect are discussed extensively in T.L. BEAUCHAMP

& J.E CHILDRESS, PINCIPLES OF BIOMEDICAL ETHICS(41h ed. 1994).

11. T.E. Quill et al., The Rule of Double Effect-ACritique of Its Role in End-of-Life Decision Mak-ing, 337 NEw ENG. J. MED. 1768-71 (1997).

12. Id.

13. J. Hardwig, Spiritual Issues at the End of Life, 30HASTINGS CENTER REP. 28-30 (2000).

14. D. Orentlicher, The Alleged Distinction BetweenEuthanasia and the Withdrawal of Life-SustainingTreatment: Conceptually Incoherent and Impossibleto Maintain, 3 U. ILL. L. REv. 837-59 (1998).

15. Olmstead v. United States, 277 U.S. 438, 478(1928) (Brandeis, J., dissenting).

16. In re Quinlan, 355 A.2d 647, 662-64 (N.J. 1976);Cruzan v. Director, Missouri Department of Health,497 U.S. 261, 278 (1990).

17. Bouvia v. Superior Court of Los Angeles County,225 Cal. Rptr. 297, 300-01 (Cal. Ct. App. 1986).

18. Vacco v. Quill, 521 U.S. 793, 800 (1997).

19. Compassion in Dying v. Washington, 79 E3d 790,824 (9th Cir. 1996). It should be noted that somereligious views do not accept this.

20. Washington v. Glucksberg, 521 U.S. 702, 735(1997); Vacco at 797.

21. Vacco at 809 (O'Connor, J., concurring).

22. Id. at 810 (Breyer, J., concurring).

23. R.A. Burt, The Supreme Court Speaks: Not AssistedSuicide but a Constitutional Right to PalliativeCare, 337 NEW ENG. J. MED. 1234-36 (1997).

24. D. Orentlicher, The Supreme Court and Physician-Assisted Suicide: Rejecting Assisted Suicide butEmbracing Euthanasia, 337 NEw ENG. J. MED.

1236-39 (1997). See also D. Orentlicher, TheSupreme Court and Terminal Sedation: An Ethi-cally Inferior Alternative to Physician-AssistedSuicide, in PHYSICIAN ASSISTED SUICIDE: EXPANDING

THE DEBATE ch. 18 (M.P. Battin et al. eds., 1998).

ARTICLE I Terminal Sedation and Dehydration 535.t