Background and Aims: EUS-guided coil and glue of gastric varices (GV), targeting the submucosal variceal component (SVC), has always been the conventional strategy. In cases in which the feeder vessel can be delineated, targeting the feeder alone (EUS-guided feeder vessel obliteration [EUS-F]) instead of the SVC seems logical and has never been systematically compared. Methods: This study included patients with GV with a discernible feeder vessel localized on EUS and characterized as per the Sarin and Arakawa classification. Patients were divided into 2 arms, in which either the feeder or the SVC was targeted and obliterated. The primary outcome (reintervention rates) and secondary outcomes (amount of glue, number of coils, obliteration, and bleeding rates) at 6 months were analyzed. Results: A total of 135 patients (EUS-F, 32; EUS-SVC, 103) were evaluated. Gastroesophageal varices type 2 (81; 60%) and type 1 Arakawa (97; 71.85%) were the most common GV types. The strategy of EUS-F showed lower requirements of both coil (P = .0001) and glue (P = .0001), not only at the index procedure but also overall requirement of coils (1.94 +/- 0.91 vs 2.79 +/- 1.76; P = .001) and glue (1.88 +/- 0.64 vs 2.83 +/- 1.70 mL; P < .0001) at 6 months. Overall reintervention rates at 6 months were significantly lower in the EUS-F arm (3.1% vs 17.5%; P = .04), along with the number of EUS sessions needed for complete GV obliteration (P = .02). Conclusions: For EUS-guided angioembolization of GV, targeting the feeder vessel, when feasible and discernible, over SVC (conventional method) results in significantly lower requirements of coil and glue, with significantly lower reintervention rates, which might translate to a more cost-effective strategy.
Background and Aim Gastrointestinal endoscopy is resource-intensive, but no carbon footprint data have been reported from a Japanese facility, and published estimates differ widely in system boundary. We quantified the operational carbon footprint of an endoscopy unit at a Japanese regional hospital and identified which emission sources are amenable to unit-level intervention.Methods In this retrospective single-center study, we analyzed 794 consecutive procedures performed over 5 months at a 150-bed regional public hospital. Emissions from electricity, patient travel, medical waste, consumables, reprocessing chemicals, medical gas, and water were estimated according to the Greenhouse Gas Protocol using Japanese national emission factors and published life-cycle data. Patient travel was aggregated at the procedure level under two scenarios: procedure-day travel only (Scenario A) and visit-count-corrected travel (Scenario B). Sensitivity analyses addressed transport mode, air-conditioning load factor, and waste composition.Results The operational carbon footprint was 6537.5 kg CO2e (CO2e) (8.23 kg CO2e per procedure) in Scenario A and 8042.1 kg CO2e (10.13 kg CO2e per procedure) in Scenario B. In Scenario B, electricity was the largest source (38.8%), followed by patient travel (35.3%) and medical waste (13.5%). Air conditioning accounted for 69.3% of electricity use. Waste generation was 0.80 kg per procedure, of which suction fluid bags represented 36.7%.Conclusions Electricity, dominated by air conditioning, was the largest operational emission source and is directly modifiable by the unit. Patient travel was of comparable magnitude but reflects healthcare access geography rather than endoscopic practice.
