Wanted, Dead or Alive Frank Chessa, Thomas I. Cochrane, Joan MacGregor, and Kenneth Leeds To the Editor: In “Rethinking the Ethics of Vital Organ Donation” (Nov-Dec 2008), Franklin Miller and Robert Truog offer a provocative argument for overturning status quo justification for medical decision-making at the end of life. They do this in the name of providing a more “honest” justification for the procurement of vital organs for transplant. They correctly point out that incoherencies and practical difficulties involved in diagnosing death have led to an unsatisfactory justification of current practices in organ donation. But their proposed solution ignores equally troublesome issues in obtaining informed consent from patients and surrogates. Miller and Truog argue that when a patient dies after life-sustaining treatment is withdrawn, the withdrawal is the cause of death, and therefore, the physician who orders it has caused the death. This contrasts with the status quo view that withdrawing life-sustaining treatment allows death to occur, but that death itself is caused by the underlying illness. On Miller and Truog’s view, the key element distinguishing a justified causing of death by removal of life-sustaining treatment from criminal homicide is that the patient or surrogate consents to removing life-sustaining treatment. This allows them to argue that causing death by removing vital organs can also be justified by patient or surrogate consent, freeing organ procurement practices from the need to be justified by suspect neurological or cardiovascular criteria for death. The most important turn in this argument is to make patient or surrogate consent the primary gatekeeper for permitting organ procurement, jettisoning the dead donor rule in the process. Unfortunately, there are many reasons to believe that obtaining valid informed consent is at least as problematic as formulating defensible criteria for death. There are three ways to obtain consent to withdraw life-sustaining treatment: (1) a patient with decisional capacity provides contemporaneous consent; (2) a legally authorized surrogate decision-maker provides consent using either the substituted judgment or best interest standard; or (3) the patient has provided an advance directive stating that she does not wish to have her life prolonged in certain situations. The problems with each have been well discussed. The most common approach to determining patient capacity (see Paul Appelbaum and Thomas Grisso, “Assessing Patients’ Capacities to Consent to Treatment,” New England Journal of Medicine vol. 319 (1988): 1635–38) uses four criteria: ability to communicate choice; understanding; ability to manipulate information rationally; and appreciation of the situation and its consequences. These criteria are open to subjective interpretation in all cases but are especially difficult to evaluate in a critically ill person. When communication is limited, for example, to gestures, mouthing words, or pointing to letters on a board, it is difficult to know in any but the broadest terms what a patient really wants. By contrast, an apnea test and clinical exam for brain death are pleasingly unambiguous. If the bar for meeting the criteria is kept high to avoid relying on the responses of persons with compromised capacity, we move to relying on surrogates or advance directives. In a systematic review, David Shalowitz and colleagues (“The Accuracy of Surrogate Decision Makers: A Systematic Review,” Archives of Internal Medicine vol. 166 (2006): 493–97) show that patient-designated and next-of-kin surrogates incorrectly identify patients’ end-of-life wishes one-third of the time, which calls into serious question whether substituted judgment is ethically appropriate. Advance directives are also notoriously difficult to apply in actual situations, when doctors may be hesitant to predict death and thereby activate them. Studies also suggest that persons tend to change their minds about what they would judge an acceptable quality of life as they become sicker, so advance directives may not reflect current wishes. Of course, these problems with consent will be with us whether or not we accept Miller and Truog’s indictment of the current criteria for death. It is tempting to think that when faced with two or more flawed concepts, it is always good practice to eliminate what we can. But in the current instance, if we jettison the dead donor rule because we are less than clear about...