duodenoscopy (OGD). Those who did, had higher red cell transfusion requirements than those who did not (median 1050 [IQR 150-1650] vs 0 [IQR 0-675]ml; p=0.019). 67% (8/12) had oesophageal or peptic ulceration but only 25%(2/8) required endotherapy. There was one re-bleed which required further endotherapy and one required embolisation. Overall mortality rate in the UGIB cohort is 41.3%. Increased age, heart rate and urea levels were significantly associated with death but not the UGIB status.

more »

... S was prolonged in the UGIB group compared to the non-UGIB group (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27) vs 6 [IQR 3-14] days; p<0.001). Subgroup analysis of UGIB patients who underwent OGD compared to those managed conservatively showed no significant difference in mortality (44%vs 33%; p=0.735) or vs 16 [IQR 8-27] days; p=0.349). Conclusions The incidence of concurrent UGIB in patients with COVID-19 infection is low and our data suggests that UGIB does not increase the mortality rate in patients with COVID-19 but does increase LOS. Endoscopic intervention did not appear to affect mortality or LOS. Although numbers are small, this data suggests selection criteria used for endoscopy was justified.