METHODS: CSF samples were collected from 2 neurosurgical centers in Kuala Lumpur, Malaysia, between January 2022 and October 2023. Conventional markers and CLGR were quantified using standard laboratory methods, with BGA utilized for measurement when feasible. Samples were categorized into confirmed CBI-positive (CBI+) and CBI-negative (CBI-) groups. Marker performance was compared, and receiver operating characteristic analysis conducted. Pearson correlation assessed the agreement between BGA and laboratory measurements.
RESULTS: Among the 130 CSF samples, 11 were CBI+. Both cerebrospinal fluid lactate (cLac) and CLGR were significantly elevated in the CBI + group (P < 0.001). The area under the curve for cLac and CLGR was 0.990 and 0.994, respectively. Using a cutoff of 6.0 mmol/L, cLac demonstrated sensitivity of 100%, specificity of 93.3%, positive predictive value of 57.9%, negative predictive value of 100%, and diagnostic accuracy of 93.9%. CLGR ≥20.0 showed even higher accuracy: 100.0% sensitivity, 98.6% specificity, 84.6% positive predictive value, 100% negative predictive value, and overall accuracy of 98.5%. Both markers maintained excellent performance in blood-stained CSF. BGA measurements correlated well with laboratory results (r = 0.980 and 0.999, respectively, P < 0.001).
CONCLUSIONS: CLac levels ≥6.0 mmol/L and CLGR ≥20.0 accurately identified CBI in neurosurgical patients, with CLGR exhibiting superior efficacy. The potential for instant BGA measurement suggests promise for point-of-care testing.