The old adage, “When you hear hoof beats, think of horses, not zebras,” is a great principle for our students and residents. Unfortunately, for many common conditions or presenting complaints, we lack robust epidemiologic evidence based on unselected patients from the community with which to inform our decisions. Consider irritable bowel syndrome.
In medical school, I was taught that this condition was a diagnosis of exclusion, even for younger patients. (“You wouldn’t want to miss a diagnosis of cancer, would you?”) Countless patients endured barium enemas, barium swallows, and a host of evaluations before receiving (largely ineffective) therapies.
Like so many “pearls” I embraced then, this approach has been discarded, as Dr Keith Holten outlines in this issue of The journal of family practice (“Irritable bowel syndrome: Minimize testing, let symptoms guide treatment,”). What we really should be doing, according to latest recommendations, is to stratify patients on the basis of risk (pretest probability) of serious conditions, and test (or treat presumptively) on this basis.
Of course this appeals to my practical FP genes, but I am discomforted by how thin the evidence is to warrant this recommendation. How can we provide cost-effective, evidence-based care when so little of our “knowledge” is derived from the patients who come to our offices?
If America’s health care system is to truly reform and become value-driven and evidence-based, we need a common electronic health record and data repository that can generate representative epidemiologic data. The Netherlands, I understand, has such vision—why not the US? We must fund further primary care research efforts that look at common conditions and the “meaning” of presenting symptoms over time. As we wade through increasingly byzantine MediCare rules, we practice a medieval alchemy of divining hoof beats.
Don’t our patients deserve better?
