Siranart N.Chumpangern Y.Pajareya P.Nakaphan P.Kassels T.Chuanchai W.Sattawatthamrong S.Fangsaard P.Tangpradubkiat P.Steinway S.N.Tantitanawat K.Chirapongsathorn S.Mahidol University2026-05-012026-05-012026-01-01American Journal of Gastroenterology (2026)00029270https://repository.li.mahidol.ac.th/handle/123456789/116472Background: – Non-variceal upper gastrointestinal bleeding (NVGIB) represents an urgent medical condition with high mortality. Standard endoscopic therapies (ST) still carry a high rebleeding rate. Novel modalities such as the over-the-scope clip (OTSC) and hemostatic powder (HP) could be promising alternatives, but their efficacy and cost-effectiveness remain to be explored.Methods: – A network meta-analysis and cost-effectiveness analyses were conducted to compare OTSC vs. HP vs. ST. Primary endpoints include 30-day rebleeding and 30-day further bleeding. Markov models with a 12-month time horizon were created across 3 different healthcare settings: the United States (US), Thailand with Boston Scientific products (TH-BOS), and Thailand with China-manufactured alternatives (TH-CHN). Incremental cost-effectiveness ratio (ICER) was evaluated at a willingness-to-pay (WTP) of $100, 000 per quality-adjusted life year (QALY) for the US model and THB150, 000 per QALY for the TH-BOS and TH-CHN models.Results: – Ten RCTs with 1, 120 NVGIB patients were included in the NMA. Lowest 30-day rebleeding was seen in OTSC (P-score:0.996), followed by ST (0.296) and HP (0.208). Similarly, OTSC also achieved lowest 30-day further bleeding (0.906), followed by HP (0.570) and ST (0.024). In CEA, the US model reported OTSC as a dominant approach compared to HP (ICER: –$314) and ST (–$3, 112). In the TH model, OTSC still dominated ST (ICER:–THB28, 423 for TH-BOS and –22, 858 for TH-CHN) and was more cost-effective than HP (THB1, 833 for TH-BOS and THB2, 026 for TH-CHN).Conclusion: – OTSC was superior to HP and ST in terms of efficacy and cost-effectiveness, favoring its use as a first-line endoscopic hemostasis intervention for NVGIB.MedicineArticleSCOPUS10.14309/ajg.00000000000040002-s2.0-1050366822231572024141873865