Incidence and Predictors of GI Bleeding Among Septic Shock Patients of United States: A 10-Year Analysis of Nationwide Inpatient Population

Abstract

SESSION TITLE: Sepsis & Septic Shock SESSION TYPE: Original Investigation Slide PRESENTED ON: Sunday, October 29, 2017 at 01:30 PM - 03:00 PM PURPOSE: Gastrointestinal bleeding (GIB) complicating septic shock(SS) presents a therapeutic challenge in the intensive care units. Contemporary large-scale data regarding utilization, length of stay, and cost outcomes of this association are lacking. The purpose of this study is to analyze trends, outcomes and risk factors associated with GIB in patients with SS. METHODS: We queried the Healthcare Cost and Utilization Project's Nationwide Inpatient Sample from 2003 to 2012 we identified all adult patients aged ≥18 years hospitalized for septic shock by the following criteria previously validated by recently presented data of the same database (BMC Infectious diseases 2016 10.1186): 1) primary ICD-9 diagnosis of infection associated with sepsis plus vasopressor use, or 2) primary ICD-9 diagnosis of infection associated with sepsis plus a non-primary diagnosis of septic shock, or 3) primary ICD-9 diagnosis of septic shock. We compared the baseline characteristics and outcomes among patients with SS plus GIB to patients with SS without GIB. Measured outcomes included length-of-stay, mortality and cost of hospitalization. Comorbidities were assessed with a univariate logistic regression to determine associated risk factors for GIB among SS patients. A multivariate analysis was also performed to address any confounders. RESULTS: The weighted sample size from 2003 to 2012 was 119684 admissions for SS. Among them, 6571(5.4%) patients were found to have a GIB. Mean age of the SS population with and without GIB was Age (mean/SEM) {70.85 (0.43) vs. 67.43 (0.13) P<0.001} respectively. The incidence of GIB among black (6.8%) septic patients was higher compared to white (5.3%) or Hispanic (5.9%) P<0.001. No significant difference in the occurrence of GIB between males and females respectively {5.6% vs. 5.7% P=0.653}. Patients with SS and GIB compared to patients with SS and no GIB were found to have longer length of stay {20.56 (0.61) vs. 15.76 (0.13) P<0.001}, higher mortality {54% vs. 45% P<0.001} and higher admission costs (mean/SEM) {192524.89 (7378.20) vs. 142688.55 (1336.65) P<0.001}. Univariate analysis demonstrated that major comorbid conditions associated with an increased incidence of GIB in septic patients included peptic ulcer disease and cirrhosis which both had significant odds ratios {1.56 and 1.709, P=0.016 and 0.046 respectively}. Gastroesophageal reflux disease was found to be associated with a lower incidence of GIB {OR: 0.57, P=0.0008}. The cause of sepsis (pneumonia, urinary tract infection or abdominal infections) was not a significant distinguishing factor for the incidence of GIB in SS. CONCLUSIONS: GIB among SS patients is associated with 9% increase in mortality compared to no GIB, with increased length of stay and cost burden. The risk of GIB in Septic Shock patients is perhaps more related to the baseline patient characteristics and comorbid illness than to the etiology of sepsis. CLINICAL IMPLICATIONS: To the author's knowledge this is the first nationwide extensive study analyzing the association of GIB and SS in US inpatient population. Awareness of the risk factors and association of GIB among SS patients may help to improve the outcomes by institution of agressive prophylactic measures among high risk patients. DISCLOSURE: The following authors have nothing to disclose: Abdul Siddiqui, Faraz Siddiqui, Saqib Abbasi, Tahir Khan, Moiz Ahmed, Sana Alhajri, Shimshon Wiesel, Naureen Narula, Asma Memon, Michel Chalhoub No Product/Research Disclosure Information