Impact of Body Mass Index on Long-Term All-Cause Mortality after Percutaneous Coronary Intervention in African-Americans

Transcription

1 Original Contribution Impact of Body Mass Index on Long-Term All-Cause Mortality after Percutaneous Coronary Intervention in African-Americans Shyam Poludasu, MD, Erdal Cavusoglu, MD, Waqas Khan, MD, Jonathan D. Marmur, MD ABSTRACT: Background. Studies on body mass index (BMI) and post-percutaneous coronary intervention (PCI) outcomes have suggested an obesity paradox (lower post- PCI mortality rates in obese patients compared to patients of normal weight). Hypothesis. We assessed the hypothesis that BMI is an independent predictor of post-pci long-term mortality in African-Americans. Methods. We evaluated 777 patients (146 with normal BMI [BMI >18.5 to < 25 kg/m 2 ], 261 overweight patients [BMI 25 to < 30 kg/m 2 ], and 370 obese patients [BMI 30 kg/m 2 ]) who underwent PCI during January 2003 to August Results. After a median follow up of 4 ± 1 years, the overall mortality rate was 10.5% (82 deaths). The survival rate was 84%, 90% and 92% in the normal BMI, overweight and obese groups, respectively (p = by log-rank test; hazard ratio relative to the obese group was 2.2 for the normal weight and 1.2 for the overweight groups). After adjustment for baseline clinical and procedural characteristics (age, smoking history, chronic renal insufficiency, end-stage renal disease, left main coronary artery intervention, preprocedure hemoglobin, left ventricular ejection fraction and successful PCI) using a Cox proportional hazards model, there was no significant difference in the long-term all-cause mortality rate among the three groups (p = 0.93). Conclusion. BMI does not appear to be an independent predictor of long-term mortality after PCI in African-Americans. J INVASIVE CARDIOL 2009;21:20 25 Key words: Body mass index, African-Americans, percutaneous coronary intervention Obesity is a well-established risk factor for coronary artery disease (CAD). 1 Accordingly, a higher number of patients with obesity are undergoing percutaneous coronary intervention (PCI). 2 Obese patients are assumed to have worse outcomes after PCI. However, recent studies 3,4 and a meta-analysis 5 have suggested an apparent protective effect of obesity on post-pci mortality. Because obesity would be predicted to increase the From SUNY Downstate Medical Center, Brooklyn, New York. The authors report no conflicts of interest regarding the content herein. Manuscript submitted July 14, 2008, provisional acceptance given September 8, 2008, final version accepted October 3, Address for correspondence: Jonathan D. Marmur, MD, FACC, SUNY Downstate Medical Center, Cardiology, 450 Clarkson Avenue, Box 1257, Brooklyn, NY jonathan@marmur.com likelihood of adverse outcomes, the term obesity paradox has been used if obese patients are found to have lower mortality compared to the non-obese group. 3,4 Obesity is a rapidly growing epidemic in the United States and African-Americans are disproportionately affected. Approximately 45% of African-Americans were reported to be obese in , compared to only 30.6% and 36.8% obese Caucasians and Hispanics, respectively. 6 The long-term mortality after PCI is unknown in this population subgroup. Accordingly, the present study evaluated the effect of body mass index (BMI) on long-term all-cause mortality in a cohort of African-American patients who underwent PCI at an inner city single-center institution. Methods Study cohort. This is a retrospective analysis of 777 consecutive African-American patients from our bolus-only glycoprotein IIb/IIIa (GPI) database and activated clotting time (ACT)-guided intravenous (IV) dalteparin dosing during PCI. We retrospectively analyzed 1397 (1001 consecutive patients during January 2003 to August 2004, 292 consecutive patients during December 2004 to August 2005 in two separate studies 7,8 on GPI bolus-only therapy during PCI, and 104 consecutive patients who underwent PCI with IV dalteparin from December 2005 to August 2006) 9 patients who underwent PCI at a single center. From this database we included all African- Americans in the current study. The first PCI occurring during this time period is taken as the index procedure. Patients without a valid social security number had to be excluded from the study, as the longer-term follow up was obtained using the Social Security Death Index. The hospital s institutional review board approved the study. Demographic data including height and weight of the patient, periprocedural and laboratory data were collected by reviewing charts and hospital records. The in-hospital events and the length of stay are also recorded. Patients with ST-segment elevation myocardial infarction (STEMI) were not treated with bolus-only GPI and therefore were excluded from the bolus-only GPI database. Periprocedural medication. All patients were loaded with aspirin 325 mg and clopidogrel mg prior to PCI. All patients in our GPI bolus-only studies received an initial bolus of IV 40 units/kg unfractionated heparin (UFH) plus a bolus of GPI (0.25 mg/kg abciximab, 25 g/kg tirofiban or 180 g/kg x 2 eptifibatide) at the beginning of the intervention, which is 20 The Journal of Invasive Cardiology

2 BMI and Mortality after PCI in African-Americans Cumulative Survival Kaplan-Meier survival curves for normal weight, overweight and obese patients undergoing PCI Log-rank test p = BMI = > 30 kg/m 2 BMI = > 25 to < 30 kg/m 2 BMI = > 18.5 to < 25 kg/m Follow Up (days) Figure 1. Kaplan-Meier survival curves according to body mass index (BMI) divided as normal (> 18.5 to < 25 kg/m 2 ) overweight ( 25 to < 30 kg/m 2 ), and obese ( 30 kg/m 2 ). At 4 ± 1 years, the survival rate was 84% in the group with normal weight, 90% in the overweight group, and 92% in the obese group (p = by log-rank test). the standard practice at our institution. 7,8 If necessary, supplemental boluses of UFH were administered to achieve a target activated clotting time (ACT) of at least 200 seconds. In the ACT-guided IV dalteparin study, all patients received an initial IV bolus of 50 IU/kg of dalteparin. A GPI bolus at the beginning of the intervention was given at the discretion of the operator. Those patients who did not reach a target ACT with an initial bolus of dalteparin were given supplemental boluses as described previously. 9 The ACT was measured using the Hemochron device (ITC Technidyne Corp., Edison, New Jersey). Coronary intervention. Coronary interventional procedures were performed according to standard techniques via a femoral approach. The choice of drug-eluting stent (DES) or bare-metal stent (BMS) implantation was at the discretion of the operator. Femoral vascular closure devices (Angio-Seal, St. Jude Medical, Minnetonka, Minnesota, or Perclose, Perclose, Inc., Redwood City, California) were used unless contraindicated. Serial monitoring of cardiac biomarkers was performed every 8 hours for 24 hours after PCI, and hemoglobin levels were measured every 24 hours until the patient was discharged. At the time of discharge, all patients were instructed to take aspirin indefinitely. Clopidogrel was recommended for at least 1 month for the patients who received BMS and for at least 9 12 months for those who received DES. Follow up. We used all-cause mortality as our long-term outcome. The Social Security Death Index 10 was used to determine the time of death on April 24, Definitions. Body mass index was calculated as weight in kilograms divided by the square of the patient s height in meters (kg/m 2 ). Based on BMI, the study cohort was divided into normal (> 18.5 to < 25 kg/m 2 ), overweight ( 25 to <30 kg/m 2 ), and obese ( 30 kg/m 2 ). Hyperlipidemia was diagnosed in patients who received lipidlowering medication or who had a history of low-density lipoprotein (LDL) > 130 mg/dl. Smoking was defined as the inhaled use of cigarettes, cigars or pipes in any quantity in the last 2 weeks prior to the procedure. Left ventricular dysfunction was defined as left ventricular ejection fraction (LVEF) < 50% on echocardiography or coronary angiography in the past. Chronic renal insufficiency (CRI) was defined as patients with creatinine levels > 1.5 mg/dl, but who were not dependent on chronic dialysis. Patients were considered to have end-stage renal disease (ESRD) if they were dependent on chronic dialysis. Non-STsegment elevation myocardial infarction (NSTEMI) was defined as chest pain and enzymatic evidence of myocardial infarction (MI), but without ST-segment elevation on the presenting electrocardiogram (ECG). Unstable angina pectoris was defined according to the classification scheme based on the guidelines of the American College of Cardiology/American Heart Association. 11 Post-PCI MI was defined according to the Thrombolysis in Myocardial Infarction (TIMI) criteria. 12 A new MI was defined by biochemical or ECG criteria: the MB isoform of creatinine kinase (CK-MB) at least three times the upper limit of the normal range in at least one blood sample, or the finding of abnormal Q-waves 2 contiguous leads. For patients with a recent MI who had an elevated CK-MB level before the procedure, a value of more than three times the upper limit of normal and at least 50% above the baseline value was required to meet the definition. Bleeding was defined according to the Randomized Evaluation in the PCI Linking Angiomax to Reduced Clinical Events (REPLACE-2) criteria. 13 Although many studies in the cardiovascular literature used a definition of bleeding based on the Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) bleeding, 14 in the current study, we defined bleeding based on the REPLACE-2 criteria, as these definitions are more sensitive and were well validated as predictors of higher mortality on long-term follow up. 15 Major bleeding was defined as either any intracranial, intraocularor, retroperitoneal or clinically overt bleeding with a drop of hemoglobin by 3 g/dl, or any drop of hemoglobin by 4 g/dl, or the transfusion of 2 more units of packed red blood cells. Minor bleeding was defined as clinically overt bleeding not meeting the above criteria. Thrombocytopenia was defined as a fall in the platelet count below 100,000/ l, or a decrease by 25% below baseline values, in the event the initial platelet count was less than 100,000/ l. Statistical analysis. Continuous variables are presented as mean ± standard deviation (SD). Categorical variables are presented as percentages. The chi-square test was used to compare the differences between categorical variables. The Kruskal Wallis test was used to compare continuous variables without a normal distribution. For in-hospital outcomes, logistic regression analyses were Vol. 21, No. 1, January

3 POLUDASU, et al. Table 1. Baseline characteristics. Normal BMI (n = 146) Over weight (n = 261) (> 18.5 to < 25 kg/m 2 ( 25 to < 30 kg/m 2 ) Obese (n = 370) ( 30 kg/m 2 ) Age (mean ± SD in years) 66 ± ± ± 10 < Females, n (%) 83 (56.8) 142 (54.4) 146 (39.5) < Indication, n (%) NSTEMI 26 (17.8) 45 (17.2) 51 (13.8) 0.37 Unstable angina 66 (45.2) 97 (37.2) 151 (40.8) 0.28 Stable angina and others 54 (37) 121 (46.4) 170 (45.9) 0.13 Medical history, n (%) Hypertension 131 (89.7) 227 (87) 335 (90.5) 0.35 Diabetes mellitus 67 (45.9) 121 (46.4) 209 (56.5) LV dysfunction (LVEF < 50) 54 (37) 72 (28) 105 (28) 0.10 Smoking 49 (33.6) 82 (31.4) 101 (27.3) 0.30 Family history 28 (19.2) 65 (24.9) 119 (32.2) Previous CAD 74 (50.7) 154 (59) 201 (54.3) 0.24 CRI (creatinine > 1.5 mg/dl) 44 (30.1) 71 (27.2) 71 (19.2) 0.01 ESRD 13 (8.9) 14 (5.4) 16 (4.3) 0.12 Medications, n (%) Beta-blocker 114 (78.1) 218 (83.5) 298 (80.5) 0.38 Statin 119 (81.5) 207 (79.3) 306 (82.7) 0.60 ACE-inhibitor 97 (66.4) 191 (73.2) 260 (70.3) 0.35 GP IIb/IIIa inhibitor, n (%) 136 (93.2) 241 (92.3) 337 (91.1) 0.7 Abciximab 33 (22.6) 81 (31) 116 (31.4) 0.12 Eptifibatide 49 (33.6) 102 (39.1) 147 (39.7) 0.41 Tirofiban 54 (37) 59 (22.6) 74 (20) < Ejection fraction (mean ± SD) 48 ± ± ± Dalteparin during PCI, n (%) 12 (8.2) 23 (8.8) 36 (9.7) 0.84 BMI = body mass index; SD = standard deviation; NSTEMI = Non-ST-segment elevation myocardial infarction; CAD = coronary artery disease; CRI = chronic renal insufficiency; ESRD = end-stage renal disease; GP = glycoprotein conducted to assess the univariate effect of predictors of interest and the extent of confounding or interaction of these predictors with covariates statistically significant in univariate analysis. Event-free survival rates during the follow-up period were estimated by the Kaplan-Meier method and tested by the log-rank statistic. Multivariable Cox proportional-hazards regression modeling was used to adjust for the covariates that had significant impact (p < 0.05) on all-cause mortality in the study population. Sequential models were fit with the initial model including no covariates (unadjusted); the final model included covariates selected by step-wise methods that had significant impact on all-cause mortality. Covariates tested in a step-wise fashion in the multivariable Cox proportional-hazards regression model include age, gender, BMI, indication for the procedure (NSTEMI, unstable angina and stable angina), history of any medical problems (hypertension, diabetes mellitus, CAD, congestive heart failure, chronic renal insufficiency, end-stage renal disease), smoking history, family history of CAD, medications (beta-blockers, ACE-inhibitors and statins), type of platelet glycoprotein IIb/IIIa inhibitor used, LVEF, target coronary artery, number of diseased vessels, multivessel intervention, type of stent used, length and diameter of the stents used, percentage stenosis of the vessels, maximum ACT, successful PCI and in-hospital outcomes (post-pci MI, bleeding complications and thrombocytopenia). P-values < 0.05 were considered to indicate statistical significance. All statistical analyses are performed utilizing SPSS software Version 15.0 (SPSS, Inc., Chicago, Illinois). Results Baseline characteristics. The baseline characteristics are listed in Table 1 and procedural characteristics are presented in Table 2. There were 146 patients in the normal BMI arm, 261 patients in the overweight arm and 370 patients in the obese arm. Higher BMI groups had younger patients, fewer proportions of women, higher proportions of patients with diabetes, family history of CAD, fewer proportions of patients who received tirofiban during PCI and had higher mean LVEF (Table 1). In the procedural characteristics, higher BMI groups had a lower mean peak ACT (Table 2). In-hospital outcomes. The in-hospital outcomes are presented in Table 3. There was no difference in the in-hospital composite endpoint of death, non-fatal MI, repeat revascularization and bleeding complications (adjusted p = 0.86; covariates in the logistic regression model were NSTEMI, mean diameter of the stent, mean length of the stent and number of stents used). We did not perform a logistic regression analysis for the in-hospital individual complications due to a very low incidence of these complications, which might not yield meaningful results. However, the unadjusted non-fatal MI rate was lower in the normal BMI arm (2.1% in the normal BMI arm 22 The Journal of Invasive Cardiology

4 BMI and Mortality after PCI in African-Americans Table 2. Procedural characteristics. Normal BMI (n = 146) (> 18.5 to < 25 kg/m 2 Overweight (n = 261) ( 25 to < 30 kg/m 2 ) Obese (n = 370) ( 30 kg/m 2 ) Target coronary artery, n (%) Left main 2 (1.4) 5 (1.9) 5 (1.4) 0.84 Left anterior descending 55 (37.7) 104 (40) 149 (40.3) 0.86 Left circumflex 36 (24.7) 60 (23) 89 (24.1) 0.92 Others (ramus intermedius 56 (38.4) 81 (31) 114 (30.8) 0.22 LIMA and SVG) Number of vessels involved, n (%) One 43 (29.5) 93 (35.6) 136 (36.8) 0.28 Two 49 (33.6) 88 (33.7) 111 (30) 0.55 Three 50 (34.2) 72 (27.6) 113 (30.5) 0.37 Stent type, n (%) Bare-metal 59 (40.4) 106 (40.6) 174 (47) 0.19 Drug-eluting 91 (62.3) 150 (57.5) 204 (55.1) 0.33 Multivessel intervention, n (%) 6 (4.1) 9 (3.4) 10 (2.7) 0.69 Mean peak ACT ± SD 291 ± ± ± PCI successful, n (%) 141 (96.6) 259 (99.2) 360 (97.3) 0.14 Mean number of stents 1.3 ± ± ± Hospital stay (mean ± SD in days) 2.9 ± ± ± Mean length of stent 16.3 ± ± ± Mean diameter of stent 3.0 ± ± ± LIMA = left internal mammary artery; SVG = saphenous venous graft; ACT = activated clotting time; PCI = percutaneous coronary intervention; SD = standard deviation compared to 6.1% in the overweight arm, and 5.1% in the obese arm), and the in-hospital bleeding complications were higher in the normal BMI arm (4.8% in normal BMI arm compared to 1.9% in the overweight and obese arms). All-cause mortality. After a median follow up of 4 ± 1 years, the overall mortality rate was 10.5% (82 deaths). The survival rate was 84% in the group with a normal BMI, 90% in the overweight group, and 92% in the obese group (p = by log-rank test; hazard ratio relative to the obese group was 2.2 for the normal weight and 1.2 for the overweight groups). However, after adjustment for baseline clinical and procedural characteristics (age, smoking history, chronic renal insufficiency, end-stage renal disease, left main coronary artery intervention, preprocedure hemoglobin, LVEF and successful PCI) using a Cox proportional hazards model, there was no significant difference in long-term all-cause mortality between the three groups (p = 0.93). Table 3. In-hospital outcomes. Variable Normal BMI (n = 146) (> 18.5 to < 25 kg/m 2 Over weight (n = 261) ( 25 to < 30 kg/m 2 ) Obese (n = 370) ( 30 kg/m 2 ) * Composite endpoint, % (n) 10 (6.8) 21 (8) 26 (7) 0.86 Death, % (n) Non-fatal MI, % (n) 3 (2.1) 16 (6.1) 19 (5.1) Bleeding, % (n) Any 7 (4.8) 5 (1.9) 7 (1.9) Major bleeding 4 (2.7) 2 (0.8) 3 (0.8) Minor bleeding 3 (2.1) 3 (1.1) 4 (1.1) Repeat revascularization Thrombocytopenia, % (n) 1 (0.7) 4 (1.5) 5 (1.4) *Composite endpoint includes in-hospital death, non-fatal myocardial infarction, repeat revascularization and bleeding complications; MI = myocardial infarction Vol. 21, No. 1, January

5 POLUDASU, et al. Table 4. Multivariate predictors of all-cause mortality at the end of follow-up period (4 ± 1 years). Variable Hazard Ratio 95% Confidence Interval Age Chronic renal insufficiency (creatinine > 1.5) End stage renal disease < Left ventricular ejection fraction < Left main intervention Unsuccessful PCI Baseline hemoglobin Smoking (in past 2 weeks) PCI = percutaneous coronary intervention Discussion Our study shows that in the African American population, elevated BMI is associated with reduced mortality after PCI. This apparent protective effect of higher BMI on Kaplan Meier survival analysis disappeared after adjustment for covariates with significant impact on mortality. However, overweight and obese patients having similar long-term mortality compared to normal BMI patients can still be considered an obesity paradox as one would speculate that higher BMI patients are at increased risk for long-term mortality due to the detrimental effects of obesity. We believe that the current report is the first study to evaluate the effect of BMI on long-term mortality after PCI in African-American population. A recent meta-analysis by Oreopoulos et al 5 showed that the obese patients undergoing PCI had lower short-term and longterm mortality compared to patients with normal BMI. However, this protective effect of higher BMI seemed to wear off with increasing duration of follow up. Among the studies with longer term follow up (follow-up period over 3 years), the posthoc analysis of the BARI (Bypass Angioplasty Revascularization Investigation) trial by Gurm et al 16 and the post hoc analysis of ARTS (Arterial Revascularization Therapies Study) by Gruberg et al 17 did not show any difference in long-term mortality in the obese group compared to the non-obese groups. In our study the median follow up period was 4 ± 1 years and our results are consistent with the above-referenced studies with longer-term follow up. The probable reason for not having mortality benefit in obese patients in the long-term follow-up studies compared to non-obese patients is that the better shortterm mortality after PCI in the obese group is probably offset by the adverse outcomes during the long-term follow up due to other comorbid conditions associated with obesity. An additional finding in our study is that lower BMI patients had higher mean peak ACT during the procedure and these patients had a trend towards lower non-fatal MI and higher in-hospital bleeding complications. On the other hand, higher BMI patients had lower mean peak ACT during the procedure and these patients had a trend towards higher non-fatal MI and lower in-hospital bleeding. From these findings, it can be postulated that despite weight-adjusted dosing of unfractionated heparin, higher BMI patients probably achieve subtherapuetic levels of anticoagulation during PCI and tend to have lower in-hospital bleeding complications. Recent studies suggest that bleeding complications post PCI are stronger predictors of long-term mortality compared to periprocedural MI. 15 This might partly explain the apparent protective effect of obesity on long-term mortality. This finding of higher bleeding complications in lower BMI patients was also noticed in the study by Gurm et al 3 and Powell et al. 18 Patients with higher BMI in our study were significantly younger compared to patients with lower BMI, a finding similar to other studies, 5 which could also be one of the reasons for similar long-term mortality in patients with higher BMI. Gurm et al 15 have postulated that probably large vessel size and the relatively favorable artery/device ratio could be one of the reasons for better outcomes in higher BMI patients. However, in our patient population the mean length of the stents, the mean diameter of the stents used to treat the lesions, and the procedural success rates were not significantly different between the three groups. This is a subanalysis of the data collected to evaluate the outcomes of patients undergoing PCI with a GPI bolus-only strategy. In our initial two studies on platelet GPI bolus-only with in-hospital outcomes as major endpoints, abciximab, eptifibatide and tirofiban given as a high-dose bolus (25 g/kg) had similar in-hospital ischemic endpoints. 7,8 However, patients who received abciximab as bolus-only during PCI had higher in-hospital major bleeding complications compared to the patients who received eptifibatide. 7 In the long-term follow up of the patients undergoing PCI with a bolus-only GPI strategy included in the initial two studies, abciximab was associated with higher long-term all-cause mortality at 4 ± 1 years compared to eptifibatide. Tirofiban was not significantly different from either eptifibatide or abciximab in regard to long-term mortality. 19 In the current study, the fraction of patients receiving abciximab or eptifibatide were similar but a higher proportion of patients in the normal BMI arm received tirofiban compared to the other two arms (p = for normal BMI vs. overweight patients; p < for normal BMI vs. obese patients). However, 24 The Journal of Invasive Cardiology

6 BMI and Mortality after PCI in African-Americans in our GPI bolus-only long-term follow-up study, patients who received tirofiban did not have a significant difference in longterm mortality compared to the other two agents. Also, in the multivariate analysis done in the current study, we tested the type of GPI as a covariate and found that it was not a significant predictor of long-term mortality, probably due to the smaller sample size. We believe that the GPI study might not have had any impact on the outcome of the current study. Study limitations. This study is a retrospective analysis and has all the inherent limitations of a retrospective study. Our sample size might be too small to assess the differences in longterm mortality in three subgroups. Our study population had only 9 patients with a lower BMI (underweight patients, BMI < 18.5 kg/m 2 ) and had to be excluded from the analysis, as this group could not be analyzed separately. The baseline and procedural characteristics between the three arms were reasonably similar, but there were certain differences that had to be adjusted. Despite adjustment, there could be residual confounding variables that would lead to the observed differences in outcomes between the three arms. We excluded patients with STEMI, as these patients were not included in our database. We had to exclude patients without valid Social Security numbers because we included all-cause mortality as our endpoint, which was obtained using the Social Security Death Index. 10 The exclusion of patients without Social Security numbers might have introduced a selection bias. However, all-cause mortality is considered the most unbiased endpoint in longterm outcome studies, especially if they are retrospective, as other chronic medical conditions may affect the attributed cause of death. 20,21 Despite these limitations, we conclude that BMI may not be an independent predictor of long-term all-cause mortality in African-Americans after PCI. These preliminary findings would best be confirmed in large, prospective PCI studies. References 1. Hubert HB, Feinleib M, McNamara PM, Castelli WP. Obesity is an independent risk factor for cardiovascular disease: A 26-year follow-up of participants in the Framingham Heart Study. Circulation 1983;67: Minutello RM, Chou ET, Hong MK, et al. Impact of body mass index on in-hospital outcomes following percutaneous coronary intervention (report from the New York State Angioplasty Registry). Am J Cardiol 2004;93: Gurm HS, Brennan DM, Booth J, et al. Impact of body mass index on outcome after percutaneous coronary intervention. Am J Cardiol 2002;90: Gruberg L, Weissman NJ, Waksman R, et al. The impact of obesity on the short-term and long-term outcomes after percutaneous coronary intervention: The obesity paradox?. J Am Coll Cardiol 2002;39: Oreopoulos A, Padwal R, Norris CM, et al. Effect of obesity on short- and long-term mortality postcoronary revascularization: A meta-analysis. Obesity (Silver Spring). 2008;16: Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of overweight and obesity in the United States, JAMA 2006;295: Marmur JD, Poludasu S, Agarwal A, et al. Bolus-only platelet glycoprotein IIb-IIIa inhibition during percutaneous coronary intervention. J Invasive Cardiol 2006;18: Marmur JD, Poludasu S, Agarwal A, et al. High-dose tirofiban administered as bolus-only during percutaneous coronary intervention. J Invasive Cardiol 2008;20: Marmur JD, Poludasu S, Feit A, et al. Activated clotting time (ACT)-guided intravenous dalteparin dosing during percutaneous coronary intervention. J Invasive Cardiol 2008;20: Curb JD, Ford CE, Pressel S et al. Ascertainment of vital status through the National Death Index and the Social Security Administration. Am J Epidemiol 1985; 121: Braunwald E, Antman EM, Beasley JW, et al. ACC/AHA guideline update for the management of patients with unstable angina and non ST-segment elevation myocardial infarction 2002: Summary article: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina). Circulation 2002;102: Bovill EG, Terrin ML, Stump DC, et al. Hemorrhagic events during therapy with recombinant tissue-type plasminogen activator, heparin, and aspirin for acute myocardial infarction. Results of the Thrombolysis in Myocardial Infarction (TIMI), Phase II Trial. Ann Intern Med 1991;115: 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. JAMA 2003;289: The GUSTO investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N Engl J Med 1993;329: Lincoff AM, Kleiman NS, Kereiakes DJ, et al for the REPLACE-2 Investigators. Longterm efficacy of bivalirudin and provisional glycoprotein IIb/IIIa blockade vs. heparin and planned glycoprotein IIb/IIIa blockade during percutaneous coronary revascularization: REPLACE-2 randomized trial. JAMA 2004;292: Gurm HS, Whitlow PL, Kip KE, Investigators B. The impact of body mass index on short- and long-term outcomes inpatients undergoing coronary revascularization. J Am Coll Cardiol 2002;39: Gruberg L, Mercado N, Milo S, et al. Impact of body mass index on the outcome of patients with multivessel disease randomized to either coronary artery bypass grafting or stenting in the ARTS trial. Am J Cardiol 2005;95: Powell BD, Lennon RJ, Lerman A, et al. Association of body mass index with outcome after percutaneous coronary intervention. Am J Cardiol 2003;91: Marmur JD, Poludasu S, Lazar J, Cavusoglu E. Long-term mortality after bolus-only administration of abciximab, eptifibatide or tirofiban during percutaneous coronary intervention. Catheter Cardiovasc Interv (in press). 20. Gottlieb SS. Dead is dead artificial definitions are no substitute. Lancet 1997;349: Lauer MS, Blackstone EH, Young JB, Topol EJ. Cause of death in clinical research: Time for a reassessment? J Am Coll Cardiol 1999;34: Vol. 21, No. 1, January