Background
The objective was to explore factors associated with early outcomes in patients with spinal dural arteriovenous fistulae (sDAVF) treated by open surgery (Open) or endovascular procedures (Endo).

Methods
Data sources: MEDLINE, Web of Science, and Ovid to March 9, 2024. Study selection: adult sDAVF cohorts with ≥11 patients reporting on at least one of the outcomes of interest. Data extraction and synthesis: PRISMA guidelines were used to screen studies/extract data. Fistula closure rates, complication rates, and prevalence of the affected spine segments were analysed based on summary data. The post- vs. pre-procedural difference in clinical disability was based on individual patient data. Main Outcome(s) and Measure(s): (1) fistula closure rate; (2) early complications rate; (3) clinical disability; (4) prevalence of sDAVF across spine segments/levels.

Results
We identified 115 cohorts. Odds of closure (106 reports on Open, 82 on Endo, adjusted for covariates) were higher with Open vs. Endo (OR = 7.68, 95%CI 5.48–11.0). Odds of complications (59 reports on Open, 48 on Endo, adjusted) were similar for Open vs. Endo (OR = 1.02, 0.77–1.35, prediction 0.77–1.35). With adjustment (21 reports with individual patient data, 288 Open and 134 Endo procedures), reduction in disability scores was larger with Open vs. Endo (difference= −0.55, 95%CI −0.95, −0.15), consistently across the spinal segments, age, and sex. All estimates were resistant to bias (E-values = 4.99, 2.00, and 2.70).

Conclusions
Compared to Endo, Open is more likely to result in fistula closure and reduction of disability, with a similar probability of complications.