Early versus late diagnosis in community-acquired bacterial meningitis: a retrospective cohort study.

OBJECTIVES

To examine clinical characteristics and outcome of patients with late diagnosis of community-acquired bacterial meningitis (CABM).

METHODS

We conducted a chart review of all adults with proven CABM in three centres in Denmark from 1998 through to 2014. Patients were categorized as early diagnosis of CABM immediately on admission, or late diagnosis if CABM was not listed in referral or admission records and neither lumbar puncture nor antibiotic therapy for meningitis was considered immediately on admission. We used modified Poisson regression analysis to compute adjusted relative risks with 95% CIs for predictors of late diagnosis and in-hospital mortality.

RESULTS

A total of 113/358 (32%) patients were categorized as late diagnosis demonstrating a variety of tentative diagnoses of which 81/113 (72%) were non-infectious. We observed several statistically significant baseline differences (p <0.05) in patients with late versus early diagnosis including age >65 years (56/113, 50% versus 67/245, 27%), neck stiffness (35/97, 36% versus 183/234, 78%), concomitant pneumonia (26/113, 23% versus 26/245, 11%), and meningococcal meningitis (6/113, 5% versus 52/245, 21%). These variables remained statistically significant in multivariate analysis. Moreover, late diagnosis was associated with increased in-hospital mortality (41/113, 36% versus 43/245, 18%; adjusted relative risk 1.7, 95% CI 1.2-2.5).

CONCLUSIONS

Late diagnosis of CABM was common and patients were admitted with mostly non-infectious diagnoses. Absence of neck stiffness did not rule out CABM and special attention should be given to patients with pneumonia and the elderly. Late diagnosis was associated with incorrect patient management and increased mortality.