The world's most widely known physician is probably not William Osler, but Sir Arthur Conan Doyle, the Scottish general practitioner who penned the tales of Sherlock Holmes at the turn of the century (25). I had long been dimly aware that the famous detective's fantastic observational and deductive skills were drawn from a medical model (26) (see also the Extracts), but when, as a first-year medical student at Johns Hopkins, I heard upperclassmen relate amazing stories of the diagnostic acumen of Tumulty and the encyclopedic knowledge of McKusick, Holmes' feats suddenly did not seem so stratospheric after all (27).

With this frame of mind, the specific impetus that launched Zebra Cards occurred the next year, during the introductory course in clinical skills. I still remember clearly the feeling I had when I first learned of the standard review of systems question, "How many pillows do you sleep with?" The idea that physicians could somehow assess cardiac function by using pillows was, in my mind, fantastic. That I was now privy to this secret, and would soon be let in on others like it, was exhilarating. It was then I began collecting "elegant" clinical signs, which soon coalesced into the concept of zebras.

The succeeding years have made it clear that zebras are not static: They vary as one's knowledge accumulates. No one but a second-year medical student would think of the association between heart failure and orthopnea as a zebra. A third-year medical student thinks the association of gingival hypertrophy with phenytoin use is a zebra, while the third-year medical resident does not. Zebras also change as medicine changes. In a 1928 address, Archibald Garrod (28) had hoped "to have spoken of that malady involving the bone-marrow, of which the copious excretion of Bence Jones protein is a diagnostic sign, for it is one of the most obscure of rare diseases." Today, "obscure" is not a word that describes multiple myeloma. Syphilis, the great imitator of a generation ago, is now relegated to zebra in most differential diagnoses; immunodeficiency is not.

We have been careful to distinguish zebras from mere rare diseases, but a brief foray into the literature of rare diseases is instructive, as the field is closely allied with our own. As the Extracts show, some of the brightest stars in the history of medicine have had a special interest in rare diseases, including Hutchinson, Harvey, Paget, and Garrod. Books specifically adressing rare diseases and lesions have appeared (29-34), and others are so enriched with uncommon entities that they serve as invaluable resources, as their frequent citation in these cards attests (35-38). Most of these works devote at least some space to justifying the study of rare diseases, a practice that continues even today, in the era of the omnivorous biomedical research establishment (39).

The points made in these justifications tend to fall into one of a few general arguments. McKusick (7), in his Foreword to Bean's 1967 Rare Diseases and Lesions, enumerates them. He lists four reasons (one of which was alluded to previously) "that rare conditions are, or should be of interest to physicians":

• Rare disorders can teach us much about the normal or about more common disorders.
• Rare manifestations are sometimes valuable clues to the existence of grave internal disease.
• People have them.
• They are a break in the routine and "keep [the physician's] powers of observation from undergoing atrophy."

It is simple to see that these justifications may be applied equally well to zebras; for the proof of the first three arguments, one need only look in the cards (for example, HE-004 for the first argument, and GI-001 for the second).

Though it has long been known that the rarity of a disease actually makes it rarer (40), and recently stated why "classic cases" are rare (41), there is a mathematical question regarding rare diseases that I have not seen discussed. The devoted student of rare conditions, F. Parkes Weber (42), noted that "For one common disease or syndrome there are several rare ones." Might it then be true that, lumping these rare disorders into a single entity, it is a common occurrence for someone to be suffering from a rare disease (43)?

---

[The Fat Man] was the only resident to have his own reference file on every disease there was, on three-by-five cards. He loved references on three-by-five cards. He loved everything that was on a three-by-five card.

The House of God (47)

---

Back