Akram Alnounou, Brian Nohomovich, Mohammad Alabbas, Austin Brubaker, Dimitar Bahariev, John Brown, Amy Welch, Tooba Tariq, Neil Hughes, Santhosh K G Koshy, Eric Martin Sieloff, Neil Sengupta
GIB during PCI hospitalizations was predominantly UGIB, and the rise in incidence was driven by UGIB. UGIB was associated with higher adjusted in-hospital mortality and greater resource utilization than LGIB, suggesting UGIB and LGIB carry different associated mortality risk in this population, though residual confounding cannot be excluded. These findings support attention to upper-GI risk factors and gastroprotection in PCI patients.
BACKGROUND: Gastrointestinal bleeding (GIB) during a percutaneous coronary intervention (PCI) hospitalization can interrupt antithrombotic therapy and worsen outcomes. National data comparing upper GIB (UGIB) versus lower GIB (LGIB) in PCI hospitalizations are limited.
AIMS: To compare incidence, clinical characteristics, and in-hospital outcomes of UGIB versus LGIB events in PCI hospitalizations.
METHODS: We analyzed PCI hospitalizations in the 2016-2022 National Inpatient Sample. UGIB and LGIB were identified from site-specific ICD-10-CM diagnosis codes. Hospitalizations with both UGIB and LGIB codes, or with unspecified gastrointestinal hemorrhage (K92.2) without a site-specific code, were excluded. Survey-weighted regression adjusted for demographics, comorbidities, acuity markers, hospital and PCI procedural characteristics.
RESULTS: Among 3,227,805 weighted PCI hospitalizations, 41,225 (1.28%) had GIB: 34,215 UGIB-only (83.0%) and 7010 LGIB-only (17.0%). GIB incidence rose from 1.10% to 1.36% (2016-2022), driven by UGIB (0.90% to 1.13%; P < 0.001); LGIB was stable (0.20-0.26%; P = 0.1022). Compared with LGIB, UGIB was independently associated with higher in-hospital mortality (12.35% vs. 6.14%; adjusted odds ratio 1.74, 95% confidence interval 1.36-2.23), longer median stay (7.27 vs. 5.96 days), and higher median cost ($44,712 vs. $36,046; both P < 0.001), with no significant difference in transfusion (P = 0.2134). Excluding variceal-source UGIB did not alter this association.
CONCLUSIONS: GIB during PCI hospitalizations was predominantly UGIB, and the rise in incidence was driven by UGIB. UGIB was associated with higher adjusted in-hospital mortality and greater resource utilization than LGIB, suggesting UGIB and LGIB carry different associated mortality risk in this population, though residual confounding cannot be excluded. These findings support attention to upper-GI risk factors and gastroprotection in PCI patients.