ABSTRACT
RESUMO Objetivo: Avaliar o Escore para Meningite Bacteriana (EMB) isolado e associado ao valor do lactato no líquor para diferenciar meningite bacteriana (MB) e meningite asséptica (MA). Métodos: Foram selecionadas crianças com meningite atendidas em hospital terciário privado entre janeiro de 2011 e dezembro de 2014. Os dados foram obtidos na admissão. Utilizou-se o EMB com: coloração de Gram no líquor (2 pontos); neutrófilos no líquor ≥1.000 células/mm3 (1 ponto); proteína no líquor ≥80 mg/dL (1 ponto); neutrófilos no sangue periférico ≥10.000 células/mm3 (1 ponto); e convulsão durante/antes da chegada (1 ponto). Analisou-se também o lactato no líquor (elevado: ≥30 mg/dL). Avaliaram-se sensibilidade, especificidade e valor preditivo negativo de diversos valores de corte do EMB e do EMB associado ao lactato elevado para prever MB. Resultados: Dos 439 pacientes elegíveis, 94 não tinham todos os dados necessários para o escore, sendo 345 pacientes selecionados: 7 no grupo de MB e 338 no de MA. Como preditivos de MB, o EMB ≥1 mostrou sensibilidade de 100% (intervalo de confiança de 95% - IC95% 47,3-100), especificidade de 64,2% (58,8-100) e valor preditivo negativo de 100% (97,5-100), enquanto o EMB ≥2 ou EMB ≥1 associado a lactato liquórico ≥30 mg/dL mostrou sensibilidade de 100% (47,3-100), especificidade de 98,5% (96,6-99,5) e valor preditivo negativo de 100% (98,3-100). Conclusões: O EMB com 2 pontos associado à dosagem de lactato no líquor manteve a sensibilidade e o valor preditivo negativo, ao passo que aumentou a especificidade para identificar meningites bacterianas em relação à utilização do EMB com 1 ponto.
ABSTRACT Objective: To evaluate Bacterial Meningitis Score (BMS) on its own and in association with Cerebrospinal Fluid (CSF) lactate dosage in order to distinguish bacterial from aseptic meningitis. Methods: Children diagnosed with meningitis at a tertiary hospital between January/2011 and December/2014 were selected. All data were obtained upon admission. BMS was applied and included: CSF Gram staining (2 points); CSF neutrophil count ≥1,000 cells/mm3 (1 point); CSF protein ≥80 mg/dL (1 point); peripheral blood neutrophil count ≥10,000 cells/mm3 (1 point) and seizures upon/before arrival (1 point). Cutoff value for CSF lactate was ≥30 mg/dL. Sensitivity, specificity and negative predictive value of several BMS cutoffs and BMS associated with high CSF lactate were evaluated for prediction of bacterial meningitis. Results: Among 439 eligible patients, 94 did not have all data available to complete the score, and 345 patients were included: 7 in bacterial meningitis group and 338 in aseptic meningitis group. As predictive factors of bacterial meningitis, BMS ≥1 had 100% sensitivity (95%CI 47.3-100), 64.2% specificity (58.8-100) and 100% negative predictive value (97.5-100); BMS ≥2 or BMS ≥1 associated with high CSF lactate also showed 100% sensitivity (47.3-100); but 98.5% specificity (96.6-99.5) and 100% negative predictive value (98.3-100). Conclusions: 2 point BMS in association with CSF lactate dosage had the same sensitivity and negative predictive value, with increased specificity for diagnosis of bacterial meningitis when compared with 1-point BMS.
Subject(s)
Humans , Male , Female , Infant , Child, Preschool , Child , Cerebrospinal Fluid/chemistry , Meningitis, Bacterial/diagnosis , Lactic Acid/analysis , Meningitis, Aseptic/diagnosis , Retrospective Studies , Cohort Studies , Diagnosis, DifferentialABSTRACT
PURPOSE: Bacterial meningitis score (BMS) has been introduced as a clinical predictive parameter for diagnosing cerebrospinal fluid (CSF) pleocytosis in children at very low risk of bacterial meningitis in the postconjugate vaccine era. This study aimed to examine the usefulness of the BMS and to identify an additional index to distinguish between enteroviral meningitis and bacterial meningitis. METHODS: We retrospectively included 289 patients with enteroviral meningitis and 10 patients with bacterial meningitis between the aged 2 months to 16 years. We applied the BMS to all the included patients, and compared the initial laboratory and clinical characteristics between the two groups. RESULTS: Of the 210 patients categorized as having a very low risk of bacterial meningitis based on BMS, 2 (1%) had bacterial meningitis, both of whom were younger than 3 months. The sensitivity, specificity, and negative predictive value of the BMS for bacterial meningitis were 80%, 72%, and 99%, respectively. Compared with patients with enteroviral meningitis, those with bacterial meningitis were younger (0.3 years vs. 5.0 years, P<0.01), had higher leukocyte count (530/mm³ vs. 43/mm³, P=0.04), neutrophil count (490/mm³ vs. 20/mm³, P<0.01), and protein level (106 mg/dL vs. 30 mg/dL, P<0.01) in the CSF, and increased serum C-reactive protein (CRP) level (79 mg/L vs. 4.5 mg/L, P<0.01). CONCLUSION: Although the BMS is a useful clinical predictive parameter for identifying children with CSF pleocytosis who are at very low risk of bacterial meningitis, it should be applied with caution in young infants. In addition to applying BMS, clinical parameters such as CSF profiles and serum CRP levels are useful in clinical decision making for the management of children with CSF pleocytosis.
Subject(s)
Child , Humans , Infant , C-Reactive Protein , Cerebrospinal Fluid , Clinical Decision-Making , Leukocyte Count , Leukocytosis , Meningitis , Meningitis, Aseptic , Meningitis, Bacterial , Neutrophils , Retrospective Studies , Sensitivity and SpecificityABSTRACT
PURPOSE: Bacterial meningitis score (BMS) has been introduced as a clinical predictive parameter for diagnosing cerebrospinal fluid (CSF) pleocytosis in children at very low risk of bacterial meningitis in the postconjugate vaccine era. This study aimed to examine the usefulness of the BMS and to identify an additional index to distinguish between enteroviral meningitis and bacterial meningitis. METHODS: We retrospectively included 289 patients with enteroviral meningitis and 10 patients with bacterial meningitis between the aged 2 months to 16 years. We applied the BMS to all the included patients, and compared the initial laboratory and clinical characteristics between the two groups. RESULTS: Of the 210 patients categorized as having a very low risk of bacterial meningitis based on BMS, 2 (1%) had bacterial meningitis, both of whom were younger than 3 months. The sensitivity, specificity, and negative predictive value of the BMS for bacterial meningitis were 80%, 72%, and 99%, respectively. Compared with patients with enteroviral meningitis, those with bacterial meningitis were younger (0.3 years vs. 5.0 years, P<0.01), had higher leukocyte count (530/mm³ vs. 43/mm³, P=0.04), neutrophil count (490/mm³ vs. 20/mm³, P<0.01), and protein level (106 mg/dL vs. 30 mg/dL, P<0.01) in the CSF, and increased serum C-reactive protein (CRP) level (79 mg/L vs. 4.5 mg/L, P<0.01). CONCLUSION: Although the BMS is a useful clinical predictive parameter for identifying children with CSF pleocytosis who are at very low risk of bacterial meningitis, it should be applied with caution in young infants. In addition to applying BMS, clinical parameters such as CSF profiles and serum CRP levels are useful in clinical decision making for the management of children with CSF pleocytosis.