I ran across a recent review in the New Yorker by Canadian journalist Malcolm Gladwell, commenting on The Death of Cancer, a book by Dr. Vincent DeVita. Dr. DeVita is a leading cancer specialist in the United States who has developed groundbreaking chemotherapy regimens which are still in use. This very influential physician advocates in his book one side of a difficult controversy that we see all the time in healthcare: should we insist on treating patients according to guidelines based in best-quality research studies, or should we encourage physicians to be “creative” and use intuitively reasonable but unproven treatment strategies?

In some ways this controversy is an example of a more abstract question: should we be focusing in healthcare on the individual patient, or on a population of which the patient is presumed to be a member? In my practice caring for frail elderly people at home I find, because my patients never seem to fit into labelled-box categories, I nearly always end up favouring the “individual” side of that question. As far as insisting on guidelines versus trying out unproven treatments is concerned, I am forced almost daily to be creative about what kind of treatment I offer my frail elderly patients.

I think though that it’s easier to make a case for that kind of practice with someone in their 90s, an example of the famous heterogeneity principle in geriatrics. Not only does this old person not quite fit the definition of any study population, he or she is pretty much completely unique, and often needs a unique solution to a unique problem. But what about 30-year-olds with HIV, 40-year-olds with breast cancer, or 60-year-olds with heart attacks? They are much more like the others in their group and likelier to benefit from a treatment that has been proven to be effective in big studies of very similar people. And of course “borderline” situations exist. An 80-year-old with new heart disease who drives a car, runs a business, and has never been sick a day in her life, or an eccentric young cancer patient whose beliefs preclude standard care? What about them?

My solution has turned by default to giving my patients a choice. Many of them have to delegate that choice to a substituted decision-maker, but it’s still the patient’s (not my) choice to make. Once I take the time to let them know the potential risks and benefits of guidelines-based care versus some once-off treatment I believe might help (and if I’m careful not to spin the information I provide) each patient can think over the options in their own circumstance and with their own biases and preferences.

This could be a way out of the dilemma Dr. DeVita discusses. Evidence-based care, clinical guidelines and hospital policies and procedures need to be recognized and treated as suggestions, not tyrannical laws of practice. It is after all the patient who suffers or enjoys the real consequences of what we do and don’t do, not me, my reputation, an institution I work for, or a regulatory health authority.