Wilkinson & Savulescu on Medical Futility

Last month, I blogged about the early publication of "Knowing When to Stop: Futility in the ICU" by Dominic Wilkinson and Julian Savulescu.  The final article has been published and is (apparently because of its funding source) freely available on the Journal site.  This is primarily a review article.  As such, it is first-rate.  It succinctly summarizes the history and core issues surrounding medical futility (as well as citing several articles by Professor Pope).  



Still, I have one concern.  The authors write that there are two basic justifications for unilateral refusal:  (1) when further treatment is not in the patient's interests, and (2) when further treatment is harmful to others.  With respect to the first reason, the authors repeatedly refer to this as a "paternalistic judgment" about the value of treatment and life.  Yes, futility seems analogous to paternalism because empowering physicians to unilaterally refuse life-sustaining treatment brings us back to the strong-physician, weak-patient model of the 1950s.  But I do not think that providers hardly ever make a "paternalistic judgment" in refusing treatment either because it is unwanted by the patient or because it is not in her best interests.   



Calling the unilateral refusal a "paternalistic judgment" seems inaccurate on three levels.  First, unilateral refusals do not necessarily involve any paternalism.  Second, they actually do not usually  involve paternalism.  It is not the autonomy of the patient that is overridden but rather the decision of the surrogate.  The surrogate is often not an accurate spokesperson for the patient.  Therefore, overriding the surrogate actually promotes, not undermines, the patient's autonomy.  Third, the authors themselves do not advocate overriding patient autonomy.  If the patient herself really wants the treatment, the authors (following Truog) argue that physicians should provide it (unless the harm-to-others rationale obtains).  Ultimately, the patient's autonomy is respected.  Rarely do we have robust evidence of the patient's preferences applicable to her present circumstances and decide to override those.