Antibiotics & adjuvant corticosteroids in management of pneumococcal meningitis: a retrospective case-notes audit

Abstract

Background Pneumococcus remains the most common cause of bacterial meningitis with high morbidity and mortality. Adjuvant corticosteroids with early antibiotics have been shown to reduce the neurological morbidity and mortality respectively and this is reflected in British Infection Society (BIS) guidance. Aim To assess how closely BIS guidelines were followed regarding antibiotic and adjuvant corticosteroid administration in management of pneumococcal meningitis. Methods Newcastle Upon Tyne Hospital case-notes of pneumococcal meningitis from a 7-year-period(2012-2019) were audited. Patients were identified using microbiological records and case-notes. Data was collected on intervals from initial-assessment to commencing antibiotics and corticosteroids. Results Eighteen cases were identified of whom three(17%) presented with the classic triad (fever, meningism, reduced GCS). All patients received appropriate antibiotics: 3/18(17%) within first hour of assessment. The median time to antibiotics was 5h 8mins (range:21-7129min). Eight patients(44%) received antibiotics >6h after assessment. Twelve patients(67%) received corticosteroids; only six(33%) at the recommended dose and duration. Mean time from antibiotics to corticosteroids was 6h1min. Five deaths occurred in the cohort with three attributable to pneumococcal sepsis (all had late presentations). 6/18 had significant neurological sequelae, irrespective of whether they received corticosteroids. Discussion The significant morbidity and mortality of pneumococcal meningitis demands a high index of suspicion. BIS guideline targets are repeatedly not met; long delays exist between assessment and antibiotic and corticosteroid administration. Integrated electronic prescribing and clinical Early Warning Systems have potential to ameliorate this with meningitis-tailored order sets to prompt consideration of meningitis and guide correct prescribing.