Meningococcal disease is normally suspected on clinical grounds and is
confirmed by isolation of Neisseria meningitidis bacteria from blood
or cerebrospinal fluid or, more recently, by serology or PCR of cerebr
ospinal fluid. Achieving confirmation of a clinical diagnosis of menin
gococcal disease has become more difficult in the last few years. The
prehospitalization administration of parenteral benzylpenicillin norma
lly renders blood cultures sterile, and lumbar puncture is undertaken
less frequently, especially in young children. We evaluated PCR for th
e detection of meningococcal DNA in 80 blood samples taken from patien
ts with known or suspected meningococcal disease or from patients with
other diagnoses (negative controls). Both the sensitivity and the spe
cificity of the test were 100% for patients with confirmed cases of me
ningococcal disease when the blood buffy coat was used (83 to 100% sen
sitivity and 87 to 100% specificity with 95% confidence limits). Posit
ive PCR results could be obtained from both blood buffy coat and serum
samples. Sensitivity was unaffected by prior antibiotic treatment. PC
R is a rapid, sensitive test that may be used to confirm a diagnosis o
f meningococcal disease by using peripheral blood samples. Introductio
n of this test into clinical laboratories may in some cases obviate th
e need for lumbar puncture to be performed on patients with suspected
meningococcal disease. Our results demonstrate that a substantial numb
er of cases of meningococcal disease are not confirmed by conventional
techniques and remain undiagnosed. If the PCR test described here was
widely applied, the number of cases of meningococcal disease ascertai
ned might rise by as much as 60% greater than that recognized at prese
nt. It is likely that we are in a prevaccination era for meningococcal
disease. Better case ascertainment is urgently required to assess the
need for vaccines, to determine their costs and benefits, and to moni
tor their efficacies.