Background: Despite the great morbidity and mortality that childhood bacterial meningitis (BM) is experiencing in Africa, diagnosis of BM in resource-limited contexts is still a challenge. Several algorithms and clinical predictors have been proposed to help physicians in decision-making but a lot of these markers used variables that are calculable only in well-equipped laboratories. Predictors or algorithm based on parameters that can be easily performed in basic laboratories can help significantly in BM diagnosis, even in resource-limited settings, rural hospitals or health centers.
Results: This retrospective study examined 145 cerebral-spinal fluid (CSF) specimens from children from 2 months to 14 years. CSF specimens were divided into two groups, according to the presence or not of a clinical diagnosis of BM. For each specimen, CSF aspect, CSF white blood cells (WBC) count, CSF glucose and protein concentration were analyzed and statistical analysis were performed. CSF WBC count ≥10/µl is no more a valuable predictor of BM. CSF protein concentration ≥50 mg/dl has a better sensitivity for BM diagnosis and when used with CSF glucose concentration ≤40 mg/dl, can help to diagnose correctly almost all the BM cases. An algorithm including CSF protein concentration, glucose concentration and WBC count has been proposed to rule out BM and to correctly diagnose it.
Conclusions: In resource-limited health centers, the availability of a combination of easy-to-obtain parameters can significantly help physicians in BM diagnosis. The prompt identification of a BM case can be rapid treated or transferred to adequate structures and can modify the outcome in the patient.