Original Contribution

The American Journal of Gastroenterology (2007) 102, 2411–2416; doi:10.1111/j.1572-0241.2007.01460.x

Predictive and Protective Factors Associated With Upper Gastrointestinal Bleeding After Percutaneous Coronary Intervention: A Case-Control Study

Marcus WS Chin MBBS (Hons)1, Gerald Yong MBBS, FRACP2, Max K Bulsara BSc (Hons), MSc3, Jamie Rankin MBBS, FRACP4 and Geoffrey M Forbes MBBS, MD, FRACP5

  1. 1Department of Gastroenterology and Hepatology, Royal Perth Hospital, Perth, Western Australia
  2. 2Department of Cardiology, Royal Perth Hospital, Perth, Western Australia
  3. 3School of Population Health, Faculty of Medicine and Dentistry, University of Western Australia, Perth, Western Australia
  4. 4Department of Cardiology, Royal Perth Hospital, Perth, Western Australia
  5. 5Department of Gastroenterology and Hepatology, Royal Perth Hospital, and School of Medicine and Pharmacology, University of Western Australia, Perth, Western Australia

Correspondence: Dr. Marcus W. S. Chin , Department of Gastroenterology and Hepatology, Royal Perth Hospital, Box X2213 GPO Perth WA 6000, Western Australia.

Received 19 January 2007; Accepted 13 June 2007.





Hemorrhagic complications of acute coronary syndromes and percutaneous coronary intervention (PCI) are associated with increased mortality. Upper gastrointestinal (UGI) bleeding after PCI is a potential target for preventative strategies.



To evaluate the risk factors for UGI bleeding in a large cohort of contemporary PCI patients and assess the outcomes of medical and endoscopic management.



A case-control study evaluating UGI bleeding in the 30 days following PCI for stable angina and acute coronary syndromes, at one institution between 1998 and 2005. Cases were identified and outcomes assessed using linkage analysis of data from institutional PCI and endoscopy databases, statewide vital statistics and hospital discharge registries, and a detailed review of medical notes for each case and three matched controls. Analysis of the case and control groups for risk and protective factors was performed using the chi2 test with Fisher's exact P value and logistic regression.



The incidence of UGI bleeding following PCI was 1.2% (70 of 5,673 patients). The etiologies of these bleeds were diverse. Risk factors for UGI bleeding were primary PCI (OR 27.80, 95% CI 6.28–123.05, P < 0.001), cardiac arrest (OR 6.17, 95% CI 1.82–20.84, P = 0.003), inotropic requirement (OR 5.85, 95% CI 1.98–17.27, P = 0.001), thienopyridine use before PCI (OR 2.40, 95% CI 1.04–5.53, P = 0.02), and advanced age (OR 1.08, 95% CI 1.04–1.12, P < 0.001). Proton pump inhibitor use after PCI (OR 0.08, 95% CI 0.02–0.40, P = 0.002) was accompanied by a reduced risk of UGI bleeding. Endoscopy provided therapeutic intervention in 33% of patients. There were no serious complications of endoscopy. The 30-day mortality for cases was 11.9% and 0.5% for controls (P = 0.001).



UGI bleeding after PCI is relatively common and associated with increased mortality. Those undergoing PCI for acute myocardial infarction or in the presence hemodynamic instability are at highest risk. Proton pump inhibition following PCI may reduce the bleeding risk, though when UGI bleeding occurs, therapeutic endoscopy is safe.