Background: The prompt diagnosis of meningitis is essential for a good outcome.
Aim: This study was designed to differentiate children with meningococcal meningitis, aseptic meningitis, and no meningitis.
Patients and Methods: A total of 150 children aged 1 month to 12 years and adolescents aged 12 to 14 years, who had had a lumbar puncture, were recruited. In all patients, demographic characteristics and clinical and laboratory findings were recorded, and polymerase chain reaction for enterovirus was performed. On discharge, the patients were divided into 3 groups: children and adolescents with meningococcal meningitis, those with aseptic meningitis, and those without the disease.
Results: Statistical analysis revealed that headache, vomiting, and clinical stiffness could differentiate the combined groups of meningitis irrespective of etiology with the group without meningitis. The presence of headache increased the probability of meningitis by 76%, the presence of vomiting doubled the probability, whereas the presence of neck stiffness increased the respective possibility by 67%. The number of leukocytes, neutrophils, and platelets as well as cerebrospinal fluid cells differed significantly among the 3 groups. An increase of 10 cerebrospinal fluid cells corresponds to a 12% elevation of the odds ratio for meningitis being meningococcal, whereas logistic regression analysis did not show any other significant variable among those of laboratory findings. The result of the polymerase chain reaction for enterovirus was negative in all samples.
Conclusion: Clinical characteristics should contribute to the physician's decision to perform a lumbar puncture for the diagnosis of meningitis.
From the *Third Department of Pediatrics, Athens University Medical School, Athens, Greece; †Hellenic Pasteur Institute, and ‡Second Department of Pediatrics, Athens University Medical School, Athens, Greece.
Reprints: Andrew Fretzayas, MD, PhD, 3rd Department of Pediatrics, Attikon University Hospital, Rimini St, 12462 Haidari, Athens, Greece. E-mail: firstname.lastname@example.org.
The authors report no conflicts of interest.