The Role of Outcomes Research in Reducing Bleeding After Percutaneous Coronary Intervention

Complicating events that occur intra-procedurally during percutaneous coronary intervention (PCI) demand the full attention and skill of the interventional cardiologist. However, complications that occur after the patient leaves the lab may be managed by a variety of clinicians, including other physicians and nurses. Thus, post procedural events, particularly bleeding events, tend to be underestimated in frequency and type unless they are carefully tracked and reported back to clinicians. Database-driven outcomes research can give an accurate picture of post procedural events and their rates. While post-procedure ischemic complications are often viewed as urgent, bleeding complications, unless of hemorrhagic proportions, are often viewed as less serious. Possibly because of this long held belief, bleeding events are not routinely tracked and tabulated after PCI. Yet recent publications have shown that bleeding is, in fact, a very serious complication that can lead to further ischemic complications, as well as an alarming increase in mortality.¹ The relationship between blood loss and resulting ischemia to an already weakened coronary muscle is being closely explored, but what is certain is that patients who do not experience bleeding complications recover better, and faster. In this study, we evaluated post-PCI bleeding complications during a one-year period in patients who were treated with heparin plus glycoprotein (GP) IIb/IIIa inhibitors. Outcomes showed a higher than acceptable rate of bleeding events. Bivalirudin, a direct thrombin inhibitor, has consistently shown a superior safety profile over heparin, particularly in regard to bleeding events, and was therefore instituted as a substitution for heparin in a subset of patients to investigate whether reductions in bleeding could be realized. Event rates for heparin with GP IIb/IIIa inhibitors and bivalirudin monotherapy were tracked for a one-year period. Significant reductions in bleeding were found in the bivalirudin patients. As physicians became aware of the consistent efficacy and bleeding reductions associated with bivalirudin, utilization of the drug increased steadily as bleeding events continued to decrease. This article describes the system used at Christ Hospital in Oak Lawn, Illinois to monitor and reduce bleeding events in its post-PCI patient population. Methods This study consisted of two year-long study periods. The first was between August 2002 and July 2003, during which time bleeding outcomes of patients who underwent PCI with heparin plus GP IIb/IIIa inhibitors were tracked. The second was between August 2003 and July 2004, for which outcomes of patients treated with heparin plus GP IIb/IIIa blockade were collected and compared to those treated with bivalirudin and provisional use of GP IIb/IIIa inhibitors. To assure all bleeding was recorded in the database, each patient was visited the day after the procedure by a clinical outcome Registered Nurse (RN) to assess the patient’s status and monitor the puncture site. Post-procedure bleeding events collected included retroperitoneal bleeds with and without transfusion, hematomas >10cm, and hematomas with transfusion. American College of Cardiology (ACC) definitions were used in defining a bleeding event (Appendix 1). If a complication was found, the clinical outcome RN completed a data sheet and the event was entered into the Velos ACC database (version 3.04) by a data entry secretary. The data were extrapolated on a monthly basis to be discussed at the Performance Improvement Committee meeting. Members of the committee included the cardiovascular director, cardiologists, cardiovascular surgeons, as well as the cath lab, telemetry, and post-intervention unit managers. Using this methodology, it was determined that between August 2002 and July 2003, total bleeding rate for patients anticoagulated with heparin (n=1437) was 2.5%, and the transfusion rate was 2%. Major adverse cardiac events (MACE) in this same time period included 4 deaths (0.28%), and 11 myocardial infarctions (0.77%). Target vessel revascularization rates are not available. Prior to these results, heparin alone was used in approximately 10% of PCI cases and heparin with GP IIb/IIIa inhibitors in the remaining 90%. Understanding the detrimental effects of bleeding and transfusion, it was determined that efforts were required to decrease bleeding events. Physicians decided to test bivalirudin, a direct thrombin inhibitor shown to decrease bleeding, instead of heparin in a subset of PCI patients. Bivalirudin, a thrombin-specific anticoagulant, overcomes many of the limitations posed by heparin. Bivalirudin binds specifically to thrombin, resulting in a linear and predictable dose-proportional anticoagulant response throughout the procedure.² Bivalirudin inhibits both circulating and clot-bound thrombin, thereby providing more effective inhibition of thrombin and thrombin generation.² Further, bivalirudin has a shorter half-life than heparin (25 minutes), which may contribute to reduced bleeding risk.³ From August 2003 through July 2004, interventional cardiologists used bivalirudin monotherapy in all elective and urgent PCIs, adding GP IIb/IIIa inhibitors only on a provisional basis. The system described above for tracking and recording bleeding events was utilized. Over the course of the study year, as a widening reduction in bleeding was experienced in the bivalirudin patients, more physicians chose to treat a greater number of patients with bivalirudin instead of heparin (72% vs. 28%). Results A total of 1465 patients underwent angioplasty, with 1055 receiving bivalirudin, and 410 receiving heparin. Baseline and procedural characteristics are detailed in Table 1. In the majority of cases, bivalirudin was used alone, with GP IIb/IIIa inhibitors used in only 7.2% of patients, compared to 45% of those patients treated with heparin. A significant reduction in bleeding was seen in the bivalirudin-treated patients compared to patients who received heparin (Figure 1). Of all patients treated between August 2003 and July 2004, only one patient required a transfusion in the bivalirudin group, and hematomas >10 cm were reduced by 66% compared to heparin. Bleeding outcomes for 1474 interventions performed during this study period are detailed in Table 2. Total MACE events for 2003-2004 were determined retrospectively through database analysis and were not available to track by anticoagulant. Of a total of 1465 patients, there was one death (0.07%) and 20 MIs (1.4%). While the rate of MI almost doubled over that in 2002-2003, the number of deaths was reduced from 4 to 1. Discussion True rates of bleeding associated with PCI procedures often are unknown by physicians and hospital administrators. It takes a diligent effort to not only track and tabulate bleeding prospectively, but also to devise an effective communication system to report the actual rate of bleeding in their patients back to physicians and nurses caring for PCI patients. There is an increasing body of evidence that suggests that bleeding leads to poor outcomes, prolonged hospital stays, and increased mortality. Slater reported that hematoma with transfusion was associated with a 9-fold increase in hospital death and a 6-fold increase in hospital death/MI. In his study, the procedural mortality rate and death/MI rate among patients with hematoma requiring transfusion was significantly higher than those who did not experience this event (10.3% vs. 1.2%, p 4 Kinnaird et al found that bleeding increased risk of in-hospital cardiac events and mortality, and even increased the risk of mortality out to one year post procedure.¹ While it is well-known that anticoagulant therapy increases the risk of bleeding, some drugs are associated with increased bleeding, including GP IIb/IIIa inhibitors and low molecular weight heparins.¹ Kuchulakanti, in a retrospective study of 10,669 PCI patients, found that despite careful use of antiplatelet and antithrombotics, vascular complications continue to be a problem and are predictors of both adverse coronary events and one-year mortality.⁵ Clearly efforts to reduce bleeding lead to improved outcomes for patients. Bivalirudin is the only anticoagulant that is associated with reduced bleeding events. In the REPLACE-2 trial,⁶ overall bleeding was reduced 42% (2.4% vs. 4.1%, p=0.001) and access site bleeding was reduced 68% (0.8% vs. 2.5%, p=0.001) with bivalirudin compared to heparin plus GP IIb/IIIa inhibitors, while ischemic protection was equal to heparin plus GP IIb/IIIa. Our data confirm the REPLACE-2 findings. Bleeding events were significantly reduced in the bivalirudin patients, and use of GP IIb/IIIa was reduced from 45% to 7.2%. There was no discernable difference in ischemic complications, although they were not tracked as part of this study. Presently, bivalirudin with provisional GP IIb/IIIa is the drug of choice in among physicians in our institution due to their experience with this study. Any effort to reduce bleeding must start with an objective measure of actual bleeding events that are occurring. A second critical component to a successful program is a regular feedback mechanism to inform clinicians of actual bleeding event rates. A monthly meeting with interventional cardiologists and the clinical outcomes RN to discuss outcomes data is an effective way to share the actual rates of complication experienced by angioplasty patients, and allows for a discussion in a group setting of efforts that can be made to reduce complications. Conclusion Concentrated efforts at tabulating and evaluating bleeding events prospectively in angioplasty patients is the first step in understanding true rates of bleeding complications. While these efforts may be labor-intensive, they are the only way to truly understand the consequences of anticoagulant use. Bivalirudin has shown to provide ischemic protection for patients undergoing PCI, while significantly reducing bleeding complications in patients. Bleeding will always be present due to the need to anticoagulate patients during angioplasty procedures; however, reducing bleeding to the lowest rate possible means providing the best of care for PCI patients. Patricia Busch can be contacted at patricia.busch@advocatehealth.com She discloses that she is a member of the speaker’s bureau for The Medicines Company. The remaining authors have nothing to disclose. Appendix Bleeding and vascular complications were defined according to the ACC National Cardiovascular Data Registry, Cardiac Catheterization Module 3.04 Data Definitions, outlined in the table below.⁷ Bleeding “ Percutaneous Entry Site The bleeding requires a transfusion and/or prolong the hospital stay, and/or cause a drop in hemoglobin >3.0 gm/dl. Bleeding at the percutaneous entry site can be external or a hematoma >10 cm for femoral access or >2 cm for radial access; or >5 cm for brachial access. Bleeding - Retroperitoneal The retroperitoneal bleeding occurred during or after the cath lab visit until discharge. The bleeding requires a transfusion and/or prolongs the hospital stay, and/or cause a drop in hemoglobin > 3.0 gm/dl. Bleeding -Gastrointestinal Gastrointestinal bleeding occurred during or after the cath lab visit until discharge. The bleeding requires a transfusion and/or prolong the hospital stay, and/or cause a drop in hemoglobin > 3.0 gm/dl. Bleeding - Genital/Urinary Genital or urinary bleeding occurred during or after the cath lab visit until discharge. The bleeding should require a transfusion and/or prolong the hospital stay, and/or cause a drop in hemoglobin >3.0gm/dl. Bleeding - Other/Unknown Bleeding occurred at other or unknown locations during or after the cath lab visit until discharge. The bleeding should require a transfusion and/or prolong the hospital stay, and/or cause a drop in hemoglobin >3.0 gm/dl. Vascular - Access Site Occlusion An access site occlusion occurred at the site of percutaneous entry during the procedure or after lab visit but before any subsequent lab visits. Access Site Occlusion is defined as total obstruction of the artery usually by thrombus (but may have other causes) usually at the site of access requiring surgical repair. Occlusions may be accompanied by absence of palpable pulse or Doppler. Vascular - Peripheral Embolization A peripheral embolization occurred distal to the arterial access site during the procedure or after lab visit but before any subsequent lab visits, requiring therapy. Peripheral embolization is defined as a loss of distal pulse, pain and/or discoloration (especially the toes). This can include cholesterol emboli. Vascular - Dissection A dissection occurred at the site of percutaneous entry during the procedure or after lab visit but before any subsequent lab visits. A dissection is defined as a disruption of an arterial wall resulting in splitting and separation of the intimal (subintimal) layers. Vascular - Pseudoaneurysm A pseudoaneurysm occurred at the site of percutaneous entry during the procedure or after lab visit but before any subsequent lab visits. Do not code for pseudoaneurysms noted after discharge. Pseudoaneurysm is defined as the occurrence of a disruption and dilation of the arterial wall without identification of the arterial wall layers at the site of the catheter entry demonstrated by arteriography or ultrasound.

1. Kinnaird TD, Stabile E, Mintz GS, et al. Incidence, predictors, and prognostic implications of bleeding and blood transfusion following percutaneous coronary interventions. <i>Am J Cardiol</i> 2003;92:930-935.<p>2. Topol EJ, Bonan R, Jewitt D, et al. Use of a direct antithrombin, Hirulog, in place of heparin during coronary angioplasty. <i>Circulation</i> 1993;87(5):1622-1629. </p><p>3. Weitz JI and Bates SM. Acute coronary syndromes: a focus on thrombin. <i>J Invas Cardiol</i> 2002;14(Suppl B):2B-7B. </p><p>4. Slater J, Selzer F, Feit F, et al. The impact of adverse access site hematoma with transfusion in patients undergoing percutaneous coronary intervention: A report from the NHLBI Dynamic Registry. [Abstract] <i>Circulation</i> 2003;356. Abstract 1667. </p><p>5. Kuchulakanti PK, Satler LF, Suddath WO, et al. Vascular complications following coronary intervention correlate with long-term cardiac events. <i>Catheter Cardiovasc Interv</i> 2004;62:181-185. </p><p>6. Lincoff AM, Bittl JA, Harrington RA, et al. Bivalirudin and provisional glycoprotein IIb/IIIa blockade compared with heparin and planned glycoprotein IIb/IIIa blockade during percutaneous coronary intervention. REPLACE-2 Randomized Trial. <i>JAMA</i> 2003;289(7):853-863. </p><p>7. ACC National Cardiovascular Data Registry Cardiac Catheterization Module 3.04 Data Definitions. Accessed on May 4, 2005 at http://www.accncdr.com/WebNCDR/NCDRDocuments/datadictdefsonlyv30.</p&gt;