Antibiotic treatment delay and outcome in acute bacterial meningitis

Summary

Objectives

To identify to what degree in-hospital delay of antibiotic therapy correlated to outcome in community acquired bacterial meningitis.

Methods

All cases of culture-positive cerebrospinal fluids in east Denmark from 2002 to 2004 were included. Medical records were collected retrospectively with 98.4% case completeness. Glasgow Outcome Scale was used. Multiple regression outcome analyses included the hypothesised factors: delay of therapy, age, bacterial aetiology, adjuvant steroid therapy, coma at admission and the presence of risk factors.

Results

One hundred and eighty seven cases were included. Adult mortality was 33% and the proportion of unfavourable outcome in adults was 52%, which differed significantly from that of children (<18 years) with a mortality of 3% (OR = 15.8, 95% confidence interval: 3.7–67.6) and an unfavourable outcome of 14% (OR = 12.7, CI: 4.3–37.2). Delay of antibiotic therapy correlated independently to unfavourable outcome (OR = 1.09/h, CI: 1.01–1.19) among the 125 adult cases. In the group of adults receiving adequate antibiotic therapy within 12 h (n = 109), the independent correlation between antibiotic delay and unfavourable outcome was even more prominent (OR = 1.30/h, CI: 1.08–1.57). The median delay to the first dose of adequate antibiotics was 1 h and 39 min (1 h and 14 min in children vs. 2 h in adults, p < 0.01), and treatment delay exceeded 2 h in 21–37% of the cases with clinically evident meningitis.

Conclusion

The delay in antibiotic therapy correlated independently to unfavourable outcome. The odds for unfavourable outcome may increase by up to 30% per hour of treatment delay.

Introduction

The mortality in community acquired acute bacterial meningitis has been described as 13–27% in countries comparable to Northern Europe1, 2, 3, 4, 5, 6 and severe sequelae were found in 6–26% of the survivors depending on bacterial aetiology and patient age.3, 7 Early steroid treatment has shown a beneficial correlation to outcome,⁸ whereas co-morbidities and coma have been associated to an adverse clinical outcome.1, 4, 9

In-hospital delay of antibiotic therapy may influence outcome5, 6, 10, 11 and there seems to be a consensus that the prognosis worsens, if initiation of treatment is delayed until a late stage of the disease.1, 6, 10, 12, 13 Recognized textbooks of infectious disease and internal medicine recommend administration of antibiotics within 60 min of arrival to hospital⁷ or no later than 90–120 min,¹² despite the fact that no clinical studies exist to support these specified recommendations. The recent IDSA guidelines recommend administration of antibiotics as soon as possible.¹³ Aronin et al found that delay of antibiotic therapy was associated with adverse clinical outcome, if the patient's condition had progressed to the highest stage of severity,¹ but a general correlation between delayed therapy and clinical outcome was not demonstrated.1, 14 Miner et al suggested such a possible correlation, but the groups compared had a difference in median treatment delay of 5 h.¹⁰ In ICU patients with community acquired pneumococcal meningitis, in-hospital antibiotic delay exceeding 3 h was identified as an independent risk factor for mortality.¹¹ In adult meningitis, one study demonstrated a sharp rise in the case-fatality rate after 6 h of antibiotic therapy delay, where treatment delay correlated to the absence of the meningitis symptom triad and to the need of further patient transferral for head CT scan.⁶

Studies have demonstrated a large variation in antibiotic delay in bacterial meningitis with a lower treatment delay among children.10, 15, 16, 17, 18, 19 The median treatment delay in various publications was found to be between 1 h and 5.5 h.6, 10, 15, 16, 17, 18, 19 Talan et al described in an analysis of an emergency room department that 90% of the total delay of antibiotics occurred after the initial physician encounter,¹⁶ and that prompt administration of antibiotics was associated with general clinical signs of severe disease rather than more specific symptoms of meningitis.¹⁷

The intention of the present study was to do an outcome analysis focusing on in-hospital antibiotic delay and the mentioned hypothesised factors of importance in a complete population of patients with community acquired bacterial meningitis.