Introduction Non-variceal upper GI bleeding (NVUGIB) is a common problem with significant morbidity and 30-day mortality. Hemospray (HS) is a recently introdued endoscopic treatment for NVUGIB. It is a proprietary hemostatic spray propelled by carbon dioxide under pressure, which can achieve rapid hemostasis when it comes in contact with blood. The drying effect of the spray and activation of clotting, in addition to pure tamponade, underlies its mechanism of action.

Method A retrospective single centre study was conducted of patients in whom HS was used as monotherapy or part of dual therapy for control of NVUGIB by different endoscopists. The aim of the study was to evaluate efficacy of HS for treatment of NVUGIB.

Results A total of 19 patients; mean age 68 years (range 42 to 89 years), 16 males; with NVUGIB were treated with HS between Oct 13 to Oct 14. Clinical presentation included hematemesis and/or melena in 18 (95%) and anaemia in 1 (5%). The mean Rockall score (post endoscopic) was 6.5 and mean Blatchford score was 11.6. Medication review found use of aspirin in 3, NSAID’s in 3 and warfarin in 1 patient. Investigations prior to endoscopy revealed a mean haemoglobin of 79.4 gm/L (range 51 to 136), thrombocytopenia (platelets <150) in 6 (32%) and coagulopathy (INR ≥ 1.3) in 4 (21%) patients. Endoscopic findings included bleeding from oesophageal banding induced ulceration in 2 (10.5%), peptic ulcers in 10 (53%); Forrest class Ib 6, IIa 3 and IIb 1; bleeding tumours in 4 (21%), post-sphincterotomy bleed in 2 (10.5%) and hemobilia in 1 (5%). HS was administered as monotherapy in 8 (42%) patients, 8/8 achieved immediate haemostasis, however 2/8 had recurrent bleed (1 was treated with adrenaline injection and 1 patient was too unwell for further intervention). HS was used as second modality (post adrenaline injection) in 11 (58%) patients where thermal modalities or clips could not be applied for technical reasons or were not considered at the discretion of the endoscopists. 2/11 failed to achieve immediate hemostasis (1 needed immediate laprotomy and 1 died next day) while 3/11 had recurrent bleed (2 needed laprotomy and 1 was treated with OGD and further HS). Average length of stay was 34.5 days (range 5 to 152) and 30 day mortality was 21%.

Conclusion The data shows that when HS was used as monotherapy immediate hemostasis was achieved in 100% while re-bleeding occurred in 25% cases. As part of dual therapy HS achieved immediate hemostasis in 82% of cases while re-bleeding was seen in 27% of cases. This can be compared to the conventional endoscopic hemostatic methods which achieve overall success rates of 85–95% for immediate hemostasis with re-bleeding in 5–10% cases. The high re-bleeding could be accounted for by the more complex pathologies (tumours, banding induced ulcers) treated with HS.

Disclosure of interest None Declared.