568517
research-article2015
AOPXXX10.1177/1060028014568517Annals of PharmacotherapyBaek et al
Research Report
Evaluation of Procedure-Related Bleeding Risk in Patients Receiving Bivalirudin During Percutaneous Coronary Intervention
Annals of Pharmacotherapy 2015, Vol. 49(4) 427–430 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028014568517 aop.sagepub.com
Vivian S. Baek, PharmD1, Jeff Hurren, PharmD1, and Stephanie B. Edwin, PharmD1
Abstract Background:Bivalirudin has historically been considered an attractive anticoagulant during percutaneous coronary intervention (PCI) because of reduced bleeding complications reported by early trials. Bivalirudin use during PCIs has been a subject of controversy because of conflicting data and recent findings. Objective: To evaluate the clinical characteristics of patients receiving bivalirudin to determine if an opportunity to improve use exists based on risk of procedure-related bleeding. Methods: This was a single-center, retrospective, observational study (n = 100) of all adult patients who received bivalirudin during cardiac catheterization at St John Hospital and Medical Center from June to August 2013. The risk of bleeding complications associated with PCI was estimated using a clinical risk algorithm developed from the National Cardiovascular Data Registry (NCDR). Results: Treatment with bivalirudin was safe and effective. Of the 100 patients who received bivalirudin, only 34% were identified as having a high risk of procedure-related bleeding according to the NCDR clinical risk algorithm. There was no incidence of stent thrombosis noted and only 1 case of provisional glycoprotein IIb/IIIa inhibitor use. No episodes of Thrombolysis in Myocardial Infarction (TIMI) major bleeding were noted in the study population; however, 1 patient met TIMI minor bleeding criteria. Limitations of this study include small sample size and retrospective nature of the study. Conclusion: Opportunities to establish a more cost-effective use of bivalirudin may exist through implementation of protocols incorporating the NCDR risk assessment model. Keywords bivalirudin, percutaneous coronary intervention, bleeding
Introduction Approximately 1 million percutaneous coronary intervention (PCI) procedures are performed annually in the United States, which have been associated with a mean hospitalization cost of $70 176 per patient.1 Bleeding complications resulting from PCI can lead to prolonged hospitalization, morbidity, mortality, and increased health care costs.2 Recognizing these challenges as an opportunity for improvement, the American College of Cardiology (ACC) routinely evaluates performance and outcome measures related to PCI in-hospital risk-adjusted rate of bleeding through the National Cardiovascular Data Registry (NCDR) CathPCI registry.3 Bivalirudin, a synthetic reversible direct thrombin inhibitor, has become an attractive anticoagulant over unfractionated heparin and glycoprotein IIb/IIIa inhibitors (GPIs) because of reduced bleeding complications, as demonstrated by early trials.4-6 The high cost of bivalirudin, however, may be a prohibitive factor for widespread use. Furthermore, the presumed benefits of bivalirudin are under
increasing scrutiny because of conflicting data and recent findings. A controversial, single-center, randomized study recently demonstrated a lower rate of major adverse cardiac events in patients treated with unfractionated heparin, without any difference in bleeding complications, versus bivalirudin.7 Similarly, data from the Ottawa STEMI registry revealed no statistical difference in bleeding or the composite end point of in-hospital major bleeding, stroke, reinfarction, or death for patients receiving bivalirudin versus unfractionated heparin alone.8 To assist with identification of patients at high risk of periprocedural bleeding, a clinical risk algorithm was developed using the NCDR CathPCI registry in 2009.2 This model uses 9 preprocedural clinical factors to predict risk of 1
St John Hospital and Medical Center, Detroit, MI, USA
Corresponding Author: Stephanie B. Edwin, Department of Pharmacy, St John Hospital and Medical Center, 22101 Moross Rd, Detroit, MI 48236, USA. Email: [email protected]
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
428
Annals of Pharmacotherapy 49(4)
Variable ACS Type ST-elevation MI Non-ST elevation MI / Unstable angina Cardiogenic shock Female gender Previous CHF No previous PCI NYHA class IV CHF Peripheral vascular disease Age (years) 66-75 76-84 > 85
Points Assigned 10 3 8 6 5 4 4 2
Estimated glomerular filtration rate
2 5 8 1 (per 10 unit decrease if < 90 mL/min)
Score
Risk Category
Estimated Bleeding Risk
Less than 8 points
Low Risk
0.63%
8-17 points
Intermediate Risk
1.77%
Greater than 17 points
High Risk
5.08%
NYHA: New York Heart Association; CHF: Congestive Heart Failure Adapted from Mehta SK. Circ Cardiovac Intervent. 2009.
Figure 1. National Cardiovascular Data Registry bleeding clinical risk algorithm.
bleeding events following PCI (Figure 1). The purpose of this study is to evaluate clinical characteristics of patients receiving bivalirudin, based on risk of bleeding, to determine if an opportunity to improve use exists.
Methods This retrospective, observational study was performed at St John Hospital and Medical Center, a 772-bed hospital in Detroit, Michigan. Institutional review board approval was granted prior to the start of data collection; informed consent was waived because of the retrospective nature of the study. A convenience sample of patients who received bivalirudin was generated from the pharmacy information system over the period from June 2013 to August 2013. Adult patients who received PCI were eligible for inclusion. Only the first PCI procedure was included in the analysis for patients with several PCI procedures during the study period. Patients with missing bleeding or efficacy data were excluded from the study. Baseline demographics, including anticoagulant and antiplatelet medications prior to PCI, were collected for each patient. The risk of bleeding complications associated with PCI was estimated using a clinical risk algorithm developed from the NCDR. This model comprises 9 baseline clinical factors that are used to categorize risk of experiencing a bleeding event following PCI (Figure 1).2 Data on vascular access site and adjunctive anticoagulant/antiplatelet medications were collected. Creatinine clearance (CrCl) was calculated for each patient to determine the appropriateness of the bivalirudin dosing strategy, according to manufacturer recommendations, using the Cockcroft-Gault formula and serum creatinine closest to the time of PCI.
Data on the incidence of stent thrombosis and use of provisional GPIs during the index hospitalization were collected to assess the efficacy of bivalirudin. Bleeding events were assessed using the Thrombolysis in Myocardial Infarction (TIMI) bleeding criteria, and site of bleeding was recorded when available. Major bleeding events were defined as intracranial hemorrhage, clinically overt signs of hemorrhage with hemoglobin drop ≥5 g/dL, or fatal bleeding. Minor bleeding events were defined as clinically overt bleeding resulting in hemoglobin drop ≥3 to 5 g/dL.9 Data are reported using descriptive statistics. Continuous variables are described as means with SDs. Categorical variables are described as frequency distributions. All data were analyzed using Microsoft Excel 2007.
Results During the study period, 100 consecutive patients were identified and included for analysis. The majority of patients receiving bivalirudin were Caucasian (69%) and male (58%) and had an average age of 65.1 ± 13.1 years (Table 1). Three patients had thrombocytopenia (platelets < 100 000/ mm3) prior to cardiac catheterization; however, no patient had a documented history of heparin-induced thrombocytopenia. The majority of patients had a CrCl >30 mL/min; bivalirudin dose adjustment for renal dysfunction was required in only 13% of patients. Despite the higher bleeding risk in STEMI patients, the majority of patients had a diagnosis of unstable angina (73%; Table 2). Vascular access was obtained through the radial artery in 46% of patients. Unfractionated heparin was commonly administered (92%) to patients prior to PCI. Antiplatelet agents were administered immediately prior to or during PCI in
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
429
Baek et al Table 1. Baseline Characteristics (n = 100).a
40
Age (years) Female Caucasian History of HIT Platelets < 100 000/mm3 Estimated CrCl (mL/min) CrCl ≥ 30 mL/min CrCl = 10-29 ESRD/Dialysis Anticoagulation prior to admission Warfarin LMWH Dabigatran
65.1 ± 13.1 42 69 0 3 87 8 5 3 1 1
Abbreviations: HIT, heparin-induced thrombocytopenia; CrCl, creatinine clearance; ESRD, end-stage renal disease; LMWH, low-molecular-weight heparin. a Continuous data are displayed as mean ± SD. Categorical data are displayed as numbers.
Table 2. Procedural Data (n = 100). Procedure Characteristics
n (%)
ACS type Unstable angina STEMI NSTEMI Vascular access site Femoral Radial Adjunctive antiplatelet (n = 74) Clopidogrel 600 mg Clopidogrel 300 mg Prasugrel 60 mg Ticagrelor 180 mg Other Heparin (n = 92) Prior to catheterization During catheterization
73 (73%) 19 (19%) 8 (8%) 54 (54%) 46 (46%) 13 (18%) 12 (16%) 4 (5%) 36 (49%) 9 (12%) 41 (45%) 51 (55%)
Abbreviation: ACS, Acute Coronary Syndrome.
74% of patients; ticagrelor (49%) was the most frequently used agent in this setting. Only 34% of the patients who received bivalirudin were classified as high risk for PCI-related bleeding according to the NCDR clinical risk algorithm (Figure 2). The risk of procedure-related bleeding was calculated as low in 23% of patients. Of note, the NCDR risk score was unable to be calculated retrospectively for 11 patients because of a lack of complete NCDR variable documentation in the electronic medical record. Treatment with bivalirudin was determined to be relatively safe and effective. No incidents of acute stent
Number of Patients (%)
Characteristics
32
34
30 23 20 11 10 Low
Intermediate
High
NCDR Risk Category
Unable to Calculate
Figure 2. National Cardiovascular Data Registry (NCDR) risk stratification (n = 100).
thrombosis were identified in the study population. Only 1 case of provisional GPI use was documented. Three patients did not receive appropriate doses of bivalirudin because of discrepancies in weight and/or renal function, none of which resulted in a procedural complication. No episodes of TIMI major bleeding were noted in the study population; 1 patient met criteria for TIMI minor bleeding. Per the NCDR bleeding risk assessment, this patient was high risk and experienced bleeding at a femoral vascular access site.
Discussion This retrospective, observational study of patients undergoing PCI demonstrates that bivalirudin is a safe and effective antithrombotic agent. The rate of bailout GPI and incidence of bleeding events was noted to be lower than reported in previous studies.6-8 It is interesting to note that bivalirudin was frequently utilized in patients with a low or moderate risk of procedure-related bleeding in our study population. Because bivalirudin is a significantly more expensive antithrombotic compared with unfractionated heparin, an opportunity may exist to establish a more cost-effective use strategy. Bivalirudin has historically been considered an attractive anticoagulation option during PCI based on the improved safety profile and similar efficacy demonstrated by early landmark trials.9,10 This has recently been challenged by the results of the recently published HEATPPCI. This open-label, single-center, randomized controlled trial is unique because previous studies used outcomes that combined safety and efficacy. The primary efficacy outcome of all-cause mortality, cerebrovascular accident, reinfarction, or unplanned target lesion revascularization was significantly lower in the cohort of patients receiving unfractionated heparin compared with bivalirudin. With similar frequency of GPI use in each group, no significant difference in major bleeding events was noted between agents.10 Changes in clinical practice since the publication of the landmark bivalirudin trials should be considered. Vascular
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
430
Annals of Pharmacotherapy 49(4)
access through the radial artery is an important advance, used with increasing frequency, to reduce the rate of procedurerelated complications. In patients with radial access, the lower rate of bleeding may obviate the need for an antithrombotic agent, which reduces bleeding complications. Radial access was used in almost half of the patients who underwent PCI in our study. Further research regarding the role of bivalirudin in patients with radial access is needed. One small, retrospective study evaluating patients who received bivalirudin by radial or femoral access found no difference in net clinical adverse events and ischemic or bleeding complications between groups.10 Another important change in clinical practice since the publication of landmark bivalirudin trials is the use of novel P2Y12 inhibitors. A recent trial that used novel P2Y12 inhibitors in approximately half of all patients noted a significantly higher incidence of acute stent thrombosis in patients receiving bivalirudin. These results suggest that acute stent thrombosis remains a concern with bivalirudin, regardless of antiplatelet selection.6 Limitations of this study must be acknowledged. Given that this was a retrospective, observational study, the results of the study depended on the accuracy of the chart. Lack of documentation may have affected the study results, such as bleeding complications. TIMI bleeding criteria do not encompass all types of bleeding; hence, bleeding events may have occurred without being captured by our definition. The study also had a nonrandom convenience sample, which can increase the potential for selection bias and limit external validity. The small sample size may lead to larger uncertainties in measurement of bleeding and safety outcomes. We acknowledge that this study cannot ascertain how well bivalirudin is performing at our institution as compared with patients receiving heparin because our study was not controlled. Finally, it should be noted that this study was designed based on the 2009 NCDR bleeding risk assessment model. A newer bleeding risk assessment model was released in 2013, which incorporates 10 clinical factors. Because each bleeding risk assessment model was validated in a large cohort of patients, we would expect to see similar trends in bivalirudin use using the new bleeding risk assessment. Based on the results of this study, our institution has implemented criteria for bivalirudin use. Prior to implementation of these criteria, selection of the anticoagulation strategy was at the discretion of the interventional cardiologist. According to the newly implemented criteria, patients may receive anticoagulation with bivalirudin during PCI if any of the following criteria are met: high risk per NCDR bleeding risk assessment, documented history of heparininduced thrombocytopenia, platelet count
research-article2015
AOPXXX10.1177/1060028014568517Annals of PharmacotherapyBaek et al
Research Report
Evaluation of Procedure-Related Bleeding Risk in Patients Receiving Bivalirudin During Percutaneous Coronary Intervention
Annals of Pharmacotherapy 2015, Vol. 49(4) 427–430 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028014568517 aop.sagepub.com
Vivian S. Baek, PharmD1, Jeff Hurren, PharmD1, and Stephanie B. Edwin, PharmD1
Abstract Background:Bivalirudin has historically been considered an attractive anticoagulant during percutaneous coronary intervention (PCI) because of reduced bleeding complications reported by early trials. Bivalirudin use during PCIs has been a subject of controversy because of conflicting data and recent findings. Objective: To evaluate the clinical characteristics of patients receiving bivalirudin to determine if an opportunity to improve use exists based on risk of procedure-related bleeding. Methods: This was a single-center, retrospective, observational study (n = 100) of all adult patients who received bivalirudin during cardiac catheterization at St John Hospital and Medical Center from June to August 2013. The risk of bleeding complications associated with PCI was estimated using a clinical risk algorithm developed from the National Cardiovascular Data Registry (NCDR). Results: Treatment with bivalirudin was safe and effective. Of the 100 patients who received bivalirudin, only 34% were identified as having a high risk of procedure-related bleeding according to the NCDR clinical risk algorithm. There was no incidence of stent thrombosis noted and only 1 case of provisional glycoprotein IIb/IIIa inhibitor use. No episodes of Thrombolysis in Myocardial Infarction (TIMI) major bleeding were noted in the study population; however, 1 patient met TIMI minor bleeding criteria. Limitations of this study include small sample size and retrospective nature of the study. Conclusion: Opportunities to establish a more cost-effective use of bivalirudin may exist through implementation of protocols incorporating the NCDR risk assessment model. Keywords bivalirudin, percutaneous coronary intervention, bleeding
Introduction Approximately 1 million percutaneous coronary intervention (PCI) procedures are performed annually in the United States, which have been associated with a mean hospitalization cost of $70 176 per patient.1 Bleeding complications resulting from PCI can lead to prolonged hospitalization, morbidity, mortality, and increased health care costs.2 Recognizing these challenges as an opportunity for improvement, the American College of Cardiology (ACC) routinely evaluates performance and outcome measures related to PCI in-hospital risk-adjusted rate of bleeding through the National Cardiovascular Data Registry (NCDR) CathPCI registry.3 Bivalirudin, a synthetic reversible direct thrombin inhibitor, has become an attractive anticoagulant over unfractionated heparin and glycoprotein IIb/IIIa inhibitors (GPIs) because of reduced bleeding complications, as demonstrated by early trials.4-6 The high cost of bivalirudin, however, may be a prohibitive factor for widespread use. Furthermore, the presumed benefits of bivalirudin are under
increasing scrutiny because of conflicting data and recent findings. A controversial, single-center, randomized study recently demonstrated a lower rate of major adverse cardiac events in patients treated with unfractionated heparin, without any difference in bleeding complications, versus bivalirudin.7 Similarly, data from the Ottawa STEMI registry revealed no statistical difference in bleeding or the composite end point of in-hospital major bleeding, stroke, reinfarction, or death for patients receiving bivalirudin versus unfractionated heparin alone.8 To assist with identification of patients at high risk of periprocedural bleeding, a clinical risk algorithm was developed using the NCDR CathPCI registry in 2009.2 This model uses 9 preprocedural clinical factors to predict risk of 1
St John Hospital and Medical Center, Detroit, MI, USA
Corresponding Author: Stephanie B. Edwin, Department of Pharmacy, St John Hospital and Medical Center, 22101 Moross Rd, Detroit, MI 48236, USA. Email: [email protected]
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
428
Annals of Pharmacotherapy 49(4)
Variable ACS Type ST-elevation MI Non-ST elevation MI / Unstable angina Cardiogenic shock Female gender Previous CHF No previous PCI NYHA class IV CHF Peripheral vascular disease Age (years) 66-75 76-84 > 85
Points Assigned 10 3 8 6 5 4 4 2
Estimated glomerular filtration rate
2 5 8 1 (per 10 unit decrease if < 90 mL/min)
Score
Risk Category
Estimated Bleeding Risk
Less than 8 points
Low Risk
0.63%
8-17 points
Intermediate Risk
1.77%
Greater than 17 points
High Risk
5.08%
NYHA: New York Heart Association; CHF: Congestive Heart Failure Adapted from Mehta SK. Circ Cardiovac Intervent. 2009.
Figure 1. National Cardiovascular Data Registry bleeding clinical risk algorithm.
bleeding events following PCI (Figure 1). The purpose of this study is to evaluate clinical characteristics of patients receiving bivalirudin, based on risk of bleeding, to determine if an opportunity to improve use exists.
Methods This retrospective, observational study was performed at St John Hospital and Medical Center, a 772-bed hospital in Detroit, Michigan. Institutional review board approval was granted prior to the start of data collection; informed consent was waived because of the retrospective nature of the study. A convenience sample of patients who received bivalirudin was generated from the pharmacy information system over the period from June 2013 to August 2013. Adult patients who received PCI were eligible for inclusion. Only the first PCI procedure was included in the analysis for patients with several PCI procedures during the study period. Patients with missing bleeding or efficacy data were excluded from the study. Baseline demographics, including anticoagulant and antiplatelet medications prior to PCI, were collected for each patient. The risk of bleeding complications associated with PCI was estimated using a clinical risk algorithm developed from the NCDR. This model comprises 9 baseline clinical factors that are used to categorize risk of experiencing a bleeding event following PCI (Figure 1).2 Data on vascular access site and adjunctive anticoagulant/antiplatelet medications were collected. Creatinine clearance (CrCl) was calculated for each patient to determine the appropriateness of the bivalirudin dosing strategy, according to manufacturer recommendations, using the Cockcroft-Gault formula and serum creatinine closest to the time of PCI.
Data on the incidence of stent thrombosis and use of provisional GPIs during the index hospitalization were collected to assess the efficacy of bivalirudin. Bleeding events were assessed using the Thrombolysis in Myocardial Infarction (TIMI) bleeding criteria, and site of bleeding was recorded when available. Major bleeding events were defined as intracranial hemorrhage, clinically overt signs of hemorrhage with hemoglobin drop ≥5 g/dL, or fatal bleeding. Minor bleeding events were defined as clinically overt bleeding resulting in hemoglobin drop ≥3 to 5 g/dL.9 Data are reported using descriptive statistics. Continuous variables are described as means with SDs. Categorical variables are described as frequency distributions. All data were analyzed using Microsoft Excel 2007.
Results During the study period, 100 consecutive patients were identified and included for analysis. The majority of patients receiving bivalirudin were Caucasian (69%) and male (58%) and had an average age of 65.1 ± 13.1 years (Table 1). Three patients had thrombocytopenia (platelets < 100 000/ mm3) prior to cardiac catheterization; however, no patient had a documented history of heparin-induced thrombocytopenia. The majority of patients had a CrCl >30 mL/min; bivalirudin dose adjustment for renal dysfunction was required in only 13% of patients. Despite the higher bleeding risk in STEMI patients, the majority of patients had a diagnosis of unstable angina (73%; Table 2). Vascular access was obtained through the radial artery in 46% of patients. Unfractionated heparin was commonly administered (92%) to patients prior to PCI. Antiplatelet agents were administered immediately prior to or during PCI in
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
429
Baek et al Table 1. Baseline Characteristics (n = 100).a
40
Age (years) Female Caucasian History of HIT Platelets < 100 000/mm3 Estimated CrCl (mL/min) CrCl ≥ 30 mL/min CrCl = 10-29 ESRD/Dialysis Anticoagulation prior to admission Warfarin LMWH Dabigatran
65.1 ± 13.1 42 69 0 3 87 8 5 3 1 1
Abbreviations: HIT, heparin-induced thrombocytopenia; CrCl, creatinine clearance; ESRD, end-stage renal disease; LMWH, low-molecular-weight heparin. a Continuous data are displayed as mean ± SD. Categorical data are displayed as numbers.
Table 2. Procedural Data (n = 100). Procedure Characteristics
n (%)
ACS type Unstable angina STEMI NSTEMI Vascular access site Femoral Radial Adjunctive antiplatelet (n = 74) Clopidogrel 600 mg Clopidogrel 300 mg Prasugrel 60 mg Ticagrelor 180 mg Other Heparin (n = 92) Prior to catheterization During catheterization
73 (73%) 19 (19%) 8 (8%) 54 (54%) 46 (46%) 13 (18%) 12 (16%) 4 (5%) 36 (49%) 9 (12%) 41 (45%) 51 (55%)
Abbreviation: ACS, Acute Coronary Syndrome.
74% of patients; ticagrelor (49%) was the most frequently used agent in this setting. Only 34% of the patients who received bivalirudin were classified as high risk for PCI-related bleeding according to the NCDR clinical risk algorithm (Figure 2). The risk of procedure-related bleeding was calculated as low in 23% of patients. Of note, the NCDR risk score was unable to be calculated retrospectively for 11 patients because of a lack of complete NCDR variable documentation in the electronic medical record. Treatment with bivalirudin was determined to be relatively safe and effective. No incidents of acute stent
Number of Patients (%)
Characteristics
32
34
30 23 20 11 10 Low
Intermediate
High
NCDR Risk Category
Unable to Calculate
Figure 2. National Cardiovascular Data Registry (NCDR) risk stratification (n = 100).
thrombosis were identified in the study population. Only 1 case of provisional GPI use was documented. Three patients did not receive appropriate doses of bivalirudin because of discrepancies in weight and/or renal function, none of which resulted in a procedural complication. No episodes of TIMI major bleeding were noted in the study population; 1 patient met criteria for TIMI minor bleeding. Per the NCDR bleeding risk assessment, this patient was high risk and experienced bleeding at a femoral vascular access site.
Discussion This retrospective, observational study of patients undergoing PCI demonstrates that bivalirudin is a safe and effective antithrombotic agent. The rate of bailout GPI and incidence of bleeding events was noted to be lower than reported in previous studies.6-8 It is interesting to note that bivalirudin was frequently utilized in patients with a low or moderate risk of procedure-related bleeding in our study population. Because bivalirudin is a significantly more expensive antithrombotic compared with unfractionated heparin, an opportunity may exist to establish a more cost-effective use strategy. Bivalirudin has historically been considered an attractive anticoagulation option during PCI based on the improved safety profile and similar efficacy demonstrated by early landmark trials.9,10 This has recently been challenged by the results of the recently published HEATPPCI. This open-label, single-center, randomized controlled trial is unique because previous studies used outcomes that combined safety and efficacy. The primary efficacy outcome of all-cause mortality, cerebrovascular accident, reinfarction, or unplanned target lesion revascularization was significantly lower in the cohort of patients receiving unfractionated heparin compared with bivalirudin. With similar frequency of GPI use in each group, no significant difference in major bleeding events was noted between agents.10 Changes in clinical practice since the publication of the landmark bivalirudin trials should be considered. Vascular
Downloaded from aop.sagepub.com at The University of Hong Kong Libraries on May 10, 2015
430
Annals of Pharmacotherapy 49(4)
access through the radial artery is an important advance, used with increasing frequency, to reduce the rate of procedurerelated complications. In patients with radial access, the lower rate of bleeding may obviate the need for an antithrombotic agent, which reduces bleeding complications. Radial access was used in almost half of the patients who underwent PCI in our study. Further research regarding the role of bivalirudin in patients with radial access is needed. One small, retrospective study evaluating patients who received bivalirudin by radial or femoral access found no difference in net clinical adverse events and ischemic or bleeding complications between groups.10 Another important change in clinical practice since the publication of landmark bivalirudin trials is the use of novel P2Y12 inhibitors. A recent trial that used novel P2Y12 inhibitors in approximately half of all patients noted a significantly higher incidence of acute stent thrombosis in patients receiving bivalirudin. These results suggest that acute stent thrombosis remains a concern with bivalirudin, regardless of antiplatelet selection.6 Limitations of this study must be acknowledged. Given that this was a retrospective, observational study, the results of the study depended on the accuracy of the chart. Lack of documentation may have affected the study results, such as bleeding complications. TIMI bleeding criteria do not encompass all types of bleeding; hence, bleeding events may have occurred without being captured by our definition. The study also had a nonrandom convenience sample, which can increase the potential for selection bias and limit external validity. The small sample size may lead to larger uncertainties in measurement of bleeding and safety outcomes. We acknowledge that this study cannot ascertain how well bivalirudin is performing at our institution as compared with patients receiving heparin because our study was not controlled. Finally, it should be noted that this study was designed based on the 2009 NCDR bleeding risk assessment model. A newer bleeding risk assessment model was released in 2013, which incorporates 10 clinical factors. Because each bleeding risk assessment model was validated in a large cohort of patients, we would expect to see similar trends in bivalirudin use using the new bleeding risk assessment. Based on the results of this study, our institution has implemented criteria for bivalirudin use. Prior to implementation of these criteria, selection of the anticoagulation strategy was at the discretion of the interventional cardiologist. According to the newly implemented criteria, patients may receive anticoagulation with bivalirudin during PCI if any of the following criteria are met: high risk per NCDR bleeding risk assessment, documented history of heparininduced thrombocytopenia, platelet count
