To determine the usefulness of cerebrospinal fluid (CSF) tests for syp
hilis at a large academic hospital, clinical and laboratory data on 64
4 patients in whom such testing was requested over a 12-month period w
ere analysed. In 198 cases (31%) the Treponema pallidum haemagglutinat
ion (TPHA) screening test could not be performed because of insufficie
nt fluid. Thirty-eight of the remaining patients were diagnosed as hav
ing active neurosyphilis. Examination of 22 files of patients who had
a positive TPHA and fluorescent treponemal antibody absorption (FTA-Ab
s) test together with a negative CSF Venereal Disease Research Laborat
ory (VDRL) test revealed that other CSF measures indicating disease ac
tivity (CSF protein, cells or IgG index) were not utilised optimally.
In 10 (45%) of these patients neurosyphilis was not diagnosed despite
either abnormal or incomplete CSF biochemical analysis, indicating tha
t if the CSF VDRL is used as the sole marker for disease activity, som
e cases of neurosyphilis are likely to be missed.