Abstract: Objectives. We sought to describe trends in the use of preprocedural P2Y12 inhibitors and their clinical impact in patients undergoing percutaneous coronary intervention (PCI). Background. Oral P2Y12 inhibitors are ubiquitously used medications; however, the specific timing of initial P2Y12 inhibitor administration remains intensely debated. Methods. Our study population comprised 74,053 consecutive patients undergoing PCI at 47 hospitals in Michigan from January 2013 through June 2015. In-hospital outcomes included stent thrombosis, bleeding, need for transfusion, and death. Hierarchical logistic regression, propensity matching, and targeted maximum likelihood estimation were used to adjust for baseline patient differences and clustering, and to minimize bias. Results. Of 24,733 patients who received a preprocedural P2Y12 inhibitor, 82% received clopidogrel, 8% prasugrel, and 10% ticagrelor. Preprocedural administration of P2Y12 inhibitors declined during the study (49.3% to 24.8%; P<.001), and varied greatly across hospitals (14.5%-95.9%). No significant differences in outcomes were observed between patients receiving preprocedural clopidogrel and a matched cohort of those not receiving any preprocedural P2Y12 inhibitor (stent thrombosis: adjusted odds ratio [OR], 1.55; 95% confidence interval [CI], 0.30-7.84; bleeding: OR, 0.96; 95% CI, 0.63-1.46; transfusion: OR, 1.03; 95% CI, 0.69-1.55; and death: OR, 0.95; 95% CI, 0.38-2.37). Similar findings were demonstrated for preprocedural ticagrelor and prasugrel. Results from a subgroup analysis of patients with non-ST segment elevation acute coronary syndrome (n = 28,072) were consistent with the overall findings. Conclusions. There was a substantial decline in the rate of preprocedural P2Y12 inhibitor administration during the study. Furthermore, there were no significant differences in outcomes between patients treated with preprocedural P2Y12 inhibitors and those who were not.
J INVASIVE CARDIOL 2017;29(10):340-351. Epub 2017 April 15.
Key words: P2Y12 inhibitor, pretreatment, percutaneous coronary intervention, cardiovascular outcomes
Oral antiplatelet medications, such as aspirin and P2Y12 inhibitors, are important cornerstones of pharmacologic therapy in patients undergoing percutaneous coronary intervention (PCI). Historically, clopidogrel has been the most widely used P2Y12 inhibitor, though it has well-recognized limitations including differential metabolism and efficacy due to genetic polymorphisms as well as a relatively delayed onset of action.1-3 Partially due to these limitations, two new P2Y12 inhibitors, ticagrelor and prasugrel, were developed. These drugs exhibited higher levels of platelet inhibition and a more rapid onset of action compared with clopidogrel.4-6 In large, randomized, controlled trials, ticagrelor and prasugrel demonstrated superior outcomes compared with clopidogrel in patients with acute coronary syndrome (ACS) undergoing PCI.7,8
Although there have been significant advances in antiplatelet therapy in the past decade, the optimal timing of P2Y12 inhibitor administration has not been conclusively established and remains a topic of continued debate.9-12 Results from the PCI-CURE13 and CREDO14 studies suggested benefit with clopidogrel pretreatment in patients with non-ST segment elevation acute coronary syndrome (NSTE-ACS) undergoing PCI. More recently, publication of the ACCOAST trial15 and meta-analyses10,12 have called the practice of P2Y12 inhibitor pretreatment into question, with the ACCOAST study demonstrating increased bleeding without a significant reduction in ischemic outcomes in patients presenting with non-ST segment elevation myocardial infarction (NSTEMI) pretreated with prasugrel. Although clinical practice guidelines have historically recommended P2Y12 inhibitor administration at the time of NSTE-ACS presentation, in light of the findings from the ACCOAST trial along with the ongoing debate over the optimal timing of P2Y12 inhibitor administration, recent clinical practice guidelines have generally de-emphasized recommendations on the optimal timing of P2Y12 inhibitor administration.16-19
According to a prior study, approximately 70% of patients undergoing PCI receive their initial dose of clopidogrel during or after PCI.20 It is unclear, however, if these practice patterns persist in contemporary practice given the increasing use of ticagrelor and prasugrel.21,22 Therefore, using a registry of all PCIs performed at non-federal hospitals in the state of Michigan, we sought to: (1) describe patterns in the utilization of specific P2Y12 inhibitors; (2) describe trends in the timing of P2Y12 inhibitor administration relative to PCI; and (3) assess the association between preprocedural P2Y12 inhibitor administration and clinically important in-hospital outcomes.
Study population. Our study population comprised all patients undergoing PCI between January 1, 2013 and June 30, 2015 at every non-federal hospital in the state of Michigan and enrolled in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) registry. This included a total of 47 hospitals, of which 14 performed primary PCI in ST-segment elevation myocardial infarction (STEMI) patients only without on-site cardiac surgery back-up. This registry has been previously described in detail elsewhere.23,24
We excluded patients who received >1 unique P2Y12 inhibitor preprocedurally; those without a recorded femoral or radial access site; and individuals with a contraindication to P2Y12 or glycoprotein IIb/IIIa inhibitors, bivalirudin, heparin, and/or where treatment was given as part of a clinical trial. The timing of medication administration was classified as occurring in the preprocedural, intraprocedural, or postprocedural setting, relative to the patient’s visit to the catheterization lab when PCI was performed. The preprocedural period was defined as the time from hospital admission to the time of entry into catheterization lab. The intraprocedural period was defined as the time from entry to the catheterization lab to discharge from the catheterization lab. The postprocedural period included the time from catheterization lab discharge to hospital discharge.
The classification of P2Y12 inhibitor timing was determined by when the patient received their first P2Y12 inhibitor. For example, a patient who received clopidogrel in the preprocedural setting and ticagrelor in the postprocedural setting was classified as having received preprocedural clopidogrel, not postprocedural ticagrelor, for the purposes of this study. We did not collect detailed information regarding the specific P2Y12 inhibitor dosage administered.
Study outcomes. All outcomes were measured during the incident hospitalization when PCI was performed. In-hospital outcomes included bleeding, need for transfusion, stent thrombosis, and death due to any cause. Bleeding was defined as per the National Cardiovascular Data Registry,25 and included an event within 72 hours of PCI that was associated with any of the following: a drop in hemoglobin ≥3 g/dL; transfusion of whole blood or packed red blood cells; and an intervention or surgery at the site of bleeding to reverse, stop, or correct the bleeding. The need for transfusion was defined as the receipt of ≥1 unit of red blood cell or whole blood transfusion after PCI. Stent thrombosis was defined as thrombosis at the site of original stent placement demonstrated on repeat angiography.
Statistical analysis. A hierarchical logistic mixed-effects regression model was utilized to assess whether the incidence of preprocedural medication administration varied between hospitals after accounting for patient clinical and demographic differences as well as temporal trends. The model included preprocedural demographic and clinical patient characteristics as fixed effects (see Supplementary Table S1; supplementary tables available at end of article), a normally distributed hospital-level random intercept, and an outcome of preprocedural use of either clopidogrel, prasugrel, or ticagrelor. From the fitted model, intraclass correlation (ICC) and median odds ratio (MOR) values along with corresponding 95% confidence intervals (CIs) were estimated using parametric bootstrap simulation.26
To assess outcomes associated with preprocedural P2Y12 inhibitor administration, patients were classified by receipt of a preprocedural P2Y12 inhibitor, and patients who received a preprocedural P2Y12 inhibitor were further stratified by the specific drug administered: clopidogrel, prasugrel, or ticagrelor. Baseline characteristics were reported as means for continuous variables and proportions for categorical variables. Differences between treatment groups were compared using Fisher’s exact testing for categorical variables and Wilcoxon rank-sum and Student’s t-tests for continuous variables. Additionally, absolute standardized differences (ASDs) were estimated and a 10% threshold for ASD was used as an indicator of baseline imbalance.
Propensity-score matching and targeted maximum likelihood estimation (TMLE) were used for the assessment of whether administration of preprocedural P2Y12 inhibitors was independently associated with the incidence of clinical outcomes. For each preprocedural P2Y12 inhibitor, a logistic regression model was utilized to estimate propensity scores among patients either receiving the specific agent or no P2Y12 inhibitor preprocedurally, with the variables in Supplementary Table S1 included as predictors. Patients receiving preprocedural P2Y12 inhibitors were then matched on a 1:1 basis to similar patients who did not receive a preprocedural P2Y12 inhibitor using a greedy matching algorithm. Matched pairs were further required to be treated at the same facility, have the same coronary artery disease presentation category (STEMI, non-STEMI, unstable angina, stable angina, and other), and have propensity scores within a caliper of 0.25 propensity score standard deviations on the logit scale in order to be included in the matched cohort. TMLE utilizing main-effects logistic regression was applied to the propensity-matched groups to estimate the causal effect of preprocedural P2Y12 inhibitor administration on outcomes. TMLE results were reported as adjusted odds ratios (aORs), adjusting for potential confounding variables listed in Supplementary Table S1, and utilizing robust standard errors accounting for within-site correlation in the outcome for inference. In a secondary analysis, we applied the same TMLE methodology to the entire study cohort without prior matching. A subgroup analysis was also performed by combining the matched cohorts constructed for each specific preprocedural P2Y12 inhibitor and stratifying by ACS presentation.
Briefly, TMLE is a methodology for evaluating the causal effect of one particular parameter of the data distribution considering the remaining parameters to be nuisance parameters, targeting the maximum likelihood estimate of the parameter of interest in a way that reduces bias.27 TMLE employs an orthogonal factorization decomposing the likelihood into an assignment mechanism specifying the likelihood of the covariate of interest values (the presumed “treatment”) given the remaining observed covariate data (g model), and an outcome mechanism specifying the probability of the outcome conditional on the entire observed covariate matrix including the covariate of interest (Q model). TMLE has the attractive quality of being doubly robust, so that if either the g or Q models are correctly specified the model will produce unbiased effect estimates.
Hierarchical generalized mixed-effects regression models were fitted using the lme4 R package.28 TMLE was performed using the TMLE package in R.29 All analyses were performed using R version 126.96.36.199
Over the course of the study period, a total of 76,433 patients underwent PCI at 47 hospitals. In total, 2380 (3.1%) patients met ≥1 exclusion criteria, of which 429 (0.6%) received >1 unique P2Y12 inhibitor preprocedurally, 251 (0.3%) did not have documentation of a femoral or radial access site, and 1955 (2.6%) had a contraindication(s) to P2Y12 inhibitors, glycoprotein IIb/IIIa inhibitors, bivalirudin, heparin, and/or received treatment as part of a clinical trial. Of the remaining 74,053 patients included in the primary analysis, 24,733 (33%) received a P2Y12 inhibitor preprocedurally (Figure 1). Clopidogrel was the most frequently administered P2Y12 inhibitor in the preprocedural setting (n = 20,275; 82%), followed by ticagrelor (n = 2436; 10%) and prasugrel (n = 2022; 8%).
Temporal and hospital-level trends. The use of preprocedural P2Y12 inhibitors declined from 49.3% to 24.8% during the study period (P for trend <.001) (Figure 2). A decline was seen across all subgroups by ACS presentation, with the most substantial decline occurring in patients with unstable angina (52.9% to 22.2%) and non-ACS presentations (53.0% to 11.8%), and less so for patients with NSTEMI (44.7% to 35.4%%) and STEMI (41.9% to 28.7%). Specifically, preprocedural clopidogrel administration declined substantially, whereas the use of ticagrelor increased (Figure 3). A later discharge date was significantly associated with decreased use of preprocedural clopidogrel (aOR, 0.945 for a 1-month difference; P<.001; 95% CI, 0.943-0.947) after adjusting for other patient characteristics and accounting for hospital clustering. Contemporaneously, as preprocedural P2Y12 inhibitor use declined, there was an increase in initial administration of P2Y12 inhibitors in the intraprocedural or postprocedural period, from 48.8% in the first quarter of 2013 to 71.4% in the second quarter of 2015 (P for trend <.001) (Figure 2). Over the study period, we noted significant increases in the initial administration of intraprocedural or postprocedural ticagrelor, from 6.4% in the first quarter of 2013 to 22.7% in the second quarter of 2015 (P for trend <.001) and clopidogrel from 30.4% to 37.4% over the same period (P for trend <.001) (Supplementary Figure S1; supplementary figures available at end of article).
There was significant variation in the frequency of preprocedural P2Y12 inhibitor use across the participating hospitals (range, 14.5%-95.9%) (Supplementary Figure S2) that could not be accounted for by patient-level variables or temporal trend (MOR, 3.33; 95% CI, 2.45-4.74; ICC, 0.22; 95% CI, 0.14-0.32) (Supplementary Figure S3). Approximately 22% of the variability in the use of preprocedural P2Y12 inhibitors was attributable to the site. Among the 33 hospitals performing both elective and primary PCI, the frequency of preprocedural P2Y12 inhibitor administration ranged from 14.5%-53.5% (Supplementary Figure S2).
Preprocedural P2Y12 inhibitor administration and clinical outcomes. Patients who received a preprocedural P2Y12 inhibitor (n = 24,733) were more likely to have hypertension, diabetes, dyslipidemia, heart failure, prior PCI, prior coronary artery bypass graft surgery, cerebrovascular disease, and peripheral artery disease (Supplementary Table S2). Prior to propensity matching and multivariable adjustment, in-hospital mortality was slightly less frequent in patients who received a preprocedural P2Y12 inhibitor vs those who did not (1.4% vs 1.7%; P<.01; ASD = 2.4%). Stent thrombosis occurred infrequently, both in patients who received preprocedural P2Y12 inhibitors and in those who did not (0.1% vs 0.3%, respectively; P<.001; ASD = 2.7%). The baseline characteristics of patients stratified by the specific preprocedural P2Y12 inhibitor administered are shown in Supplementary Table S3. Propensity matching produced generally well-balanced pairs on the matched covariates (Supplementary Figure S4); however, there were significant differences in procedural characteristics between the matched cohorts, such as increased use of procedural glycoprotein IIb/IIIa inhibitors in patients who did not receive a preprocedural P2Y12 inhibitor (Table 1).
After propensity matching, there were no significant differences in outcomes between patients who received preprocedural P2Y12 inhibitors and those who did not, regardless of the preprocedural P2Y12 inhibitor administered (Figure 4). Specifically, among patients who received preprocedural clopidogrel (n = 16,066) vs the matched cohort of patients who did not receive any preprocedural P2Y12 inhibitor (n = 16,066), there were no significant differences in stent thrombosis (aOR, 1.55; 95% CI, 0.30-7.84), bleeding (aOR, 0.96; 95% CI, 0.63-1.46), transfusion (aOR, 1.03; 95% CI, 0.69-1.55), or death (aOR, 0.95; 95% CI, 0.38-2.37). Using TMLE in the entire cohort, results were consistent, and no significant association was observed between administration of any preprocedural P2Y12 inhibitor and in-hospital outcomes (Supplementary Figure S5).
In a subgroup analysis performed by combining the three specific preprocedural P2Y12 inhibitor propensity-matched cohorts, the use of preprocedural P2Y12 inhibitors was not associated with any significant difference in outcomes among patients with NSTE-ACS (n = 28,072): stent thrombosis (aOR, 1.52; 95% CI, 0.29-7.92), bleeding (aOR, 1.03; 95% CI, 0.78-1.36), transfusion (aOR, 1.07; 95% CI, 0.78-1.48), or death (aOR, 1.03; 95% CI, 0.21-5.04) (Supplementary Table S4). Of note, reliable estimates could not be generated for patients who did not have ACS due to the sparse occurrence of clinical outcome events.
Using a large regional registry of patients undergoing PCI, we evaluated important trends and outcomes associated with the timing of P2Y12 inhibitor administration relative to PCI. Our study has three major findings. First, the administration of preprocedural P2Y12 inhibitors significantly declined over the course of the study period, primarily due to a substantial reduction in preprocedural clopidogrel administration. Second, there were no significant differences in outcomes between patients who were treated with P2Y12 inhibitors preprocedurally compared with those who were not, regardless of the specific preprocedural P2Y12 inhibitor administered. Notably, a subgroup analysis of patients with NSTE-ACS demonstrated results consistent with the overall findings. Third, there was substantial hospital-level variability in the administration of preprocedural P2Y12 inhibitors.
Consistent with a prior study assessing the timing of antiplatelet medication administration relative to PCI,20 we found only a moderate proportion of patients received preprocedural P2Y12 inhibitors at the beginning of the study period. Over the course of the study period (30 months), this number declined from 49.3% to 24.8%, consistent with a marked reduction in the use of these medications in the preprocedural setting. There are a number of potential explanations for this, the most compelling of which may be related to the concomitant increase in the use of a newer third-generation P2Y12 inhibitor, ticagrelor, in the intraprocedural and postprocedural settings. As mentioned previously, third-generation P2Y12 inhibitors have a faster onset of action and achieve higher levels of platelet inhibition than clopidogrel.4-6 Given these important pharmacodynamic differences, physicians may be less inclined to pretreat patients with clopidogrel, knowing that faster-acting agents are available. Furthermore, as the rate of ad hoc PCI increases,31 interventional cardiologists who are considering administration of a P2Y12 inhibitor prior to diagnostic angiography must carefully weigh the probability of PCI vs surgical revascularization, given that pretreatment may result in delayed surgical management in individuals with multivessel coronary artery disease.19 Similarly, cardiac surgeons may request that P2Y12 inhibitors not be given until the coronary anatomy is defined in order to reduce the risk of delaying surgery or the risk of bleeding if the patient requires urgent coronary artery bypass graft surgery.
The decline in preprocedural P2Y12 inhibitor administration may also be related to the publication of the ACCOAST trial in September 2013 and a subsequent discussion of the study’s results at the quarterly BMC2 meeting. Findings from the ACCOAST15 trial together with results from the TRITON-TIMI 388 and TRILOGY32 trials have led to current practice guidelines recommending prasugrel therapy only after the patient’s coronary anatomy is defined and a decision is made to pursue PCI in NSTE-ACS.16,19 Importantly, although the ACCOAST trial was performed in patients presenting with NSTEMI,15 the largest decline in preprocedural P2Y12 inhibitor use in our study was seen in patients presenting with unstable angina or non-ACS presentations. This suggests that physicians may have extrapolated the study’s findings from patients presenting with NSTEMI to lower-risk presentations.
Our study supports and extends the findings of ACCOAST by demonstrating no significant differences in outcomes after PCI not only in patients treated with preprocedural prasugrel, but also in those treated with preprocedural clopidogrel and ticagrelor, compared with those who were not treated with a preprocedural P2Y12 inhibitor. Of note, we did not find an increased risk of bleeding with preprocedural administration in the overall cohort or in the subgroup of matched patients experiencing NSTE-ACS, a finding that may have been driven by the relatively lower use of glycoprotein IIb/IIIa inhibitors in this population. Also, in the current study, the mean and standard deviation age and weight in the preprocedural prasugrel matched cohort were 60.6 ± 9.9 years and 93.8 ± 23.4 kg, respectively. The low frequency of patients with advanced age or low body weight may have contributed to the lack of a significant difference in bleeding outcomes. Patients with advanced age (≥75 years) and/or low body weight (<60 kg), which are known risks for bleeding with prasugrel, were included in the ACCOAST trial and contributed substantially to the excess bleeding events noted in the pretreatment arm.15,33 Lastly, the majority of bleeding seen in ACCOAST was access-site related, with approximately 60% of patients undergoing femoral artery catheterization. In the current study, the prasugrel matched cohort had femoral access rates of nearly 75%. Nevertheless, no significant difference in bleeding was noted.
We uncovered substantial variability in the use of preprocedural P2Y12 inhibitors at the 47 hospitals participating in this statewide registry. The reason for this heterogeneity remains unclear, although institutional policies, drug formularies, and individual practice patterns likely play an important role. It is worth noting that all hospitals participating in this registry are involved in statewide quality improvement initiatives.34 Therefore, the presence and degree of practice heterogeneity in regions where hospitals do not participate in such initiatives remain unknown and require further investigation.
Study limitations. The findings from this study should be interpreted with specific caveats. First, this is a registry-based cohort study; therefore, the findings represent associations and should not be interpreted as implying causation. Second, as in all observational studies, we were unable to account for unmeasured confounders that could have influenced the use and/or timing of specific P2Y12 inhibitors, although we applied sophisticated statistical methods including propensity-matching and TMLE techniques to minimize bias. Third, although prior studies have demonstrated a reduction in postprocedural myocardial infarction in patients receiving P2Y12 inhibitors prior to PCI, assessment of postprocedural myocardial infarction by biomarker ascertainment was not routinely performed in this registry and thus could not be evaluated in this analysis. Fourth, we were unable to differentiate between patients on chronic P2Y12 inhibitor therapy prior to hospitalization from those receiving it for the first time in the hospital. How this subgroup was managed (ie, re-loaded, switched to a different P2Y12 inhibitor, etc), and its effect on overall outcomes is the focus of ongoing research. Fifth, this registry does not collect data on posthospital outcomes. Since prior work has demonstrated improved long-term outcomes with a strategy that included pretreatment,13,14,35 further research evaluating posthospital outcomes is needed. Lastly, the exact dosage and timing data (ie, the number of hours prior to PCI) on P2Y12 inhibitor administration relative to PCI were not collected. These factors may have affected clinical outcomes based upon the pharmacologic properties of the specific drug used.
Given the ongoing debate over the timing of P2Y12 inhibitor administration relative to PCI, we examined the timing of P2Y12 inhibitor administration and associated outcomes in a large, contemporary cohort of patients undergoing PCI in the state of Michigan and found a substantial decline in the use of preprocedural P2Y12 inhibitors over the study period. Furthermore, preprocedural P2Y12 inhibitor administration was not associated with significant differences in outcomes compared with no preprocedural administration. This finding was consistently demonstrated across all P2Y12 inhibitors studied, including clopidogrel, prasugrel, and ticagrelor, as well as in the important subgroup of patients presenting with NSTE-ACS. Future studies are needed to determine the optimal timing of P2Y12 inhibitor administration in patients undergoing PCI.
Acknowledgments. The authors are indebted to all the study coordinators, investigators, and patients who participated in the BMC2 registry.
Supplemental Tables and Figures
1. Mega JL, Close SL, Wiviott SD, et al. Cytochrome p-450 polymorphisms and response to clopidogrel. N Engl J Med. 2009;360:354-362.
2. Frelinger AL, Bhatt DL, Lee RD, et al. Clopidogrel pharmacokinetics and pharmacodynamics vary widely despite exclusion or control of polymorphisms (CYP2C19, ABCB1, PON1), noncompliance, diet, smoking, co-medications (including proton pump inhibitors), and pre-existent variability in platelet function. J Am Coll Cardiol. 2013;61:872-879.
3. Wallentin L. P2Y(12) inhibitors: differences in properties and mechanisms of action and potential consequences for clinical use. Eur Heart J. 2009;30:1964-1977.
4. Husted S, Emanuelsson H, Heptinstall S, Sandset PM, Wickens M, Peters G. Pharmacodynamics, pharmacokinetics, and safety of the oral reversible P2Y12 antagonist AZD6140 with aspirin in patients with atherosclerosis: a double-blind comparison to clopidogrel with aspirin. Eur Heart J. 2006;27:1038-1047.
5. Wiviott SD, Trenk D, Frelinger AL, et al. Prasugrel compared with high loading- and maintenance-dose clopidogrel in patients with planned percutaneous coronary intervention: the Prasugrel in Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation-Thrombolysis in Myocardial Infarction 44 trial. Circulation. 2007;116:2923-2932.
6. Gurbel PA, Bliden KP, Butler K, et al. Randomized double-blind assessment of the ONSET and OFFSET of the antiplatelet effects of ticagrelor versus clopidogrel in patients with stable coronary artery disease: the ONSET/OFFSET study. Circulation. 2009;120:2577-2585.
7. Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2009;361:1045-1057.
8. Wiviott SD, Braunwald E, McCabe CH, et al. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2007;357:2001-2015.
9. Windecker S, Kolh P, Alfonso F, et al. 2014 ESC/EACTS guidelines on myocardial revascularization: the Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) Developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J. 2014;35:2541-2619.
10. Bellemain-Appaix A, O’Connor SA, Silvain J, et al. Association of clopidogrel pretreatment with mortality, cardiovascular events, and major bleeding among patients undergoing percutaneous coronary intervention: a systematic review and meta-analysis. JAMA. 2012;308:2507-2516.
11. Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. Circulation. 2011;124:e574-e651.
12. Bellemain-Appaix A, Kerneis M, O’Connor SA, et al. Reappraisal of thienopyridine pretreatment in patients with non-ST elevation acute coronary syndrome: a systematic review and meta-analysis. BMJ. 2014;349:g6269.
13. Mehta SR, Yusuf S, Peters RJ, et al. Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in patients undergoing percutaneous coronary intervention: the PCI-CURE study. Lancet. 2001;358:527-533.
14. Steinhubl SR, Berger PB, Mann JT 3rd, et al. Early and sustained dual oral antiplatelet therapy following percutaneous coronary intervention: a randomized controlled trial. JAMA. 2002;288:2411-2420.
15. Montalescot G, Bolognese L, Dudek D, et al. Pretreatment with prasugrel in non-ST-segment elevation acute coronary syndromes. N Engl J Med. 2013;369:999-1010.
16. Roffi M, Patrono C, Collet JP, et al. 2015 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: task force for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation of the European Society of Cardiology (ESC). Eur Heart J. 2016;37:267-315.
17. Jneid H, Anderson JL, Wright RS, et al. 2012 ACCF/AHA focused update of the guideline for the management of patients with unstable angina/non-ST-elevation myocardial infarction (updating the 2007 guideline and replacing the 2011 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2012;60:645-681.
18. Hamm CW, Bassand JP, Agewall S, et al. ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: the task force for the management of acute coronary syndromes (ACS) in patients presenting without persistent ST-segment elevation of the European Society of Cardiology (ESC). Eur Heart J. 2011;32:2999-3054.
19. Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC guideline for the management of patients with non-ST elevation acute coronary syndromes: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;64:e139-e228.
20. Dean BB, Yu HT, Bae JP, et al. Pattern of clopidogrel use in hospitalized patients receiving percutaneous coronary interventions. Am J Health Syst Pharm. 2010;67:1430-1437.
21. Sandhu A, Seth M, Dixon S, et al. Contemporary use of prasugrel in clinical practice: insights from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium. Circ Cardiovasc Qual Outcomes. 2013;6:293-298.
22. Karve AM, Seth M, Sharma M, et al. Contemporary use of ticagrelor in interventional practice (from Blue Cross Blue Shield of Michigan Cardiovascular Consortium). Am J Cardiol. 2015;115:1502-1506.
23. Kline-Rogers E, Share D, Bondie D, et al. Development of a multicenter interventional cardiology database: the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) experience. J Intervent Cardiol. 2002;15:387-392.
24. Moscucci M, Rogers EK, Montoye C, et al. Association of a continuous quality improvement initiative with practice and outcome variations of contemporary percutaneous coronary interventions. Circulation. 2006;113:814-822.
25. NCDR CathPCI Registry v4.4 Coder’s Data Dictionary. Accessed at https://www.ncdr.com/WebNCDR/docs/public-data-collection-documents/cathpci_v4_codersdictionary_4-4.pdf?sfvrsn=2.
26. Merlo J, Chaix B, Ohlsson H, et al. A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. J Epidemiol Community Health. 2006;60:290-297.
27. Gruber S, van der Laan MJ. Targeted maximum likelihood estimation: a gentle introduction. UC Berkeley Division of Biostatistics Working Paper Series. August 2009: Working Paper 252. Accessed at http://biostats.bepress.com/ucbbiostat/paper252/.
28. lme4: Linear mixed-effects models using Eigen and S4 [computer program]. Version R Package version 1.1-72014.
29. Gruber S, van der Laan MJ. tmle: an R Package for targeted maximum likelihood estimation. J Statistical Software. 2012;51:1-35.
30. R: A language and environment for statistical computing [computer program]. Vienna, Austria: R Foundation for Statistical Computing; 2015.
31. Hannan EL, Samadashvili Z, Walford G, et al. Predictors and outcomes of ad hoc versus non-ad hoc percutaneous coronary interventions. JACC Cardiovasc Intervent. 2009;2:350-356.
32. Roe MT, Armstrong PW, Fox KA, et al. Prasugrel versus clopidogrel for acute coronary syndromes without revascularization. N Engl J Med. 2012;367:1297-1309.
33. Valgimigli M. Pretreatment with P2Y12 inhibitors in non-ST segment-elevation acute coronary syndrome is clinically justified. Circulation. 2014;130:1891-1903; discussion 1903.
34. Share DA, Campbell DA, Birkmeyer N, et al. How a regional collaborative of hospitals and physicians in Michigan cut costs and improved the quality of care. Health Affairs. 2011;30:636-645.
35. Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med. 2001;345:494-502.
36. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604-612.
From the 1Division of Cardiovascular Medicine, Department of Medicine, University of Michigan, Ann Arbor, Michigan; 2Division of Cardiology, Department of Medicine, William Beaumont Hospital, Heart and Vascular, Royal Oak, Michigan; 3Division of Cardiology, Department of Medicine, Henry Ford Hospital, Detroit, Michigan; 4St. John Hospital, Detroit, Michigan; and 5Veterans Affairs Medical Center, Ann Arbor, Michigan.
Funding. This work was supported by the Blue Cross Blue Shield of Michigan and Blue Care Network as part of the Blue Cross Blue Shield of Michigan Value Partnerships program. The funding source supported data collection at each site and funded the data-coordinating center but had no role in study concept, interpretation of findings, or in the preparation, final approval, or decision to submit the manuscript.
Disclaimer. Although Blue Cross Blue Shield of Michigan (BCBSM) and BMC2 work collaboratively, the opinions, beliefs and viewpoints expressed by the authors do not necessarily reflect the opinions, beliefs, and viewpoints of BCBSM or any of its employees.
Disclosure. The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Sukul is supported by the National Institutes of Health T32 postdoctoral research training grant (T32-HL007853). Dr Gurm reports personal fees from Osprey Medical; grants from the NIH. The remaining authors report no conflicts of interest regarding the content herein.
Manuscript submitted December 23, 2016; provisional acceptance given on December 28, 2016, final version accepted January 13, 2017.
Address for correspondence: Hitinder S. Gurm, MD, University of Michigan Cardiovascular Center, 2A 394, 1500 East Medical Center Drive, Ann Arbor, MI 48109-5853. Email: email@example.com