Applying Evidence to Patient Care: From Black and White to Shades of Grey

The interface between evidence and clinical care is complicated: What appears black and white in a trial report may “rapidly become grey in practice” (1). Even when presented with internally valid, well-conducted trials, clinicians frequently face doubts about how we can apply this evidence to our individual patients. The potential problems in extrapolating from trials that are small, are of short duration, rely on surrogate outcomes, or use active controls of uncertain efficacy are well described (2-5). However, in this editorial, I focus on the dilemma at the heart of Dr. Shlipak's paper in this issue (6): Can we extrapolate from trials conducted in highly selected subsets of patients to a broader population of patients who have the same condition but do not meet the trial eligibility criteria? Indeed, as clinicians we infrequently see the ideal patients who populate clinical trials. Instead, we spend most of our time looking after patients who would have been deemed ineligible to participate in these trials because of age, comorbid conditions, or concomitant medication use.

As Dr. Shlipak points out, while one third of patients with heart failure have abnormal renal function, few of the trials that make up the evidence base for heart failure have enrolled patients with renal insufficiency (and those that did enrolled few such patients). This is not unique to renal dysfunction—a review of 59 heart failure trials documented that, compared with nontrial patients, trial participants are younger, have fewer comorbid conditions, and are more likely to be white men (7). As a result, although the evidence base in heart failure is strong …