Skip to main content

Table 2 Comparison of diagnostic accuracy of PC ae C44:6 and CSF cell count to distinguish between acute bacterial meningitis (n = 32) and nonbacterial infectious, autoimmune, and non-inflammatory CNS disorders (n = 189)

From: Phosphatidylcholine PC ae C44:6 in cerebrospinal fluid is a sensitive biomarker for bacterial meningitis

ParameterAUC (95% CI)Cut-offAccuracyaSensitivitySpecificityPPVNPVFold differenceb
CSF cell count0.93c (0.89–0.97)> 1000 cells/μL70.040.699.492.98964.7
PC ae C44:60.93c (0.87–0.99)> 5 nM90.590.690.365.997.96.8
  1. Values were obtained by ROC analysis. Sensitivity, specificity, positive and negative predictive value (PPV and NPV) were determined at the trade-off value (Youden index), corresponding to the maximal value of (sensitivity + specificity)/2 in the case of PCaeC44:6 and the clinically recommended cut-off for cell count. Diagnostic groups included in the analysis: bacterial meningitis (n = 32), HSV encephalitis (n = 9), varicella zoster virus meningoencephalitis (n = 15), enterovirus meningitis (n = 10), facial zoster (n = 16), segmental zoster (n = 14), anti-NMDA-receptor autoimmune encephalitis (n = 8), multiple sclerosis (n = 17), Tourette syndrome (n = 20), Bell’s palsy (n = 11), and normal pressure hydrocephalus (n = 35)
  2. AUC area under the curve, CI confidence interval, ROC receiver operating characteristic
  3. aDefined as (sensitivity + specificity)/2 at the cut-off value, bratio of mean values, bacterial meningitis/all others, casymptotic significance, P < 0.001