Incidence, outcomes and risk factors for mortality of symptomatic upper gastrointestinal hemorrhage in surgical critical ill patients (THAI SICU study)

Abstract

© 2016, Medical Association of Thailand. All rights reserved. Objective: This multicenter university-based study reports the incidence, outcomes and defined risk factors for mortality of upper gastrointestinal hemorrhage (UGIH) patients in the surgical intensive care units (ICU) patients in Thailand. Material and Method: This is part of a multicenter prospective observational study in the ICU in Thailand (THAI-SICU study). Patients who had a clinical presentation of upper gastrointestinal hemorrhage or an endoscopic diagnosis from April 2011 to January 2013 were enrolled into this sub-study. Results: A total of 4,652 patients were analyzed. Fifty-five patients (1.18%) had symptomatic UGIH during ICU admission. The median age (interquartile range, IQR) was 72 (63-78) years old and the median APACHE II score (IQR) was 17 (13-22). In a comparison between the UGIH patients who survived and those who non-survived APACHE II score were higher in the non-survivors. The ICU mortality rate and 28-day mortality rate in these patients were 30.91% and 40%, respectively. In multivariable model, UGIH was significantly associated with 28-day mortality [adjusted odds ratio, OR, (95% confidence interval, CI): 1.99 (1.02 to 3.88); p = 0.043] and ICU length of stay [adjusted coefficient (95% CI): 9.36 (8.03 to 10.70); p<0.001]. Regarding the exploratory model, the significant risk factors for non-survived of UGIH patients were coagulopathy especially platelet count <50,000 [OR (95% CI): 3.96 (1.07-14.67); p = 0.039] and INR >1.5 [5 (1.04-23.98); p = 0.044], renal failure [6.48 (1.37-30.61); p = 0.018], APACHE II score [1.11 (1.02-1.22); p = 0.020] and vasopressor use [5.78 (1.6- 37.18); p = 0.013]. Conclusion: The incidence of symptomatic UGIH in the THAI-SICU study was 1.18% and UGIH was associated with higher 28-day mortality rate and prolonged ICU length of stay. The risk factors for mortality were coagulopathy, renal failure, APACHE II score and vasopressor use.