Inappropriate use of the cerebrospinal fluid Venereal Disease Research Laboratory (VDRL) test to exclude neurosyphilis.

Only 3 of 2536 cerebrospinal fluid (CSF) VDRL tests ordered at the Johns Hopkins Hospital in 1980 were positive. Of patients on whom the test was ordered, 226 had a positive or borderline serum fluorescent treponemal antibody absorption test. Records from 156 (69%) of these patients, including all 3 with positive CSF-VDRL tests, were reviewed and showed that the diagnosis of neurosyphilis had been considered only in 44 (28%). One third of records lacked notations of historical and physical findings characteristic of neurosyphilis. Forty seropositive patients who had lumbar puncture to rule out asymptomatic neurosyphilis had negative CSF-VDRL tests; none had neurosyphilis diagnosed. Use of the CSF-VDRL test seemed to represent predominantly "box-checking" on the requisition to rule out neurosyphilis. Yet the test performs better when "ruling in" rather than ruling out neurosyphilis. The Centers for Disease Control should reevaluate its recommendation to rule out neurosyphilis in asymptomatic patients with untreated syphilis of greater than 1 year's duration, given the costs, risks, and dubious benefits.