This meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (PRISMA checklist in Supplementary Figure S1) and the Cochrane Handbook for Systematic Reviews of Interventions (Version 6.5) (15).

The inclusion criteria were formulated using the PICOS framework (16). Population: Adult patients with unilateral subacute or chronic subdural hematoma undergoing surgical evacuation. Intervention: Adjunctive middle meningeal artery embolization (MMAE). Comparator: Surgical evacuation without adjunctive MMAE. Outcomes: Primary outcomes included recurrence or reoperation rates at predefined time points (90-days). Study Design: Randomized controlled trials (RCTs) with sufficient reporting of outcomes stratified by surgical status.

Studies were excluded if they involved bilateral subdural hematomas, focused on conservative management without surgery, or provided insufficient data for meta-analysis.

Recurrence is defined as the reappearance or progression of subdural hematoma (SDH), confirmed by imaging or clinical symptoms, within 90 days of the index treatment. This includes an increase in hematoma thickness exceeding 10 mm, or an increase of more than 3 mm compared to baseline, with or without associated neurological deterioration. Reoperation refers to any repeat surgical intervention performed within 90 days to manage hematoma recurrence or progression, as determined by clinical or imaging criteria.

A comprehensive literature search was conducted using PubMed, the Cochrane Library, and Google Scholar. The search covered all articles published from database inception to November 2024. Search terms included a combination of keywords and MeSH terms for chronic subdural hematoma, MMAE, and surgical treatment. The detailed search strategy is provided in Supplementary Table 1.

Two independent reviewers screened titles, abstracts, and full-text articles for eligibility. Disagreements were resolved by a third reviewer. The PRISMA flow diagram (Figure 1) summarizes the study selection process.

Data were independently extracted using a predefined data collection form. Extracted data included: Study characteristics (authors, publication year, study design, country, number of centers), population demographics (sample size, age, sex distribution, surgical details), intervention characteristics (timing and technique of MMAE), and outcomes (recurrence, reoperation).

The Cochrane Risk of Bias 2.0 tool was used to assess the methodological quality of included studies (17). Discrepancies in risk of bias ratings were resolved through consensus. The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach was used to evaluate the certainty of evidence (18).

Data were analyzed using Review Manager (RevMan, Version 5.4). Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for dichotomous outcomes using a random-effects model. To assess statistical heterogeneity and inconsistency, the χ² test and I² statistic were utilized, with an I² value of 50% or higher indicating substantial heterogeneity. The relative weight of each study, determined by its sample size, was taken into account when estimating the treatment effects.

The PRISMA flow diagram details the systematic review process. A total of 1,804 records were identified through PubMed (421), Cochrane Library (23), and Google Scholar (1,360), with 351 duplicates removed. After screening titles and abstracts, 1,699 records were excluded. Of 105 full-text articles assessed, 103 were excluded for insufficient data or non-eligibility. Five RCTs were identified. The RCT by Fiorella et al. (12) was excluded because no recurrence or reoperation data stratified for surgically treated patients with or without MMAE were reported. The RCTs by Debs et al. (13) and Lam et al. (14) were excluded because bilateral SDHs were included in their study cohorts. Ultimately, 2 studies by Liu et al. (10) and Davies et al. (11) met the inclusion criteria for the meta-analysis. Figure 1 illustrates the process.

The EMBOLISE and MAGIC-MT trials evaluated middle meningeal artery embolization (MMAE) for chronic and subacute subdural hematoma (cSDH) (10, 11). Both studies excluded patients with bilateral subdural hematomas, focusing solely on unilateral symptomatic cases.

The EMBOLISE trial (11) conducted in multiple centers, included patients aged 18–90 years with symptomatic subacute or chronic SDH requiring surgical evacuation. Exclusion criteria included life expectancy <1 year, pre-existing severe neurologic impairment (Markwalder score ≥3), and significant disability (mRS ≥4). Patients were randomized 1:1 to receive surgery with adjunctive MMAE using the Onyx embolic system or surgery alone. MMAE was performed within 48 hours of randomization and targeted hematoma recurrence requiring reoperation as the primary endpoint. Fifty-three-point five percent of patients receiving adjunctive MMAE underwent burr-hole surgery, while 46.4% underwent craniotomy. Among those treated with surgery alone, 51.0% underwent burr-hole evacuation, and 49.0% underwent craniotomy. Subdural drains were used in 95.9% of patients receiving MMAE and 96.0% of those without MMAE.

The MAGIC-MT trial (10), a multicenter study in China, enrolled patients aged ≥18 years with symptomatic nonacute SDH and mass effect. Exclusions included bilateral SDH, craniotomy requirements, or emergency hematoma evacuation. In the EMBOLISE trial (11), 53.6% vs. 51.0% of patients in the treatment and control arms, respectively, were treated via burr-hole and 46.4% vs. 49.0% via craniotomy, whereas every patient in the MAGIC-MT (10) study underwent burr-hole evacuation. Subdural drains were inserted in 95.9% of EMBOLISE cases, while MAGIC-MT (10) only mandated drainage in the protocol without reporting the proportion actually placed. Participants were randomized to receive MMAE with Squid embolic agent followed by conservative management or burr-hole drainage, or usual care without MMAE. Primary endpoints were recurrence or progression of SDH within 90 days. MMAE was performed preoperatively in surgical patients. Patients who underwent craniotomy were excluded and the use of subdural drains was mandated in all cases. Table 1 summarizes the characteristics of the included studies.

The risk of bias assessment for the two RCTs shows overall low risk of bias across most domains. Both studies demonstrated robust random sequence generation, allocation concealment, and complete outcome data reporting. However, performance bias was identified due to the lack of blinding of participants and personnel in both trials. Despite this, the blinding of outcome assessment and selective reporting were adequately addressed, ensuring high methodological quality. Figure 2 summarizes the risk of bias assessment.

The meta-analysis included data from two randomized controlled trials (RCTs), comprising a total of 965 patients. Four hundred and seventy-eight (478/965; 49.5%) were allocated to the MMAE group, and 487 (487/965; 50.5%) to the control group. Twenty-five (25/478; 5.2%) in the MMA group had a recurrence, whereas 45 (45/487; 9.2%) in the control group had a recurrence. The odds ratio (OR) for recurrence or progression at 90 days was 0.52 (95% CI: 0.17–1.59; p = 0.25), indicating no statistically significant difference between MMAE and control groups. Heterogeneity was moderate (I² = 77%, p = 0.04), suggesting variability between the included studies. Figure 3 illustrates the forest plot summarizing the results.

This section presents the results of a meta-analysis evaluating the impact of MMAE on recurrence or progression requiring reoperation at 90 days after surgical evacuation of unilateral cSDH. The meta-analysis included the data from the two randomized controlled trials (RCTs), comprising a total of 965 patients. Four-hundred and seventy-eight (478/965; 49.5%) were allocated to the MMAE group, and 487 (487/965; 50.5%) to the control group. Twelve (12/478; 2.5%) in the MMA group underwent reoperation, and 29 (29/487; 6.0%) in the control group. The OR for recurrence or progression requiring reoperation at 90 days was 0.41 (95% CI: 0.20–0.82; p = 0.01), indicating statistically significant reduction in recurrence with MMAE. Heterogeneity was low (I² = 0%, p = 0.37), suggesting no variability between the included studies. Figure 4 illustrates the forest plot summarizing the results.

To ensure adequate reliability, we implemented three measures to assess potential publication bias. First, an extensive literature search strategy was employed. Second, the trials included in this meta-analysis strictly adhered to the predefined inclusion and exclusion criteria. Lastly, publication bias was analyzed using funnel plots (Figures 5) and statistical tests for key endpoints (recurrence and reoperation). The data points fell within the inverted funnel, suggesting minimal publication bias in the analysis of these endpoints.

The GRADE assessment demonstrates high certainty for reduced reoperation risk (relative risk: 0.41, absolute reduction: 34 fewer per 1,000 patients), indicating a significant clinical benefit of middle meningeal artery embolization. However, moderate certainty supports recurrence reduction due to notable heterogeneity (I² = 77%) and wide confidence intervals (OR: 0.52, CI: 0.17–1.59). Both outcomes are deemed critical for clinical decision-making. Supplementary Table 2 summarizes the GRADE approach.