(1242) Cost-Effectiveness of a Multivessel Embolization Strategy for Management of Placenta Accreta Spectrum: A Decision-Analytic Model

To estimate the cost-effectiveness of our institution’s surgical strategy using multivessel embolization (MVE) followed by immediate hysterectomy for managing placenta accreta spectrum (PAS) compared to non-MVE approaches.

Study Design:

We developed a decision-analytic model in TreeAge Pro 2025 to compare a standardized MVE-based strategy with immediate hysterectomy for PAS to alternative immediate hysterectomy approaches using a healthcare system perspective. Primary outcome was cost-effectiveness of MVE-based strategy. Primary exposure was prenatal suspicion of high-grade PAS with planned immediate hysterectomy. Costs included operative time, surgical approach, length of stay, complications, and reoperation/readmission. Estimates were obtained from institutional data and the literature, and all parameters characterized as distributions. The model was run probabilistically using a 10,000-trial microsimulation (sampling from the age distribution of PAS), and 5000 samples from each of the input parameters. We used willingness-to-pay threshold of $100,000 per life-year gained. Additional deterministic sensitivity analyses included the relative risk (RR) of prolonged hospital stay (1.0–2.0) with alternative approaches and costs associated with use of resuscitative endovascular balloon occlusion of the aorta (REBOA) ($0–$25,015, with the low vs high range corresponding to probabilities from no vs all patients undergoing REBOA).

Results:

In the base case, MVE had lower mean total costs than alternatives ($32,704 [95% CI: $32,544–$32,864] vs $48,089 [95% CI: $47,979–$48,198]), with similar discounted life expectancy (25.49 years). Higher OR costs ($19,191 vs $11,513) were offset by fewer >5-day hospitalizations (11.9% vs 23.9%). MVE was cost-saving in 97.9% of simulations, and cost-effective in 88.3% of simulations.

Conclusion:

Higher OR costs associated with an MVE-based strategy for high grade PAS may be more than offset by shorter associated hospital length of stay. Future analyses adjusting for potential patient heterogeneity and using larger sample sizes are needed.