Assessment of safety of performing percutaneous coronary intervention after a recent episode of gastrointestinal bleeding

BACKGROUND: Little literature exists on the risk of performing coronary intervention (PCI) on patients who have had recent gastrointestinal bleeding (GIB), although bleeding after PCI has been identified as a risk factor for long-term mortality. METHODS: Patients within the Cleveland Clinic PCI database who had acute GIB within 30 days preceding PCI during the same hospitalization (n = 79) were retrospectively compared to those who had PCI without recent GIB (n = 10 979) for mortality and need for revascularization. Baseline characteristics, laboratory values, procedures, morbidities, and mortality were compared using chi-square test for categorical variables and using Wilcoxon rank sum test for continuous variables. Mortality data was obtained using Social Security Death Index and demonstrated using Kaplan-Meier method. RESULTS: The GIB group had more prevalent history of peptic ulcer disease, GIB, gastrointestinal or liver disease (P < 0.0001), transient ischemic accident (P = 0.017), peripheral vascular disease (P = 0.0002), significant carotid artery occlusion (P = 0.023), and myocardial infarction (P < 0.0001). 47% of patients had upper GIB with 20% needing endoscopic intervention. This group had more anemia (P < 0.0001), heart failure (P = 0.0001), cardiogenic shock (10% versus 1.4%, P < 0.001), cardiac arrest (7.6% versus 1%, P < 0.001). GIB group had worse in-hospital mortality (P < 0.0001), long-term mortality (P < 0.001), and a 7.6% re-bleeding incidence. CONCLUSIONS: Overall, the patients who had GIB preceding PCI had higher in-hospital mortality and long-term mortality compared with those without GIB before PCI.