ABSTRACT
BACKGROUND: Difficulty in interpreting white blood cell (WBC) counts in cerebrospinal fluid (CSF) complicates the diagnosis of neonatal meningitis in traumatic lumbar punctures (LP). The aim of our study was to determine the correction factor for WBC counts in traumatic LP that offers the greatest diagnostic efficacy in meningitis. METHODS: We conducted a retrospective observational study of LP in neonates between January 2014 and December 2020. Traumatic LP was defined as a red blood cell (RBC) count ≥ 1,000 cells/mm3 CSF and pleocytosis as WBCs ≥ 20 cells/mm3 CSF. The CSF RBC:WBC ratio was analyzed by linear regression to determine a new correction factor. Cell count adjustments were also studied using the 500:1, the 1,000:1 ratio method, and the peripheral blood RBC:WBC ratio, using ROC curves and studies of accuracy (sensitivity and specificity). RESULTS: Overall, 41.0% of the 1,053 LPs included in the study were traumatic. The best results for effective WBC correction were the method based on the peripheral blood ratio (sensitivity = 1.0 and specificity = 0.9 for bacterial meningitis and sensitivity = 0.8 and specificity = 0.9 for viral meningitis) and the 400:1 ratio (sensitivity = 1.0 and specificity = 0.8 for bacterial meningitis and sensitivity = 0.8 and specificity = 0.8 for viral meningitis) obtained from linear regression (95% CI 381.7-427.4; R2 = 0.7). CONCLUSION: Both the peripheral blood correction and the 400:1 correction reduce the number of neonates classified with pleocytosis who were not eventually diagnosed with meningitis. Both methods might be a useful tool to clarify the neonatal meningitis diagnosis, offering neonatologists the possibility to assess the WBC count in traumatic LP.