The Impact of Gastrointestinal Bleeding on Six-Month Cardiovascular Outcomes in Survivors of a ST-Segment Elevation Acute Myocardial Infarction

Introduction: The occurrence of gastrointestinal bleeding (GIB) during an acute myocardial infarction (AMI) might affect the management of AMI which could impact future cardiovascular outcomes. We evaluated the impact of GIB on the occurrence of subsequent 6-month AMI and heart failure (HF) events among patients who survived an ST-segment elevation AMI (STEMI). Methods: Adult survivors of a STEMI between January and June 2014 were identified from the US Nationwide Readmissions Database. Outcomes evaluated were 1) hospitalization for an AMI and 2) HF hospitalization, both within 6-months of discharge. Patients with either upper, lower, or unspecified GIB were included in the GIB cohort. Appropriate ICD-9 codes were utilized to identify all diagnoses. Unadjusted and risk-adjusted evaluations of the impact of GIB on the outcomes were made using the chi-square and multivariable logistic regression analyses respectively. Results of the adjusted analysis are presented as odds ratio (OR) and 95% confidence intervals (CI). The regression model adjusted for age, sex, cardiovascular risk factors, prior cardiovascular diseases, current and prior coronary revascularization, heart failure, kidney disease, income, and insurance status. Results: Of 66,723 STEMI survivors, a GIB occurred in 878 (1.3%) patients. Patients who had a GIB were older (mean age 70+13 years vs 62+12 years), more likely to be female (43% vs 30%), and had a lower rate of coronary revascularization (66% vs 86%) compared with patients who did not have a GIB (all P< 0.001). A 6-month AMI event occurred in 3,232 patients (4.8%), at a 130% higher rate in those with a GIB (11.0%) compared with patients without a GIB (4.8%; P< 0.001). The higher AMI risk in patients with a GIB persisted after multivariable risk adjustment (OR 1.72, 95% CI 1.21-2.44; P< 0.001). An HF hospitalization within 6-months from discharge occurred in 2,273 patients (3.4%), at a 158% higher rate in those with a GIB (8.5%) compared with patients without a GIB (3.3%; P< 0.001). The higher HF risk in patients with a GIB did not persist after multivariable risk adjustment (OR 1.17, 95% CI 0.77-1.78; P=0.453). Conclusion: GIB occurred in a small proportion of STEMI survivors but substantially increased the risk for 6-month AMI events even after risk adjustment. The increased risk for HF hospitalizations with GIB was not evident after risk adjustment. Optimal AMI management strategies in patients who also have a GIB are needed to improve cardiac outcomes.