High case-fatality rates related to acute meningitis make accurate diagnosis critical. The clinical signs and symptoms occurring at presentation may predict prognosis, and early diagnosis may also decrease the number of adverse outcomes. Attia and associates conducted a Medline search to review the role of clinical examination in the diagnosis of acute meningitis.

Studies were included if they involved patients whose meningitis was confirmed by lumbar puncture or autopsy findings. Community-acquired meningitis is more likely to occur in patients with diabetes mellitus, otitis media, pneumonia, sinusitis or alcohol abuse. Many of the signs and symptoms of meningitis are nonspecific, such as fever, myalgia and rash. The inflammatory response of the meninges leads to some of the more protective responses, such as clinically detectable neck stiffness, or Kernig and Brudzinski signs. As the inflammation progresses, it can also cause headache, vomiting and changes in mental status (see accompanying table). Fewer than two thirds of patients present with the classic triad of neck stiffness, fever and altered mental status. On examination, the patient may have a fever or rash, and may spontaneously assume a position known as the Amoss or Hoyne sign, in which the knees and hips are flexed, the neck is extended and the back is arched. Kernig and Brudzinski signs are also described as typical findings on examination. The Kernig sign is positive if pain in the lower back or posterior thigh occurs when the knee is extended while the patient is lying in the supine position and the hip is flexed at a right angle. The Brudzinski sign is positive if knee and hip flexion occurs when the neck is flexed while the patient is in the supine position. Another commonly performed maneuver is jolt accentuation of the headache, which is positive when the headache worsens as the patient turns his or her head horizontally at two to three rotations per second.

The authors found that single symptoms had a low sensitivity for diagnosing acute meningitis. The sensitivity of neck pain was 28 percent; nonpulsatile headache, 15 percent; generalized headache, 50 percent; and nausea and vomiting, 60 percent. Therefore, clinical history alone was not useful in diagnosing acute meningitis. However, some elements of the physical examination were useful in establishing the diagnosis. Almost all patients (95 percent) had at least two of the three symptoms, and some studies found that even more patients (99 to 100 percent) presented with at least one symptom. If a patient presents without fever, neck stiffness or altered mental status, it is highly unlikely that he or she has acute meningitis. The sensitivity of fever as a single sign was 85 percent, and the specificity was 45 percent. However, it should be noted that septic patients may actually be hypothermic, so the absence of fever lowers the likelihood of acute meningitis but does not eliminate it entirely. Neck stiffness was found to have a sensitivity of 70 percent; Brudzinski sign had a 97 percent sensitivity and the Kernig sign had a sensitivity of 57 percent. Normal mental status, in patients at low risk for meningitis, may rule out this condition. There have been few studies on the specificity of these signs but, in general, the specificity is low.

[ corrected] The authors conclude that meningitis may be ruled out if a patient does not have fever, neck stiffness or changes in mental status. Fever is the most sensitive sign, followed by neck stiffness. Kernig and Brudzinski signs, although poorly tested, seem to be specific for the diagnosis of meningitis. The authors do not believe that either sign is sensitive enough to be used for making the diagnosis of meningitis. Finally, the authors found that jolt accentuation of headache may clarify the diagnosis in a patient with fever and headache by confirming a decision to perform a lumbar puncture.