Futility III
Michael A. Gillette, Ph.D.

The topic that has come under discussion for the last two months here in 'Practical Ethics' is that of futility. The discussion has surrounded various issues concerning medical care that might be considered futile, and it has become clear that a succinct definition of the term 'futility' itself is now in order. What might a reasonable definition of 'futile' be? The most obvious and common first attempt at a definition looks like this:

Definition One:
Provider X employing treatment Y to patient Z is futile if and only if: Y does not appreciably alter the course of Z's illness

I ask the reader to excuse the excessively philosophical form of this first definition, but it is imperative to develop definitions that are perfectly clear and concise. This first definition merely states, in plain English, that we should consider any type of treatment futile just in case it does not change the outcome of a disease process. This first definition seems reasonable because it cuts right to the central point of the concept of futility. Something is futile if it does not work.

On closer examination, however, definition one begins to falter. Counter examples to definition one are numerous. Consider the case of any palliative measure that is designed to alleviate patient discomfort but not designed to reverse a disease process. It would be difficult to maintain that the provision of palliative care to a terminally ill patient is futile, and yet that is what definition one entails. Surely some good comes from providing such care even if the patient's life is not saved.

Definition Two:
Provider X employing treatment Y to patient Z is futile if and only if: Y does not have the intended physiological effect on Z

Definition two has an advantage over definition one in so far as the usefulness of medical interventions are not restricted the reversal of disease processes. According to definition two, it is not necessary that a patient be cured in order for therapy to be productive, it is only necessary that the treatment in question bring about the intended physiological result. If a drug is intended to operate as an analgesic, then if that drug has the intended impact on the recipient's nervous system, the drug will have done what we intended for it to do and will not, therefore, be futile.

This definition follows the same basic approach as the first. Treatments that do not do what we want them to do are not working, and are therefore futile.

Definition two is not without its own problems, however. Take for example the use of placebos. A headache might be relieved symptomatically just because I am convinced that I have taken a powerful medication, even though the physiological mechanism for the relief of my pain is unknown. Surely there are times when we can achieve results that improve individual's behavioral or symptomatic situation without using physiological manipulation. If this is true, then definition two seems to needlessly attached to physiology. The concept of futility should be outcome oriented and not process oriented. In short, who cares if the pain relief is brought about medically or in some other way as long as the relief of pain is symptomatically real?

The proponent of definition two will respond by saying that I have taken him too literally. All that is necessary is to remove the term 'physiological' from the definition, and all will be well.

Definition Three:
Provider X employing treatment Y to patient Z is futile if and only if: Y does not perform the intended effect on Z

Unfortunately, I do not believe that this simple revision will help the definition significantly. Consider the example of a surgical procedure that is designed to eliminate a cancerous condition from a patient. Suppose that the surgery is provided, but it does not eliminate the cancer. The cancer is metastatic, and the patient is no more likely to survive the disease after the surgery than he was prior to the surgery. The patient is more comfortable, however, after having a tumor removed. According to definition three, the treatment is still futile since palliation was not our intent.

The by now very frustrated provider of definitions will respond once again by saying that I have taken the definition too literally. Surely some unintended benefits may come from our actions, and they must be considered. The proper definition of futility must be...

Definition Four:
Provider X employing treatment Y to patient Z is futile if and only if: Y does not provide any benefit to Z

While this definition may be the most workable yet, it gives up on any notion that the concept of medical futility should be attached to specifically medical or physiological effects. There may be numerous "benefits" to receiving treatment that are emotional or social in nature, but medically inconsequential. Should those benefits be considered in defining medical futility? If they should, then the fact that I would feel bad if I don't receive care would translate into a benefit if I do receive care, and no care could plausibly be defined as futile. We find now that the concept of futility, when carefully defined, becomes too broad to be of any use at all.

Is there no useful concept of futility? We'll see next month.