ST was a 74-year-old woman who presented to the hospital with syncope. She was healthy and took no medications. She was brought to the hospital by her daughter, who heard her fall in her bedroom. ST was awake but groggy when her daughter came to her aid, and she was back to herself by the time the EMTs arrived. There were no preceding symptoms, no tongue biting, no palpitations, or chest pain.
In the emergency room, she was assessed by a neurologist whose note ended, “Not a seizure, admit to medicine, consult cardiology to rule out something cardiac.”
21st-century medicine is a team sport. In the hospital, we round with teams of doctors and students. The primary, generalist team frequently consults specialist teams. Whether in outpatient or inpatient settings, doctors also work with an amazing array of other healthcare professionals: nurses, medical assistants, certified nursing assistants, pharmacists, physical therapists, occupational therapists, chaplains, social workers, case managers, child life specialists...
Good care requires the work of generalists and specialists. As a general internist, I can confidently direct the care of many diseases and diagnose and initiate treatment for many more. There are also many for which I, and my patients, rely on my subspecialty colleagues. You would not want me taking out your gallbladder, doing your bronchoscopy, treating your lupus, or deciding on the appropriate therapy for your myeloma.
Subspecialists get a lot of glory (and money) in medicine these days. It should seem surprising that the doctor who takes care of everyone and everything makes significantly less than the cardiologist who deals not with the heart but with the heart’s specialized, pacemaker cells; the oncologist who specializes in diseases caused by mutations in a single type of precursor cell; or the ophthalmologist who focuses on a single layer of cells within the eye. (OK, my bitterness slipped out there.) But in the current system, specialists spend more time in training, so they need to be compensated accordingly.
Through career choice, filtering, and evolution, generalists and specialists end up being very different people. In 1953, Isaiah Berlin published the book The Hedgehog and the Fox. The book popularized the idea, attributed to the Ancient Greek poet Archilochus, that a fox knows many things while a hedgehog knows one important thing. That dichotomy applies well to the 21st-century medical generalist foxes and specialist hedgehogs.
Generalists and specialists care for different patient populations
Patients are not triaged before seeing the generalist. There is no barrier (or even delay) between a patient of mine experiencing a symptom and me hearing about it. I occasionally have Patient Portal messages from 2 AM about a symptom that has resolved by the time I return the call the next day. Many concerns presented to the generalist do not have a demonstrable physiologic basis. Of those that do, most can be diagnosed clinically or with basic testing. A generalist carries with him hundreds of differentials, most of which consist of a handful of diagnoses.
Patients who see specialists have (usually/ideally) been referred/triaged by generalists. Patients referred to specialists fall into one of three categories:
They have been diagnosed with a disease (and possibly treated) but now require treatment that is best delivered by a specialist. I might need help managing late-stage CKD or Parkinson’s, or I need someone to do a joint replacement or a Whipple.
They have signs of serious illness that require specialized diagnostics. As testing has supplanted clinical acumen, this group of patients has become smaller. Sure, the oncologist to whom I refer a patient with lung cancer is likely to do further testing to determine the extent and genetics of the disease, but usually I know who to refer to because I know the disease I am referring. The diagnostic mysteries I refer require a diagnostic procedure I cannot perform or whose diagnosis is number 15 on my differential diagnosis; a differential that includes only 10 diagnoses.
They are people I could manage if I had had the time, interest, or bandwidth. I am embarrassed to admit that I refer patients because I sometimes just need to outsource one of their multiple problems.
Different populations lead to differences in clinical reasoning
While quizzing a medical student, a generalist will frequently mention a diagnosis, and then say, “but you never see that.” I include pheochromocytoma, relapsing polychondritis, hairy-cell leukemia, and Langerhans cell histiocytosis among these diseases. I have colleagues who have made whole careers out of these diagnoses I disrespect. The average generalist would miss diagnoses in the specialist population, while the specialist would bring down Medicare in a week caring for the generalist population.
The prognoses of the same disease differ in generalist or specialist practices
Because the generalist sees the average patient with common diseases, outcomes are dominated by those in the middle of the Bell Curve. Specialists see a selected group of patients. The patients I see with migraines have headaches that are easily controlled. My neurology colleagues will only see patients who have failed to improve with three lines of prophylactic and abortive therapy. Their view of migraine prognosis will be significantly darker than mine.
Generalists and specialists read differently
The generalists I know who work to stay abreast of the literature read broadly, consuming half a dozen journals, skimming abstracts, and diving deeply into only a few articles. Subspecialists often read one or two journals, cover to cover. This serves their purposes well. The generalist needs to have a sense of how all diseases are managed to know what their specialists can, cannot, and should not offer, while the specialist needs to be the final word on diagnosis and management.
The best in both worlds
The ideal generalist is accessible to a fault, knows how to diagnose and manage common concerns, knows what she doesn’t know, and knows what her specialists have to offer. It has been said about many fields — but I think it fits best the generalist — you want her to be a great generalist as well as the best cardiologist/neurologist/urologist who has not been trained in that field. The ideal generalist will provide good care of most everything you’re likely to experience. When you need specialty care, she will know the right person to send you to because she knows what that person can and will do.
The ideal specialist knows everything in his field that might be causing symptoms, knows the workup for even the most unusual causes, and is a pro at treating these diseases. When the ideal specialist excludes diseases in his purview, he refers back to the generalist, sometimes with suggestions regarding who might get involved next.
Both the ideal generalist and specialist communicate the intricacies of diagnosis, treatment, and prognosis with their patients and prioritize their patients’ goals and values.
Since medicine is hard and complicated, things do not always go well. Generalists fail when they manage problems outside their scope. More commonly, they fail when they overburden specialists by referring patients with problems that they should manage themselves. They also waste the time of specialists and patients alike by referring patients for whom the specialist has nothing to offer. Specialists fail when they burrow into holes that should be left unexplored, over-evaluating or overtreating. They also fail when they stray from what they know — sometimes tempted by the diffident generalist — referring to other specialists: “consult cardiology to rule out something cardiac.”
ST did well overnight on the general medicine service. Further testing was notable for a mildly elevated creatine kinase level. An echocardiogram was normal, and there was no ectopy on a cardiac telemetry. The team decided that although unlikely, a seizure was the most likely diagnosis. A brain MRI revealed a small infarct in the left temporal lobe. Carotid dopplers were normal. The neurology service was re-consulted. EEG findings increased the suspicion for a seizure.
ST was discharged on aspirin, a statin, and a seizure medication, as well as a 14-day cardiac monitor.
Photos by Ray Hennessy (fox) and Alexas_Fotos (hedgehog) on Unsplash