548 - Identifying Social Barriers to Pulmonary Arteriovenous Malformation Care: Single-Center Analysis of Procedural Utilization

Purpose: To determine the effects of social determinants of health (SDoH) on access to care in patients screened for Hereditary Hemorrhagic Telangiectasia (HHT) referred for endovascular management of pulmonary arteriovenous malformations (pAVMs) at an HHT Center of Excellence (CoE).

Materials and Methods: A retrospective analysis of patients with a documented history of pAVMs presenting for initial screening at the CoE between 1/1/2017 and 7/1/2024 was conducted. Demographics, including age, sex, insurance status, smoking history, and CDC Social Vulnerability Index (SVI) from 5-digit home ZIP code were collected. All patients underwent >1 year of follow-up from initial screening. The primary outcome was the presence of ≥1 cancelled endovascular interventions (pulmonary arteriography and/or pAVM embolization) documented within one year of screening, and the secondary outcome was time from screening to first performed intervention (time-to-treatment) at the CoE. Exploratory, univariate regression and Cox proportional hazards were performed in RStudio (Posit, Boston, MA).

Results: 310 patients underwent screening, and 277/310 (89.4%) had suspected/confirmed HHT, defined as ≥2 Curacao criteria. Of these, 108/277 (39.0%) had pAVMs, of which 62/108 (57.4%) were referred for intervention. Eleven patients with a prior pAVM embolization at the CoE and three pediatric patients (age < 18 years at screening) were excluded from subsequent analyses, leaving 48 patients (median age 49.0 years [IQR 33.8-60.5]) in the final cohort.


Twelve patients (25%) had ≥1 cancellation within 1 year of follow-up (median age 52.5 years [IQR 42.5-61]; 75% female). The first intervention resulted in 40/42 (95.2%) technically successful embolization sessions, 4 diagnostic arteriograms, and two patients with cancellation only. No baseline variables, including SVI (median 0.451 [IQR 0.342-0.656]), were significant predictors of cancellation (p >0.05). Likewise, SVI was not a significant predictor of time-to-treatment on univariate Cox proportional hazards model (median time 121 days [IQR 65.5-194.5]; p=0.61, HR 1.50 [95% CI 0.31-7.30]).

Conclusion: Preliminary findings demonstrate a complex relationship between SDoH and access to care for patients treated for pAVMs. SVI was not associated with cancellations or time-to-treatment, though further investigation is warranted, especially given that HHT remains underdiagnosed and undertreated in marginalized populations. Future studies should identify roles of area-level SDoH on quality of life and long-term clinical outcomes following embolization.