Diagnostic Stewardship and WHO AWaRe Antibiotic Exposure in Intracranial Suppurative Infections: A 10-Year Cohort from West Africa

Abstract

Background: In sub-Saharan Africa, where mortality attributable to antimicrobial resistance is the highest worldwide and only 1% of laboratories perform bacteriology, the quality of antibiotic therapy for intracranial suppurative infections has never been described. We hypothesised that the principal stewardship failure in such settings lies not in antibiotic choice but in the production of the microbiological information required to adapt it.

Methods: Single-centre observational study of 150 consecutive admissions (148 patients) managed between 2016 and 2026 at Bouaké Teaching Hospital, Côte d’Ivoire. We built a five-step microbiological cascade of care, classified every regimen by WHO AWaRe category and by a declared spectral coverage index, and analysed spectrum-pathogen concordance in both directions. Determinants were modelled with Firth penalised logistic regression using profile-likelihood intervals, bootstrap internal validation, and eleven sensitivity analyses.

Findings: Sampling occurred in 77.3% of admissions, culture was positive in 33.3%, and documented adaptation in 22.7% (95% CI 16.7-30.0). At least one Watch antibiotic featured in 92.7% of regimens; no Reserve antibiotic was used. Vancomycin was prescribed three times, and antipseudomonal coverage nineteen times, as often as the corresponding pathogens were isolated; conversely only 9 of 17 documented S. aureus were covered, all eight failures occurring in community-acquired disease. Access to sampling depended almost exclusively on surgery (OR 5.61; 2.22-14.56). No trend emerged over ten years, and unsampled patients received the longest courses (7.15 vs 6.35 weeks; p = 0.005, Kruskal-Wallis across the three groups).

Interpretation: Spectrum breadth appeared to reflect clinical uncertainty rather than documented ecology, and did so asymmetrically. Our findings support the hypothesis that strengthening diagnostic stewardship may represent a foundational component of antimicrobial stewardship in neurosurgical infections in resource-limited settings. The analytical framework we propose requires multicentre validation.