Document resource
Background Middle meningeal artery embolization (MMAe) is increasingly used to treat chronic subdural hematoma (cSDH). Both particle and liquid embolic agents are used in practice. We evaluate the cost-effectiveness of particle versus liquid embolic agents for standalone MMAe in cSDH.Methods We developed a Markov cohort model with six health states (resolved, recurrent cSDH, surgical rescue, post-surgical disability, periprocedural complication, and death) over a 5-year horizon using monthly cycles comparing polyvinyl alcohol (PVA), Onyx, and n-butyl cyanoacrylate (n-BCA). The model reflected an older, mildly frail cSDH population. Transition probabilities were informed by meta-analysis estimates, utilities drawn from cSDH-specific literature, and costs from institutional data. Probabilistic sensitivity analysis with 1,000 Monte Carlo simulations calculated incremental net monetary benefit at a willingness-to-pay threshold of $100,000/QALY.Results Over 5 years, PVA yielded 1.82 quality-adjusted life-years at a cost of $17,084 per patient, compared with $22,687 for Onyx and $24,876 for n-BCA. Quality-adjusted survival was similar across all three strategies, but total cost was lowest with PVA. In threshold analysis, Onyx would need a greater than 3.2 percentage-point improvement in surgical rescue rate versus PVA to become cost-effective, while n-BCA would need a greater than 5.0 percentage-point improvement. Alternatively, Onyx would need its cost reduced by approximately 57% (from $10,630 to ≤$4,530), and n-BCA by approximately 54% (from $14,115 to ≤$6,515), to match the value of PVA.Conclusion In this model, PVA was the most cost-effective embolic agent for standalone MMAe in cSDH, offering equivalent quality-adjusted survival at substantially lower cost than liquid embolics. Liquid embolic agents may become competitive only if they achieve meaningfully better clinical outcomes or substantial price reductions. These findings support careful consideration of embolic agent selection in value-based care and suggest that longer-term data may further clarify whether durability differences alter these results.Disclosures A. Baskaran: None. E. Goicoechea: None. R. Morsi: None. M. Hurley: None. T. Kass-Hout: None. R. Naylor: None.Abstract E-036 Figure 1Abstract E-036 Figure 2