A risk score for identifying patients at a low risk of bacterial meningitis amongst adults with cerebrospinal fluid leucocytosis and a negative gram stain result: a derivation and validation study. (March 2023)

• Record Type: Journal Article
• Authors: van Soest, Thijs M.; Horst, Liora ter; Chekrouni, Nora; Bijlsma, Merijn W.; Brouwer, Matthijs C.; Portillo, Daniela Urueta; van de Beek, Diederik; Hasbun, Rodrigo
• Is Part Of: Clinical microbiology and infection. Volume 29:Number 3(2023)
• Page Start: 360
• Page End: 365
• Publication Date: 2023-03
• Subjects: Antibiotic use -- Community-acquired bacterial meningitis -- Risk score -- Viral encephalitis -- Viral meningitis; Medical microbiology -- Periodicals; Diagnostic microbiology -- Periodicals; Communicable diseases -- Periodicals; Infection -- Periodicals; 616.01
• DOI: 10.1016/j.cmi.2022.10.001
• Languages: English
• ISSNs: 1198-743X

Abstract

Abstract: Objectives: We aimed to derive and validate a risk score to differentiate patients with bacterial meningitis from those with viral meningitis or encephalitis amongst patients presenting with cerebrospinal fluid (CSF) leucocytosis and a negative Gram staining result. Methods: We included adults with bacterial and viral meningitis or encephalitis presenting with CSF leukocyte counts of >10 per mm 3 and a negative Gram staining result from cohorts in Houston, Texas (2004–2019), and the Netherlands (2012–2021). Derivation and the first validation were performed in the American patients and further validation in the Dutch patients. Results: Derivation was performed in 109 American patients with bacterial meningitis (median age, 56 years; interquartile range [IQR], 46–66 years; 46% women) and 194 with viral meningitis or encephalitis (median age, 46 years; IQR, 33–60 years; 53% women). Serum leukocyte counts of >10.0 × 10 9 /L, CSF leukocyte counts of >2000 per mm 3, granulocyte counts of >1180 per mm 3, protein levels of >2.2 g/L, glucose levels of <1.9 mmol/L and fever on admission were included in the risk score, which was dichotomized into 'low risk' (0 present) and 'high risk' (>0 present). The first validation showed a sensitivity of 100% (95% CI, 96.6–100) and specificity of 34.0% (95% CI, 27.4–41.2). Further validation in 262 Dutch patients with bacterial meningitis (median age, 57 years; IQR 44–70 years; 45% women) and 68 with viral meningitis (median age, 34 years; IQR, 28–45 years; 60% women) showed a sensitivity of 99.6% (95% CI, 97.9–100) and specificity of 41.2% (95% CI, 29.4–53.7). Conclusions: Our risk score may be able to rule out bacterial meningitis amongst patients presenting with CSF leucocytosis and a negative Gram staining result. However, it needs prospective testing prior to clinical implementation.