The topic of nurse practitioners in the context of primary care has been resurgent of late, most notably in this post by Maggie Mahar. Much of the conversation is dominated by assertions such as this:

…Nurse Practitioners have the needed training and that, in fact, doctors who have gone through the full medical school curriculum are over-qualified for a job that, today, is more about coordinating care than medical science…

…“For most of these tasks, many primary care doctors are actually ‘over-qualified.’” … “Clearly, while they have taken on the role of health care ‘coordinators’ they have become more dependent on specialists to take care of the sickest patients. Their ‘scientific’ medical role has decreased while their ‘coordinating” role has increased. For many primary care physicians their medical training is of less importance in their new roles.

The problem is that although that description seems to be correct (and in far too many cases, is indeed all too accurate), it is because the practice of medicine in this country has deteriorated into an inappropriate emphasis on treatment while shamefully neglecting the necessary art of diagnosis.

Medicine consists of two distinct parts:

1. Diagnosis: figuring out what is wrong with the patient, and
2. Treatment: deciding what to do for the patient, and then carrying out the plan.

Far too many non-physicians seem to feel that medical diagnosis isn’t really all that complicated. Plug symptoms and exam findings into an appropriately sophisticated algorithm and out pops the answer. Hey, patients can even do it themselves on WebMD, right?

All the physicians reading this (especially those who saw a patient with a list from WebMD today) are shaking their heads sadly, knowing just how far this is off the mark. Correctly diagnosing what is wrong with a given patient is the sine qua non of practicing medicine, and although it seems simple with straightforward patients (and/or brilliant physicians), medical diagnosis is truly an art that takes years to fully master. Eliciting nuances of the medical history gleaned from extraordinary interview skills can only be demonstrated in medical school. Appreciating a subtle physical finding with painstakingly honed physical examination techniques can only be accomplished with time. These are skills only attained with copious hands-on experience.

Great physicians are great diagnosticians. And it is all those background years of education and training (the “full medical school curriculum” spoken of so disdainfully by the nurse practitioner advocates) that prepares us to master this critical skill. All that “extra” information provides us with the key knowledge patients (and nurse practitioners) are lacking when evaluating internet databases like WebMD: what to ignore. Recognition of what is not important is critical.

I am unswayed by the study quoted by Mahar, a survey of responses to a hypothetical patient with acute gastritis in which nurses were found to take a more complete history and prescribe fewer drugs than doctors. But the fact that the doctors did a lousy job (by report; might they be more complete when faced with the actual patient instead of just a study scenario?) is more a condemnation of the deterioration of American medicine than a paean to the diagnostic skills of nurse practitioners. As vital as it is to identify extraneous information, you cannot diagnose something if you’ve never heard of it.

How about an adult with a sore throat and negative strep test who is getting worse over three to five days? How many nurses have even heard of Lemierre’s syndrome? My experience with nurse practitioners (anecdotal, I know; what can I say? I’m human) is that of very limited diagnostic acumen, coupled with a significant overuse of consultants and prescription medications, especially controlled substances.

It has been argued that treatment is far more straightforward than diagnosis, and in many cases, that is very true. “Cookbook” medicine often works well, but only to the extent that the patient’s condition has been correctly diagnosed. I won’t deny that treatment often needs tweaking for individual patients, but this is seldom as complex an endeavor as diagnosis. And this is where American medicine falls on its collective ass. We may have the best treatment in the world, but in general, our diagnostic skills suck! You can have self-service gas stations every half-mile along the highway, but that’s not going to get your car started if your battery is dead.

That said, I admit that far too many doctors — both primaries and specialists — are terrible diagnosticians. Whether due to lack of time or intellectual laziness, far too many of us don’t put forth the effort to properly diagnose our patients. Shotgun studies and referrals may have become the norm, but that doesn’t make it right.

The practice of medicine is the diagnosis of disease and the treatment of patients. “Coordination” of care (diagnosis and treatment; recurring theme here?) is certainly something that could be accomplished by non-physicians, as long as recognition remains that physicians are the ones best suited to diagnosing and treating (AKA practicing medicine). Maggie Mahar may prefer the “comfort and care” approach that nurses claim to offer instead of “the scientific perspective of medical schools that teach about disease processes and bodily interactions,” but without first having an accurate diagnosis, she and many others could find themselves in deep trouble.

Lucy Hornstein is a family physician who blogs at Musings of a Dinosaur, and is the author of Declarations of a Dinosaur: 10 Laws I’ve Learned as a Family Doctor.