I wrote this last night, and then the second I was about to press the submit button, I accidentally the back button.... so this is only a pale comparison of what I originally wrote (at least in my mind).

So first of all -- doctors LOVE Sherlock Holmes. Medline has `160 articles dedicated to him, and usually there's a reference at any major medical conference. So there's actually a collection of literature of "Brain Fever" in Holmes. I'm using Caplan's "Sherlock Holmes and 'Brain Fever'" by Richard Caplan, Perspectives in Biology and Medicine, Volume 30, Number 3, Spring 1987, pp. 433-439 if you want to read more.

In Doyle's context brain fever is NOT encephalitis or meningitis but rather a psychosomatic stress reaction that involves altered consciousness, agitation, hysteria (and fever). By nature it was transient.

Doyle himself gets all meta and has a physician character give a definition (and commentary) in "A Medical Document":

"Then there is the mysterious malady called brain fever, which always attacks the heroine after a crisis, but which is unknown under that name to the textbooks. People when they are overexcited in novels fall down in a fit. In a fairly large experience I have never known anyone to do so in real life."

The term appears, with the same context, in other 19th century literature, such as Huck Finn. And when Edgar Allen Poe's wife died, he was described in the popular press as having fits of recurrent brain fever.

But was it an actual diagnosis outside of literature? Not in the time Doyle was writing. Caplan reviewed the pre-eminent American textbooks of the late 19th century (Austin Flint's Clinical Medicine: A Systematic Treatise on the Diagnosis and Treatment of Diseases, 1879), and Osler's famous Principles and Practice of Medicine (1892). There is no mention of any disease that sounds like the brain fever described in literature.

Caplan did find it described in an older work -- a general layman's textbook from 1847, called "Domestic Medicine, or a Poor Man's Friend in the Hours of Affliction, Pain, and Sickness."

In it, the author describes the various types of fever -- ague (malaria), bilious, scarlet, cholic, and nervous. Nervous fever in particular appears to be the same as brain fever:

"It affects the whole nervous system and produces a tremulous motion of the body and limbs: the system seems to be sinking; there is a clammy, cold and unnatural perspiration or sweat on the skin, and the pulse is extremely weak. ... The sleep is very much disturbed and unrefreshing; the countenance sinks or seems to change from its natural expression of feature to a ghastly appearance . . . the spirits flag, and the mind broods over the most melancholy feelings, without knowing the cause; the sight of food is very unpleasant and sometimes disgusting."

The problem here, of course, is that this description is already terribly antiquated by the time it was published in 1847. Nosology pretty much as we know it today was under development in 1847, and treating fever as a disease -- rather than as a symptom of all different sorts of conditions -- was antiquated, which is why it's appropriate that it turns up in a lay document rather than prominent medical text books.

Probably the best way to view "brain fever" in Doyle's works is as a medical trope, similar to how dissassociative identity disorder (multiple personality disorder) is used today. A convenient plot device, based on antiquated but popular medical definitions, and probably used a bit tongue-in-cheek.

Hope that answered your question! I promise, my first answer was better :). I'm currently researching an episode about the lure of Sherlock Holmes for physicians, and the link to medical decision making. It should be out next month if you're interested!