Gastrointestinal Bleeding in Patients With Mechanical Ventilation: Insights From the Healthcare Utilization Project

Abstract

Introduction: Stress ulceration of the gastrointestinal tract causes overt and clinically important bleeding in critically ill patients. It is suggested that Mechanical Ventilation (MV) is one of the most significant risk factors of Gastrointestinal Bleeding (GIB) in the intensive care unit. Very few studies have reported the incidence of GIB in MV patients. We aimed at analyzing data about this complication in MV patients in united states. Methods: We analyzed the National Inpatient Sample (NIS) database years 2002-2014 to identify hospitalizations in which MV was performed (ICD-9 CM code 96.70, 96.71 & 96.72). Hospitalizations complicated by GIB in this cohort were then identified. All primary admissions of GIB were excluded to minimize the bias. Prevalence of comorbidities was determined using Agency for Healthcare Research and Quality (AHRQ) comorbidity measures. Statistical significance of variation in the number of hospital discharges, demographics, cost of care, comorbidity measures and outcomes during the study period were determined using Cochran-Armitage trend test. Results: In 2002 there were 705806 hospitalizations with MV compared to 977560 in 2014. (P=<0.001). The percentage of MV hospitalizations complicated by GIB slightly increased from 7.43% in 2002 to 7.97% in 2014. The In-Hospital Mortality showed slightly improving trend from 46.7% in 2002 to 40.9% in 2014, but still the difference between the MV patients without GIB remains significant. The most commonly associated comorbidities in this cohort was alcoholism (14.1% Vs 8.8%, P=<0.001), congestive heart failure (24.6% Vs 22.6%, P=<0.001) coagulopathies (26.5% Vs 12.5%, P=<0.001), chronic liver disease (11.1% Vs 4.1%, P=<0.001), electrolytes disturbances (63.6% Vs 49.9%, P=<0.001) and chronic renal failure (21.3% vs 16.6%). There was a significant difference in median length of stay (8.22 Vs 11.89 days, P=<0.001) and median cost of care (73914 Vs 106740, P=<0.001) between MV hospitalizations without GIB than with GIB. Conclusion: In this analysis of inpatient sample of patients with MV, slight increase in incidence of GIB with associated high mortality is concerning, as there is more liberal use of proton pump inhibitors in these patients to prevent GIB. Further studies are needed to identify predictors of these trends and factors responsible for such outcomes to better elucidate and perhaps support our findings.