Bacterial meningitis

Steroids
Patients with acute bacterial meningitis should receive dexamethasone 10 mg IV before or with the first dose of antibiotics. This is a Neurocritical Care Society Clinical Performance Measure.

Risk factors for poor outcome

 * Patient factors
 * Age (older)
 * Altered mental status, included worsened scores on GCS, Hunt-Hess scale, and Scandinavian Stroke Scale
 * Comorbidities
 * Cranial nerve palsy
 * Hypotension
 * Seizures
 * Tachycardia
 * Labs
 * CSF
 * WBC count (low)
 * Gram stain (positive, growing bacteria)

Risk scores
Neither of these scores performs well enough for use in individual patient management.

Aronin et al. score (1998)
Derived from 176 patients from the USA, 47% S. pneumoniae, 12% N. meningitidis, and externally validated in 93 patients from the USA. However, the score did not perform well on a validation study of 2104 patients from the Netherlands, with AUROC for death or neurological deficit of 0.59 (95% CI 0.57-0.61).

Dutch Meningitis Risk Score (Weisfelt et al. score) (2008)
Derived from 696 episodes of meningitis in the Netherlands, 51% S. pneumoniae, 37% N. meningitidis. This score has been externally validated in a 301 episode cohort from the Netherlands (AUROC 0.81) but had worse performance in other countries such as Vietnam (n=426, AUROC 0.70) and Malawi (n=4654, AUROC 0.68). In 2104 patients from the Netherlands, it had an AUROC of 0.74 (95% CI 0.71-0.76).

Simplified Acute Physiology Score (SAPS) II
This general ICU mortality estimate was found to be predictive of mortality in a study of 98 adult patients with bacterial meningitis in Poland. A SAPS II value of ≥ 54 was predictive of mortality (AUROC 0.81, 95% CI 0.72-0.90).

Risk factors for poor outcome

 * Patient factors
 * Altered mental status
 * Antiepilepticus used
 * Duration of illness (prolonged)
 * Dyspnea
 * Electricity at home (none)
 * Gender (male)
 * Mechanical ventilation required
 * Seizures
 * Temperature (temp > 40°C)
 * Lab factors
 * CSF
 * Serum
 * Glucose (low or high)
 * Pathogen
 * S. pneumoniae worse than N. meningitidis in one study

Glasgow coma scale (1974)
In 723 children with bacterial meningitis in Angola (median age 11 months, IQR 6-29 months), GCS was 81% sensitive and 67% specific (AUROC 0.78) for poor outcome (death, severe neurological sequelae, or bilateral deafness).

Herson-Todd score (1977)
Derived from 73 patients with H. influenzae meningitis in Denver, Colorado and validated in 53 patients from Alaska, mostly with H. inluenzae but also with S. pneumoniae. It has since been used as a prognostic scale in bacterial meningitis of various types. In 723 children with bacterial meningitis in Angola (median age 11 months, IQR 6-29 months), the Herson-Todd scale was 61% sensitive and 69% specific (AUROC 0.69) for poor outcome (death, severe neurological sequelae, or bilateral deafness). In another population of 809 Guatemalan children with meningitis, a score of > 5 was was 70% sensitivity and 82% specific for poor outcome (AUROC 0.76). . In one study, this study was the most predictive (of several scales) for death after meningitis (sensitivity 25%, specificity 92%, AUROC 0.79)

Blantyre coma scale (1989)
In 723 children with bacterial meningitis in Angola (median age 11 months, IQR 6-29 months), BCS was 71% sensitive and 74% specific (AUROC 0.77) for poor outcome (death, severe neurological sequelae, or bilateral deafness).

Oostenbrink score (2002)
Derived from 247 patients with non-Haemophilus meningitis in the Netherlands with AUROC 0.87 (95% CI 0.78-0.96). However, in a validation set the AUROC was only 0.65 (95% CI 0.57-0.72).

Biesheuvel score (2006)
This score uses factors from the Oostenbrink score but uses a regression model for more accuracy. However, it also makes the scale more difficult to apply in clinical practice, so that even though the AUROC was 0.77 it is not so clinically useful.

Simple Luanda scale (2012)
In 723 children with bacterial meningitis in Angola (median age 11 months, IQR 6-29 months),was 71% sensitive and 74% specific (AUROC 0.77) for poor outcome (death, severe neurological sequelae, or bilateral deafness). In a study of 101 children in Sweden, this scale only had AUROC of 0.64 for neurological disabilities.

Bayesian Luanda scale (2012)
Using a Bayesian analysis and including an additional variable of blood glucose, the authors of the Luanda scale were able to improve the sensitivity and specificity of the test for predicting rish of death, severe neurological sequelae, or bilateral deafness (sensitivity 81%, specificity 70%, AUROC 0.82). However, no simple online calculator has been created, which makes the scale too cumbersome for routine use.

Meningitis Swedish Survival Score (MeningiSSS) (2020)
This score was 88% specific and 56% sensitive for a requirement for intensive care (AUROC 0.79) and performed the best compared with 4 other scales. It was very sensitive (100%) for a need for invasive ICP monitoring or management, although only 75% specific (AUROC 0.90).