Why Is “Hard” Evidence “Hard” to Accept?

Abstract

I would like to reflect on a recent editorial (1) lamenting lack of evidence for spirometry in case finding, diagnosis, and management of chronic obstructive pulmonary disease (COPD) in primary care (2). I was struck by the editorialists’ honest selfassessment that this “unimpeachable review of the evidence [was] . . . hard to accept, especially for those like ourselves, who cannot conceive of practicing without objective assessment of pulmonary function” (1). My perspective is that of a board-certified family physician primary care specialist who has cared for patients with lung conditions for over 30 years, and also that of a clinical researcher with training in epidemiology, trial design, and critical assessment ofmedical evidence.As part ofmyclinical care and research, I have personally performed hundreds of preand post-bronchodilator spirometry and peak expiratory flow measurements. It was not hard for me to accept the lack of evidence. Why should it be so for others? I believe it is hard because of a deeply held, often unstated (and even more often incorrect) assumption that permeates theAmerican referral specialist medical community. This assumption (or “faith”) is that medical technologies, clinical skills, and treatment strategies that serve referral specialists in caring for sick patients will also improve the health of the greater number of Americans who are not referred. That this assumption is patently incorrect was illustrated years ago in a humorous but insightful commentary that should be read by all who seek to understand the complementary roles played by primary care specialists and referral care specialists seeking to achieve optimum health for their patients (3). I agree with Boushy and colleagues (1) that “gaps in the evidence . . . need to be filled by carefully designed studies of sufficient size.” It should be obvious that primary care studies need to be performed on primary care patient populations. What is less obvious but most important is that primary care studies should be designed to ask and answer primary care questions. Success in answering questions about the utility of spirometry in primary care will depend on a meaningful collaboration between primary care specialists and referral specialists, who will have to share power and perspective that, in the United States, have traditionally been relegated entirely to them. As this process of collaborative research proceeds, I suggest it will be wise for all concerned to continuously reassess their underlying assumptions about similarities and differences between primary and referral care.