Nonoperative Versus Neurosurgical Treatment of Brain Abscess: An Emulated Trial Nested Within a Nationwide, Population-Based Cohort

Résumé fourni par la source

BACKGROUND: Uncertainty about the role of neurosurgery in brain abscess treatment is reflected in surveys and observational studies. METHODS: This was a nationwide population-based cohort study of all adults (≥18 years) diagnosed with brain abscess in Denmark from 2007 until 2023. Inverse probability weighting was applied to balance covariates according to neurosurgical treatment strategy. Outcomes included risks of mortality, rupture, and unfavorable outcome (Glasgow Outcome Scale score 1-4) 6 months after discharge and were assessed using modified Poisson and Cox regression with 95% confidence intervals (CIs). Unbalanced covariates were included in regression models for doubly robust estimations. RESULTS: The cohort comprised 558 patients (median age, 59 [interquartile range, 49-69] years; 367/558 [66%] male). A nonoperative strategy was assigned to 234 of 558 (42%) and neurosurgery to 324 of 558 (58%). Predisposing meningitis, diameter, and deep location of abscess were included in doubly robust estimations. Analyses of treatment strategy showed that a nonoperative approach was associated with adjusted relative risks (95% CIs) of 2.47 (1.50-4.04) for mortality, 2.25 (1.26-4.01) for rupture, and 1.24 (.95-1.61) for unfavorable outcome compared with neurosurgery. Additionally, 85 of 234 (36%) patients assigned to the nonoperative strategy required subsequent neurosurgery. Compared with neurosurgery at any time, maintained nonoperative treatment was associated with adjusted hazard rate ratios (95% CIs) of 1.53 (.84-2.76) for mortality, 2.00 (1.17-3.42) for rupture, and 1.62 (1.12-2.34) for unfavorable outcome. CONCLUSIONS: Nonoperative strategy was associated with twice the risks of mortality and rupture, and one-third of patients required subsequent neurosurgery. These results support recommendations for neurosurgical drainage of brain abscess.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.