Schneiderman, Jecker, and Jonsen define quantitative futility as the judgment that a treatment has less than a 5% chance of achieving the desired physiologic result.Select one:TrueFalse was not intended for them and they should not betreated even if they were richer than Midas.particular chemotherapy for one type of cancer basedon failures seen when used for treating similar forms ofcancer.) Or one may have accumulated empiric experi-ence insufficient to state precisely the likelihood of suc-cess, but sufficient to doubt the likelihood of success.(For example, physicians have had only a few years ofexperience with a currently popular medication to cure-baldness, but sufficient experience to be dubious of itslong-term success.)Thus, both the quantitative and qualitative aspects offutility are recognized in the most ancient traditions.Hippocrates rejects efforts that are quantitatively orprobabilistically unlikely to achieve a cure; Plato ob-jects to a cure consummating (qualitatively) in a life that"isn't worth living." Both quantitative and qualitativeaspects relate to a single underlying notion: The resultis not commensurate to the effort. The effort is, on thepart of the agent, a repeated expenditure of energy thatis consistently nonproductive or, if productive, its out-come is far inferior to that intended.Reports of one or two "miraculous" successes do notcounter the notion of futility, if these successes wereachieved against a background of hundreds or thou-sands of failures. Such rare exceptions are causallyinexplicable, because any clinical situation contains amultitude of factors-in addition to treatment-that mightaffect outcome. As Wanzer and colleagues (10) stated,"The rare report of a patient with a similar conditionwho survived is not an overriding reason to continueaggressive treatment."Defining FutilityQuantitative and Qualitative AspectsThe futility of a particular treatment may be evidentin either quantitative or qualitative terms. That is, fu-tility may refer to an improbability or unlikelihood of anevent happening, an expression that is quasi-numeric,or to the quality of the event that treatment wouldproduce. Thus, determining futility resembles using de-cision analysis-with one important distinction. In deci-sion analysis, the decision to use a procedure is basedon the joint product of the probability of success andthe quality (utility) of the outcome (11). Thus, very lowprobability might be balanced by very high utility. Inour proposal of futility, however, we treat the quanti-tative and qualitative aspects as independent thresholds,as minimal cutoff levels, either of which frees the phy-sician from the obligation to offer a medical treatment.We propose that, on the basis of these consider-ations, the noun "futility" and the adjective "futile" beused to describe any effort to achieve a result that ispossible but that reasoning or experience suggests ishighly improbable and that cannot be systematicallyproduced. The phrase, "highly improbable," impliesthat a statistical statement about probability might beapplicable. In the strict sense, such a statement cannotbe made, as proper conditions for determining probabil-ity (that is, prospective comparisons of precisely con-trolled treatment and nontreatment on identicallymatched subjects) will never be present. We introducethe concept of "systematic" to point out that if a rare"success" is not explicable or cannot be predictablyrepeated, its causality is dubious, because it is uncertainwhether treatment, some extraneous influence, or ran-dom variation caused the result.Quantitative AspectsThis independence of futility determinants can betraced back to medical antiquity (12, 13). The percep-tion of futility derived from the Hippocratic corpus.might be considered, in modern terms, to be quantita-tive or probabilistic. A book titled "The Art" (14) en-joins physicians to acknowledge when efforts will prob-ably fail: "Whenever therefore a man suffers from an illwhich is too strong for the means at the disposal ofmedicine, he surely must not even expect that it can beovercome by medicine." The writer further admonishesthe physician that to attempt futile treatment is to dis-play an ignorance which is "allied to madness."In keeping with the quantitative approach to futility,we propose that when physicians conclude (eitherthrough personal experience, experiences shared withcolleagues, or consideration of reported empiric data)that in the last 100 cases, a medical treatment has beenuseless, they should regard that treatment as futile.Technically, we cannot say that observing no successesin 100 trials means that the treatment never works.However, such an observation serves as a point esti-mate of the probability of treatment success. Althoughwe cannot say with certainty that the point estimate iscorrect, statistical methods can be used to estimate arange of values that include the true success rate with aspecified probability. For example, if there have beenno successes in 100 consecutive cases, the clinician canbe 95% confident that no more than 3 successes wouldoccur in each 100 comparable trials (3 successes per 100trials is the upper limit of the 95% CI). This confidencerange would narrow as the number of observations in-creased. If no successes were seen in 200 cases, theupper limit of the 95% CI would be 1.5 successes per100 cases and, for no successes in 1000 observations,the upper limit would be approximately 0.3 successesper 100 cases. In practical terms, because data fromcontrolled clinical trials can only rarely be called on andapplied to a specific case, practitioners usually use theirPlato's Republic (15), on the other hand, has a qual-itative notion of futility, one that emphasizes the inap-propriateness of efforts that result in patients surviving,but leading literally useless lives. According to Plato,the kind of medicine "which pampers the disease" wasnot used by the Asclepian physicians:Asclepius taught medicine for those who werehealthy in their nature but were suffering from aspecific disease; he rid them of it... then orderedthem to live as usual.... For those however,whose bodies were always in a state of inner sick-ness he did not attempt to prescribe a regimen ...to make their life a prolonged miseryMedicine15 June 1990. Annals of Internal Medicine Volume 112. Number 12 951tion of such miraculous recoveries (which have re-sulted, moreover, in incapacitating mental impairmentor total dependence) (28). This point bears on the fre-quently heard excuse for pushing ahead with futile ther-apies: "It is only by so doing that progress is made andthe once futile becomes efficacious. Remember the fu-tility of treating childhood leukemia or Hodgkin lym-phoma." These statements hide a fallacy. It is not.through repeated futility that progress is made, butthrough careful analysis of the elements of the "futilecase," followed by well designed studies, that advances.knowledge. We also point out that our proposal is in-tended for recognized illness in the acute clinical set-ting. It does not apply to preventive treatments, such asimmunizations, estrogen prophylaxis for hip fractures,or penicillin prophylaxis for rheumatic heart disease andinfectious endocarditis, all of which appear to havelower rates of efficacy because they are purposely ad-ministered to large groups of persons, many of whomwill never be at risk for or identified with the particulardiseases that their treatments are intended to prevent.Qualitative AspectsIn keeping with the qualitative notion of futility wepropose that any treatment that merely preserves per-manent unconsciousness or that fails to end total de-pendence on intensive medical care should be regardedas nonbeneficial and, therefore, futile. We do not regardfutility as "an elusive concept" (2). It is elusive onlywhen effects on the patient are confused with benefits.to the patient or when the term is stretched to includeeither considerations of 5-year survival in patients withcancer (not at all pertinent to the notion of futility) orthe "symbolic" value to society of treating handicappednewborns or patients in persistent vegetative state(which rides roughshod over patient-centered decisionmaking) (2).Here is the crux of the matter. If futility is qualita-tive, why should the patient not always decide whetherthe quality achieved is satisfactory or not? Why shouldqualitatively "futile" results not be offered to the pa-tient as an option? We believe a distinction is in order.Some qualitatively poor results should indeed be thepatient's option, and the patient should know that theymay be attainable. We believe, however, that other.sorts of qualitatively poor results fall outside the rangeof the patient's autonomy and need not be offered asoptions. The clearest of these qualitatively poor resultsis continued biologic life without conscious autonomy.The patient has no right to be sustained in a state inwhich he or she has no purpose other than mere vege-tative survival; the physician has no obligation to offerthis option or services to achieve it. Other qualitativelypoor results are conditions requiring constant monitor-ing, ventilatory support, and intensive care nursing(such as in the example at the beginning of our paper).or conditions associated with overwhelmingly sufferingfor a predictably brief time. Admittedly, these kinds ofcases fall along a continuum, and there are well knownexamples of the most remarkable achievements of lifegoals despite the most burdensome handicaps. How-ever, if survival requires the patient's entire preoccupa-extended experience as the source of their conclusions.Here, speciality practice contributes an essential ele-ment; for example, an intensive care pulmonary special-ist who sees several hundred patients who have similardisease conditions and receive similar therapy can oftengroup together "futility characteristics" better than atgeneralist who does not see cases in so focused a man-ner.Without systematic knowledge of the various factorsthat cause a therapy to have less than a 1% chance ofsuccess-knowledge that would allow the physician toaddress these factors-we regard it as unreasonable torequire that the physicians offer such therapy. To do soforces the physician to offer any therapy that may haveseemed to work or that may conceivably work. In ef-fect, it obligates the physician to offer a placebo. Onlywhen empirically observed (though not understood) out-comes rise to a level higher than that expected by anyplacebo effect (16), can a specific therapy be consideredto be "possibly helpful" in rare or occasional cases andits appropriateness evaluated according to rules of de-cision analysis. In the clinical setting, such judgmentsalso would be influenced, of course, by consideringsuch tradeoffs as how cheap and simple the interventionis and how serious or potentially fatal the disease (seeExceptions and Cautions).Although our proposed selection of proportions ofsuccess is admittedly arbitrary, it seems to comportreasonably well with ideas actually held by physicians.For example, Murphy and colleagues (17) invoked thenotion of futility in their series of patients when survivalafter cardiopulmonary resuscitation was no better than2% (upper limit of 95% CI as calculated by authors),and Lantos and colleagues (18) when survival was nobetter than 7% (upper limit of 95% CI as calculated byauthors).Obviously, as medical data on specific situations aregathered under appropriate experimental conditions,empiric uncertainty can be replaced with empiric con-fidence (19). Admittedly, some disorders may be too.rare to provide sufficient experience for a confidentjudgment of futility, even when efforts are made to pooldata. We acknowledge this difficulty but adhere to ourconservative standard to prevent arbitrary abuse ofpower. In judging futility, as in other matters, physi-cians should admit uncertainty rather than impose un-substantiated claims of certainty. Therefore, our viewof futility should be considered as encouraging ratherthan opposing well conducted clinical trials. Importantexamples of such work in progress include studies ofsurvival after cardiopulmonary resuscitation (17-24) anduse of prognostic measures in patients requiring inten-sive medical care (25, 26).Already, data on burn patients (27) and on patients inpersistent vegetative state with abnormal neurooph-thalmic signs (28) are sufficient to help with decisionmaking. The latter group of patients present a particularchallenge to presently confused notions of futility, per-haps accounting in part for why an estimated 5000 to10 000 patients in persistent vegetative state are nowbeing maintained in medical institutions (29). The myth-ologic power of the coma patient who "wakes up".apparently overrides the rarity of documented confirma-952 15 June 1990 Annals of Internal Medicine Volume 112 Number 12 tion with intensive medical treatment, to the extent thathe or she cannot achieve any other life goals (thusobviating the goal of medical care), the treatment iseffective but not beneficial; it need not be offered to thepatient, and the patient's family has no right to demandit.Specifically excluded from our concept of futility ismedical care for patients for whom such care offers theopportunity to achieve life goals, however limited.Thus, patients whose illnesses are severe enough torequire frequent hospitalization, patients confined tonursing homes, or patients with severe physical or men-tal handicaps are not, in themselves, objects of futiletreatments. Such patients (or their surrogates) have theright to receive or reject any medical treatment accord-ing to their own perceptions of benefits compared withburdens.Some observers might object, as a matter of princi-ple, to excluding patient input from assessments ofqualitative futility. Others might be concerned that suchexclusion invites abuse, neglect, and a retreat to thepaternalistic "silent world" of the past in which doctorsavoided communication with their patients (30). In re-sponse to the latter objection, we acknowledge thatpotential for abuse is present and share this concern.We would deplore the use of our proposal to excusephysicians from engaging patients in ongoing informeddialogue. Nonetheless, the alternative is also subject toabuse (for example, when legal threats made by patientsand surrogates cow hospitals into providing excessivecare). We reiterate that the distinction between medicalbenefit and effect justifies excluding patients from de-termination of qualitative futility. Physicians are re-quired only to provide medical benefits to patients. Phy-sicians are permitted, but not obligated, to offer other.non-medical benefits. For example, a physician is notobligated to keep a patient alive in an irreversible veg-etative state, because doing so does not medically ben-efit the patient. However, as noted below, a physicianmay do so on compassionate grounds, when temporarycontinuance of biologic life achieves goals of the patientor family.Exceptions and CautionsWe have attempted to provide a working definition offutility. We also have drawn attention to the ethicalnotion that futility is a professional judgment that takesprecedence over patient autonomy and permits physi-cians to withhold or withdraw care deemed to be inap-propriate without subjecting such a decision to patientapproval. Thus, we regard our proposal as representingthe ordinary duties of physicians, duties that are appli-cable where there is medical agreement that the de-scribed standards of futility are met. We recognize,however, that the physician's duty to serve the bestinterests of the patient may require that exceptions toour approaches be made under special circumstances.An exception could well be made out of compassionfor the patient with terminal metastatic cancer who re-quests resuscitation in the event of cardiac arrest tosurvive long enough to see a son or daughter who hasnot yet arrived from afar to pay last respects. Such anexception could also be justified to facilitate coping andgrieving by family members, a goal the patient mightsupport (32-36). Although resuscitation may be clearlyfutile (that is, would keep the patient alive in the inten-sive care unit for only 1 or 2 more days), complyingwith the patient's wishes would be appropriate, pro-vided such exceptions do not impose undue burdens onother patients, health care providers, and the institu-tion, by directly threatening the health care of others.We hasten to add, however, that our notion of futilitydoes not arise from considerations of scarce resources.Arguments for limiting treatments on grounds of re-source allocation should proceed by an entirely differentroute and with great caution in our present open systemof medical care, as there is no universally acceptedvalue system for allocation (31) and no guarantee thatany limits a physician imposes on his or her patientswill be equitably shared by other physicians and pa-tients in the same circumstances (37, 38).Admittedly, in cases in which treatment has begunalready, there may be an emotional bias to continue,rather than withdraw, futile measures (10). If greaterattention is paid at the outset to indicating futile treat-ments, these situations would occur less frequently;however, the futility of a given treatment may not be-come clear until it has been implemented. We submitthat physicians are entitled to cease futile measures insuch cases, but should do so in a manner sensitive tothe emotional investments and concerns of caretakers.What if a hospitalized patient with advanced cancerdemands a certain medication (for example, a particularvitamin), a treatment that the physician believes to befutile? Several aspects of this demand support its over-riding the physician's invocation of futility. Certaindeath is expected and, although an objective goal suchas saving the patient's life or even releasing the patientfrom the hospital might be unachievable, the subjectivegoal of patient well-being might be enhanced (a placebo-induced benefit). In this particular situation, the effortand resources invested to achieve this goal impose anegligible burden on the health care system and do notthreaten the health care of others. Thus, although phy-sicians are not obligated to offer a placebo, they occa-sionally do. For example, Imbus and Zawacki (27) al-lowed burn patients to opt for treatment even whensurvival was unprecedented. In this clinical situation,compassionate yielding imposes no undue burden, be-cause survival with or without treatment is measured indays. In contrast, yielding to a surrogate's demand forunlimited life-support for a patient in persistent vegeta-tive state may lead to decades of institutional care.Acknowledgments: The authors thank two anonymous reviewers andRobert M. Kaplan, PhD, for their helpful comments.Requests for Reprints: Lawrence J. Schneiderman, MD, University ofCalifornia, San Diego. School of Medicine, Department of Communityand Family Medicine, M-022, La Jolla, CA 92093.Current Author Addresses: Dr. Schneiderman: University of California.San Diego, School of Medicine. Department of Community and FamilyMedicine. M-022. La Jolla. CA 92093.Drs. Jecker and Jonsen: School of Medicine, SB-20, University ofWashington, Seattle, WA 98195.

Question

Schneiderman, Jecker, and Jonsen define quantitative futility as the judgment that a treatment has less than a 5% chance of achieving the desired physiologic result.Select one:TrueFalse was not intended for them and they should not betreated even if they were richer than Midas.particular chemotherapy for one type of cancer basedon failures seen when used for treating similar forms ofcancer.) Or one may have accumulated empiric experi-ence insufficient to state precisely the likelihood of suc-cess, but sufficient to doubt the likelihood of success.(For example, physicians have had only a few years ofexperience with a currently popular medication to cure-baldness, but sufficient experience to be dubious of itslong-term success.)Thus, both the quantitative and qualitative aspects offutility are recognized in the most ancient traditions.Hippocrates rejects efforts that are quantitatively orprobabilistically unlikely to achieve a cure; Plato ob-jects to a cure consummating (qualitatively) in a life that&#34;isn't worth living.&#34; Both quantitative and qualitativeaspects relate to a single underlying notion: The resultis not commensurate to the effort. The effort is, on thepart of the agent, a repeated expenditure of energy thatis consistently nonproductive or, if productive, its out-come is far inferior to that intended.Reports of one or two &#34;miraculous&#34; successes do notcounter the notion of futility, if these successes wereachieved against a background of hundreds or thou-sands of failures. Such rare exceptions are causallyinexplicable, because any clinical situation contains amultitude of factors-in addition to treatment-that mightaffect outcome. As Wanzer and colleagues (10) stated,&#34;The rare report of a patient with a similar conditionwho survived is not an overriding reason to continueaggressive treatment.&#34;Defining FutilityQuantitative and Qualitative AspectsThe futility of a particular treatment may be evidentin either quantitative or qualitative terms. That is, fu-tility may refer to an improbability or unlikelihood of anevent happening, an expression that is quasi-numeric,or to the quality of the event that treatment wouldproduce. Thus, determining futility resembles using de-cision analysis-with one important distinction. In deci-sion analysis, the decision to use a procedure is basedon the joint product of the probability of success andthe quality (utility) of the outcome (11). Thus, very lowprobability might be balanced by very high utility. Inour proposal of futility, however, we treat the quanti-tative and qualitative aspects as independent thresholds,as minimal cutoff levels, either of which frees the phy-sician from the obligation to offer a medical treatment.We propose that, on the basis of these consider-ations, the noun &#34;futility&#34; and the adjective &#34;futile&#34; beused to describe any effort to achieve a result that ispossible but that reasoning or experience suggests ishighly improbable and that cannot be systematicallyproduced. The phrase, &#34;highly improbable,&#34; impliesthat a statistical statement about probability might beapplicable. In the strict sense, such a statement cannotbe made, as proper conditions for determining probabil-ity (that is, prospective comparisons of precisely con-trolled treatment and nontreatment on identicallymatched subjects) will never be present. We introducethe concept of &#34;systematic&#34; to point out that if a rare&#34;success&#34; is not explicable or cannot be predictablyrepeated, its causality is dubious, because it is uncertainwhether treatment, some extraneous influence, or ran-dom variation caused the result.Quantitative AspectsThis independence of futility determinants can betraced back to medical antiquity (12, 13). The percep-tion of futility derived from the Hippocratic corpus.might be considered, in modern terms, to be quantita-tive or probabilistic. A book titled &#34;The Art&#34; (14) en-joins physicians to acknowledge when efforts will prob-ably fail: &#34;Whenever therefore a man suffers from an illwhich is too strong for the means at the disposal ofmedicine, he surely must not even expect that it can beovercome by medicine.&#34; The writer further admonishesthe physician that to attempt futile treatment is to dis-play an ignorance which is &#34;allied to madness.&#34;In keeping with the quantitative approach to futility,we propose that when physicians conclude (eitherthrough personal experience, experiences shared withcolleagues, or consideration of reported empiric data)that in the last 100 cases, a medical treatment has beenuseless, they should regard that treatment as futile.Technically, we cannot say that observing no successesin 100 trials means that the treatment never works.However, such an observation serves as a point esti-mate of the probability of treatment success. Althoughwe cannot say with certainty that the point estimate iscorrect, statistical methods can be used to estimate arange of values that include the true success rate with aspecified probability. For example, if there have beenno successes in 100 consecutive cases, the clinician canbe 95% confident that no more than 3 successes wouldoccur in each 100 comparable trials (3 successes per 100trials is the upper limit of the 95% CI). This confidencerange would narrow as the number of observations in-creased. If no successes were seen in 200 cases, theupper limit of the 95% CI would be 1.5 successes per100 cases and, for no successes in 1000 observations,the upper limit would be approximately 0.3 successesper 100 cases. In practical terms, because data fromcontrolled clinical trials can only rarely be called on andapplied to a specific case, practitioners usually use theirPlato's Republic (15), on the other hand, has a qual-itative notion of futility, one that emphasizes the inap-propriateness of efforts that result in patients surviving,but leading literally useless lives. According to Plato,the kind of medicine &#34;which pampers the disease&#34; wasnot used by the Asclepian physicians:Asclepius taught medicine for those who werehealthy in their nature but were suffering from aspecific disease; he rid them of it... then orderedthem to live as usual.... For those however,whose bodies were always in a state of inner sick-ness he did not attempt to prescribe a regimen ...to make their life a prolonged miseryMedicine15 June 1990. Annals of Internal Medicine Volume 112. Number 12 951tion of such miraculous recoveries (which have re-sulted, moreover, in incapacitating mental impairmentor total dependence) (28). This point bears on the fre-quently heard excuse for pushing ahead with futile ther-apies: &#34;It is only by so doing that progress is made andthe once futile becomes efficacious. Remember the fu-tility of treating childhood leukemia or Hodgkin lym-phoma.&#34; These statements hide a fallacy. It is not.through repeated futility that progress is made, butthrough careful analysis of the elements of the &#34;futilecase,&#34; followed by well designed studies, that advances.knowledge. We also point out that our proposal is in-tended for recognized illness in the acute clinical set-ting. It does not apply to preventive treatments, such asimmunizations, estrogen prophylaxis for hip fractures,or penicillin prophylaxis for rheumatic heart disease andinfectious endocarditis, all of which appear to havelower rates of efficacy because they are purposely ad-ministered to large groups of persons, many of whomwill never be at risk for or identified with the particulardiseases that their treatments are intended to prevent.Qualitative AspectsIn keeping with the qualitative notion of futility wepropose that any treatment that merely preserves per-manent unconsciousness or that fails to end total de-pendence on intensive medical care should be regardedas nonbeneficial and, therefore, futile. We do not regardfutility as &#34;an elusive concept&#34; (2). It is elusive onlywhen effects on the patient are confused with benefits.to the patient or when the term is stretched to includeeither considerations of 5-year survival in patients withcancer (not at all pertinent to the notion of futility) orthe &#34;symbolic&#34; value to society of treating handicappednewborns or patients in persistent vegetative state(which rides roughshod over patient-centered decisionmaking) (2).Here is the crux of the matter. If futility is qualita-tive, why should the patient not always decide whetherthe quality achieved is satisfactory or not? Why shouldqualitatively &#34;futile&#34; results not be offered to the pa-tient as an option? We believe a distinction is in order.Some qualitatively poor results should indeed be thepatient's option, and the patient should know that theymay be attainable. We believe, however, that other.sorts of qualitatively poor results fall outside the rangeof the patient's autonomy and need not be offered asoptions. The clearest of these qualitatively poor resultsis continued biologic life without conscious autonomy.The patient has no right to be sustained in a state inwhich he or she has no purpose other than mere vege-tative survival; the physician has no obligation to offerthis option or services to achieve it. Other qualitativelypoor results are conditions requiring constant monitor-ing, ventilatory support, and intensive care nursing(such as in the example at the beginning of our paper).or conditions associated with overwhelmingly sufferingfor a predictably brief time. Admittedly, these kinds ofcases fall along a continuum, and there are well knownexamples of the most remarkable achievements of lifegoals despite the most burdensome handicaps. How-ever, if survival requires the patient's entire preoccupa-extended experience as the source of their conclusions.Here, speciality practice contributes an essential ele-ment; for example, an intensive care pulmonary special-ist who sees several hundred patients who have similardisease conditions and receive similar therapy can oftengroup together &#34;futility characteristics&#34; better than atgeneralist who does not see cases in so focused a man-ner.Without systematic knowledge of the various factorsthat cause a therapy to have less than a 1% chance ofsuccess-knowledge that would allow the physician toaddress these factors-we regard it as unreasonable torequire that the physicians offer such therapy. To do soforces the physician to offer any therapy that may haveseemed to work or that may conceivably work. In ef-fect, it obligates the physician to offer a placebo. Onlywhen empirically observed (though not understood) out-comes rise to a level higher than that expected by anyplacebo effect (16), can a specific therapy be consideredto be &#34;possibly helpful&#34; in rare or occasional cases andits appropriateness evaluated according to rules of de-cision analysis. In the clinical setting, such judgmentsalso would be influenced, of course, by consideringsuch tradeoffs as how cheap and simple the interventionis and how serious or potentially fatal the disease (seeExceptions and Cautions).Although our proposed selection of proportions ofsuccess is admittedly arbitrary, it seems to comportreasonably well with ideas actually held by physicians.For example, Murphy and colleagues (17) invoked thenotion of futility in their series of patients when survivalafter cardiopulmonary resuscitation was no better than2% (upper limit of 95% CI as calculated by authors),and Lantos and colleagues (18) when survival was nobetter than 7% (upper limit of 95% CI as calculated byauthors).Obviously, as medical data on specific situations aregathered under appropriate experimental conditions,empiric uncertainty can be replaced with empiric con-fidence (19). Admittedly, some disorders may be too.rare to provide sufficient experience for a confidentjudgment of futility, even when efforts are made to pooldata. We acknowledge this difficulty but adhere to ourconservative standard to prevent arbitrary abuse ofpower. In judging futility, as in other matters, physi-cians should admit uncertainty rather than impose un-substantiated claims of certainty. Therefore, our viewof futility should be considered as encouraging ratherthan opposing well conducted clinical trials. Importantexamples of such work in progress include studies ofsurvival after cardiopulmonary resuscitation (17-24) anduse of prognostic measures in patients requiring inten-sive medical care (25, 26).Already, data on burn patients (27) and on patients inpersistent vegetative state with abnormal neurooph-thalmic signs (28) are sufficient to help with decisionmaking. The latter group of patients present a particularchallenge to presently confused notions of futility, per-haps accounting in part for why an estimated 5000 to10 000 patients in persistent vegetative state are nowbeing maintained in medical institutions (29). The myth-ologic power of the coma patient who &#34;wakes up&#34;.apparently overrides the rarity of documented confirma-952 15 June 1990 Annals of Internal Medicine Volume 112 Number 12 tion with intensive medical treatment, to the extent thathe or she cannot achieve any other life goals (thusobviating the goal of medical care), the treatment iseffective but not beneficial; it need not be offered to thepatient, and the patient's family has no right to demandit.Specifically excluded from our concept of futility ismedical care for patients for whom such care offers theopportunity to achieve life goals, however limited.Thus, patients whose illnesses are severe enough torequire frequent hospitalization, patients confined tonursing homes, or patients with severe physical or men-tal handicaps are not, in themselves, objects of futiletreatments. Such patients (or their surrogates) have theright to receive or reject any medical treatment accord-ing to their own perceptions of benefits compared withburdens.Some observers might object, as a matter of princi-ple, to excluding patient input from assessments ofqualitative futility. Others might be concerned that suchexclusion invites abuse, neglect, and a retreat to thepaternalistic &#34;silent world&#34; of the past in which doctorsavoided communication with their patients (30). In re-sponse to the latter objection, we acknowledge thatpotential for abuse is present and share this concern.We would deplore the use of our proposal to excusephysicians from engaging patients in ongoing informeddialogue. Nonetheless, the alternative is also subject toabuse (for example, when legal threats made by patientsand surrogates cow hospitals into providing excessivecare). We reiterate that the distinction between medicalbenefit and effect justifies excluding patients from de-termination of qualitative futility. Physicians are re-quired only to provide medical benefits to patients. Phy-sicians are permitted, but not obligated, to offer other.non-medical benefits. For example, a physician is notobligated to keep a patient alive in an irreversible veg-etative state, because doing so does not medically ben-efit the patient. However, as noted below, a physicianmay do so on compassionate grounds, when temporarycontinuance of biologic life achieves goals of the patientor family.Exceptions and CautionsWe have attempted to provide a working definition offutility. We also have drawn attention to the ethicalnotion that futility is a professional judgment that takesprecedence over patient autonomy and permits physi-cians to withhold or withdraw care deemed to be inap-propriate without subjecting such a decision to patientapproval. Thus, we regard our proposal as representingthe ordinary duties of physicians, duties that are appli-cable where there is medical agreement that the de-scribed standards of futility are met. We recognize,however, that the physician's duty to serve the bestinterests of the patient may require that exceptions toour approaches be made under special circumstances.An exception could well be made out of compassionfor the patient with terminal metastatic cancer who re-quests resuscitation in the event of cardiac arrest tosurvive long enough to see a son or daughter who hasnot yet arrived from afar to pay last respects. Such anexception could also be justified to facilitate coping andgrieving by family members, a goal the patient mightsupport (32-36). Although resuscitation may be clearlyfutile (that is, would keep the patient alive in the inten-sive care unit for only 1 or 2 more days), complyingwith the patient's wishes would be appropriate, pro-vided such exceptions do not impose undue burdens onother patients, health care providers, and the institu-tion, by directly threatening the health care of others.We hasten to add, however, that our notion of futilitydoes not arise from considerations of scarce resources.Arguments for limiting treatments on grounds of re-source allocation should proceed by an entirely differentroute and with great caution in our present open systemof medical care, as there is no universally acceptedvalue system for allocation (31) and no guarantee thatany limits a physician imposes on his or her patientswill be equitably shared by other physicians and pa-tients in the same circumstances (37, 38).Admittedly, in cases in which treatment has begunalready, there may be an emotional bias to continue,rather than withdraw, futile measures (10). If greaterattention is paid at the outset to indicating futile treat-ments, these situations would occur less frequently;however, the futility of a given treatment may not be-come clear until it has been implemented. We submitthat physicians are entitled to cease futile measures insuch cases, but should do so in a manner sensitive tothe emotional investments and concerns of caretakers.What if a hospitalized patient with advanced cancerdemands a certain medication (for example, a particularvitamin), a treatment that the physician believes to befutile? Several aspects of this demand support its over-riding the physician's invocation of futility. Certaindeath is expected and, although an objective goal suchas saving the patient's life or even releasing the patientfrom the hospital might be unachievable, the subjectivegoal of patient well-being might be enhanced (a placebo-induced benefit). In this particular situation, the effortand resources invested to achieve this goal impose anegligible burden on the health care system and do notthreaten the health care of others. Thus, although phy-sicians are not obligated to offer a placebo, they occa-sionally do. For example, Imbus and Zawacki (27) al-lowed burn patients to opt for treatment even whensurvival was unprecedented. In this clinical situation,compassionate yielding imposes no undue burden, be-cause survival with or without treatment is measured indays. In contrast, yielding to a surrogate's demand forunlimited life-support for a patient in persistent vegeta-tive state may lead to decades of institutional care.Acknowledgments: The authors thank two anonymous reviewers andRobert M. Kaplan, PhD, for their helpful comments.Requests for Reprints: Lawrence J. Schneiderman, MD, University ofCalifornia, San Diego. School of Medicine, Department of Communityand Family Medicine, M-022, La Jolla, CA 92093.Current Author Addresses: Dr. Schneiderman: University of California.San Diego, School of Medicine. Department of Community and FamilyMedicine. M-022. La Jolla. CA 92093.Drs. Jecker and Jonsen: School of Medicine, SB-20, University ofWashington, Seattle, WA 98195.

Accepted Answer

Quantitative futility is termed when a medical treatment provided shows a very less likely chance…

define quantitative futility as the judgment that a treatment has less than a 5% chance of achieving the desired physiologic result. Select one: True False