Don’t treat every patient like your mother

Physicians recommend different treatments for patients than they would choose for themselves.  The preceding statment is true according to a similarly titled article recently published in the Archives of Internal Medicine by Ubel et al and has, I will argue, important implications for how we view the doctor-patient relationship.  In the study, one group of physicians was asked to choose between two hypothetical treatment alternatives for either avian flu or colon cancer as if they themselves were the one with the disease.  The other group was asked to choose between the same hypothetical alternatives as if they were making a recommendation to a patient with the disease (either avian flu or colon cancer).  In the colon cancer scenario, both hypothetical treatment alternatives presented were surgical.  One surgical procedure was 4% less likely to cure the cancer, but did not carry the same 4% complication rate as did the more curative procedure.  The avian flu example involved a hypothetical treatment which decreases the chance of death due to flu from 10% to 5% and hospitalization rate from 30% to 15% but which also carried a 1% risk of a fatal reaction and a 4% risk of lower extremity paralysis.  In this example, physicians were asked to choose between the options of treating versus not treating.

What did they find?  Physicians were significantly more likely to choose the option which carries a higher mortality rate but a lower risk of complications for themselves than they were when making a recommendation to thier patient.  Why was this the case?  The authors point to cognitive bias.  They suggest that the biases of “betrayl aversion” (an exagerated feeling of harm caused by an action designed to prevent harm) and “omission” (the added regret of harm caused by a treatment when compared with the same degree of harm caused by a withholding of treatment) are more at work when doctors are choosing for themselves than they are when choosing for patients. 

The idea that we make better decisions for others than we do for ourselves and our loved one is entirely plausible.  Indeed there is other research to suggest that this is the case.  It is a big part of the reason why doctors shouldn’t operate on friends and loved ones, deliver their babies or, in my view, even prescribe them medications.  This is why I am disturbed by what I believe to be the prevailing view in medicine today – namely that we, as Dr. Wes (http://drwes NULL.blogspot NULL.com/) recommends, “Treat every patient like our mother (http://drwes NULL.blogspot NULL.com/2011/06/for-interns-ten-rules-to-go-by NULL.html).”  I will elaborate.

In the ideal situation, medical decision are based on good evidence, reflect the patient’s beliefs and values and are ethically permissible to the physician.  For this to happen a good doctor needs to a) dispassionately weigh the evidence including all attending risks and benefits of any possible intervention and b) establish a relationship with his or her patient which promotes the expression of autonomy.  Treating patients as we would ourselves or someone who is close to us (i.e. with kindness, respect and empathy) is clearly necessary for the latter of these goals.  But, as the above study demonstrates, it is likely detrimental to the former.

Many commentators worry that the rigorous nature of medical training beats the empathy out of young doctors – that medical students loose their idealism during the third year of medical school, become jaded and cynical.  This may be true, but is the wrong question to ask.  I would argue that being caring and empathic are qualities which are necessary in order to effectively perform one’s function as a physician, but are not goals to be achieved for their own sake.  Just as it helps us develop a rapport with patients and establish an effective therapeutic relationship, empathy impairs our ability to rationally weigh evidence and make decisions free of cognitive bias.  Treating every patient like your mother should not be seen as an ideal to strive for.  Rather, it should be seen as a first step which providers must then overcome in order to provide the best possible care.

Leave a Reply