Aims “Watermelon Stomach” or Gastric Antral Vascular Ectasia (GAVE), is a relatively uncommon
cause of Upper GI Bleeding, accounting for roughly 4% of non-variceal UGIB. Although
previously considered rare, GAVE is now diagnosed more often and according to recent
estimates, up to 14% of patients with cirrhosis may be affected. GAVE can be managed
by multiple treatment modalities including Argon Plasma Coagulation (APC), Radiofrequency
Ablation, Endoscopic Band Ligation (EBL), YAG laser coagulation, cryotherapy, sclerotherapy/heater
probe and surgical antrectomy with APC being the treatment of choice in most cases.
Although APC is associated with high recurrence rates (40–100%) leading to increased
healthcare burden and patient discomfort. EBL has been studied as an alternative with
better outcomes in the reported literature. Our study aims to elucidate if the use
of EBL would be cost-effective as a treatment option for GAVE compared with APC in
the United States.
Methods We conducted a cost-effectiveness analysis comparing APC, EBL and a combination of
both for GAVE, using a decision tree model with pooled data from 503 patients (277
in APC group, 206 in EBL group and 20 in APC+EBL group) from a network meta-analysis
of RCTs, prospective and retrospective studies that looked at treatment options for
GAVE. Cost of procedure, hospitalization, pRBC transfusion and post-procedure bleeding
were derived through CPT codes, average institutional reimbursements, and published
literature. Incremental Cost-Effectiveness Ratio (ICER) and Net Monetary Benefit (NMB)
was calculated comparing different treatments. Probabilistic Sensitivity Analysis
was conducted to account for real world uncertainties. We assumed the Willingness-To-Pay
(WTP) of $100,000/QALY. Analysis was performed using TreeAge Pro Healthcare 2024 [1].
Results Our study revealed EBL was associated with lower costs ($2,442) and higher efficacy
(0.06 QALY) with ICER of – 40,695 $/QALY and a higher NMB ($71,201 vs $62,760) compared
to APC. Threshold analysis revealed that EBL is cost effective compared with APC when
Probability of Bleeding Recurrence is<42.6%, Number of EBL sessions is<5.426 and Cost
of EBL procedure is<$6337.254. One-way Sensitivity Analysis revealed cost-effectiveness
of EBL is sensitive to variations in probability of recurrent bleeding with EBL and
number of sessions for bleeding cessation with EBL. At 40% probability of recurrence
with EBL, NMB is nearly $500 which reduces to -$2,000 with probability increasing
to 50%. Also when the number of sessions for EBL increases from 5 to 6 the NMB changes
from $1000 to -$2000 making it not cost-effective at the assumed WTP. In Probabilistic
sensitivity analysis, EBL was the optimal strategy for 98.50%, APC for 1.40% and APC+EBL
for 0.10% scenarios.
Conclusions At a widely accepted willingness-to-pay threshold of $100,000 per QALY, EBL may be
a cost-effective strategy compared to APC for patients with gastric antral vascular
ectasia, primarily due to fewer required treatment sessions and lower recurrence rates.
