Previous studies have questioned the external validity of randomized controlled trial results of acute coronary syndrome (ACS) because of potential selection bias toward healthier patients. We sought to evaluate differences in clinical characteristics and management of patients admitted with non–ST-elevation ACS according to participation in clinical trials over the previous decade. The Canadian ACS I (1999 to 2001), ACS II (2002-2003), GRACE (2004-2007), and CANRACE (2008) were prospective, multicenter registries of patients admitted to hospitals with ACS. We examined 13,556 patients with non–ST-elevation ACS, of whom 1,126 (8.3%) participated in clinical trials. Data were collected on baseline characteristics, medication use at admission and discharge, in-hospital procedures, and in-hospital adverse events. Patients enrolled in clinical trials were younger, more likely to be men, and had fewer co-morbidities. They were significantly more likely to be on several guideline-recommended medications and were significantly more likely to undergo invasive procedures, including coronary angiography, percutaneous coronary intervention, and coronary bypass surgery (all p values <0.001). Unadjusted in-hospital (2.1% vs 0.7%, p = 0.001) and 1-year (8.9% vs 6.3%, p = 0.037) mortality rates were higher in non-enrolled patients. In multivariable analysis, patients who were older, women, had a history of heart failure, and increased creatinine levels on presentation were less likely to be enrolled into clinical trials. In conclusion, significant differences persist in baseline characteristics, treatment, and outcomes between patients enrolled and those not enrolled in clinical trials. Consequently, generalization of ACS clinical trials over the previous decade to the “real-world” patient may remain in question.
Our objectives were to examine the differences in baseline characteristics, in-hospital and postdischarge management, and outcomes of patients with acute coronary syndrome (ACS) according to participation in randomized controlled trials (RCTs) in 4 successive registries over 1 decade. We also sought to determine the factors that are independently associated with clinical trial enrollment.
Methods
The Canadian ACS I and II registries, international Global Registry of Acute Coronary Events (GRACE/GRACE ), and the Canadian Registry of Acute Coronary Events (CANRACE) were prospective, multicenter, observational studies of patients admitted to hospitals for suspected ACS. Details of their study designs have been published. Inclusion criteria for all registries were (1) ≥18 years of age and (2) symptoms compatible with ACS within 24 hours of presentation to hospital. The GRACE/CANRACE also required ≥1 of (1) abnormal cardiac biomarkers, (2) electrocardiographic changes, and/or (3) documented history of coronary artery disease. All registries excluded patients with ACS precipitated or accompanied by a serious concurrent illness, such as trauma or gastrointestinal bleeding.
The present study included only patients with non–ST-segment elevation (NSTE) ACS (n = 14,205). Patients with STE ACS or other final non-ACS diagnosis were excluded. Patients were enrolled from 51 centers in ACS I (n = 3,125, 1999 to 2001), 36 centers in ACS II (n = 1956, 2002 to 2003), and 48 centers in GRACE/CANRACE (n = 8,475, 2004 to 2008).
Data were collected on enrollment in a clinical trial during index hospitalization, but specific details about the trials were not recorded. Information regarding clinical trial enrollment was missing for 649 patients (4.6%) who were excluded from subsequent analyses.
Baseline patient characteristics, medication use within 24 hours of presentation and at discharge, procedures during hospitalization, and adverse events (myocardial reinfarction [re-MI], death, or major bleeding) were also included. Re-MI was defined as electrocardiographic changes or increase of cardiac markers >24 hours after presentation. Major bleeding was defined as (1) necessitating ≥2 U of packed red cells to be transfused, (2) a decrease ≥10% in hematocrit, or (3) leading to death or intracranial hemorrhage. At 1 year, patients who provided consent in the ACS I and II registries were contacted by the Canadian Heart Research Centre by trained telephone interviewers to ascertain 1-year survival and medication adherence. Ethics approval was obtained at each hospital, and all patients provided informed consent.
In the present study, we stratified patients into the group participating in clinical trials and the group not participating in clinical trials. Continuous variables are displayed as median and interquartile range, whereas categorical variables are expressed as percentage. We used Wilcoxon rank-sum test and chi-square test for comparisons of continuous and categorical variables, respectively. Multivariable logistic regression analysis was used to determine independent predictors of clinical trial enrollment within the registries. We considered components of the GRACE risk score, which is a validated predictor of in-hospital mortality in these registries, and factors previously shown to be associated with clinical trial enrollment, as candidate variables in the multivariable model. We also tested for selected interaction terms including time of enrollment (registry) with age and gender in the multivariable model. Variables not significantly associated with clinical trial enrollment were removed by backward elimination. Generalized estimating equations were used to account for clustering of patients within hospitals and to estimate adjusted odds ratios and 95% confidence intervals. Model discrimination and calibration were assessed using the c-statistic and Hosmer-Lemeshow goodness-of-fit test, respectively. A 2-tailed p value <0.05 was used for statistical significance. Statistical analyses were conducted using SPSS 15.0 (SPSS, Inc., Chicago, Illinois).
Results
Overall, 1,126 patients (8.3%) were enrolled in clinical trials; this study included 388 patients in ACS I registry, 216 patients in ACS II registry, and 522 patients in GRACE/CANRACE. Baseline characteristics are listed in Table 1 .
| Variable | All Registries | ACS I | ACS II | GRACE/CANRACE | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 13,556) | (n = 3,125) | (n = 1,956) | (n = 8,475) | |||||||||
| Not Enrolled (n = 12,430) | Enrolled (n = 1,126) | p Value | Not Enrolled (n = 2,737) | Enrolled (n = 388) | p Value | Not Enrolled (n = 1,740) | Enrolled (n = 216) | p Value | Not Enrolled (n = 7,953) | Enrolled (n = 522) | p Value | |
| Age (years) ⁎ | 68 (58,77) | 65 (56,74) | <0.001 | 68 (57,76) | 65 (56,75) | 0.018 | 67 (57,75) | 66 (55,75) | 0.67 | 68 (58,78) | 64 (56,72) | <0.001 |
| Men | 65.3% | 70.4% | 0.001 | 66.1% | 71.6% | 0.029 | 67.3% | 65.7% | 0.65 | 64.6% | 71.4% | 0.002 |
| Current smoker | 22.0% | 26.7% | <0.001 | 23.9% | 28.8% | 0.035 | 21.7% | 24.5% | 0.35 | 21.4% | 26.1% | 0.014 |
| Previous bleeding † | 2.5% | 0.9% | 0.002 | 1.7% | 0.5% | 0.089 | N/A † | N/A † | N/A † | 2.8% | 1.2% | 0.022 |
| Hypertension | 61.8% | 56.4% | <0.001 | 53.5% | 48.7% | 0.079 | 59.3% | 59.3% | 0.99 | 65.2% | 60.9% | 0.049 |
| Diabetes mellitus | 28.7% | 25.6% | 0.024 | 27.6% | 21.2% | 0.008 | 27.4% | 25.9% | 0.66 | 29.4% | 28.6% | 0.70 |
| Dyslipidemia | 55.3% | 53.9% | 0.36 | 48.0% | 45.2% | 0.30 | 56.4% | 55.1% | 0.72 | 57.6% | 59.8% | 0.32 |
| Previous angina pectoris | 52.2% | 48.7% | 0.024 | 63.1% | 54.5% | 0.001 | 56.6% | 53.7% | 0.42 | 47.5% | 42.2% | 0.020 |
| Previous myocardial infarction | 36.0% | 30.9% | 0.001 | 38.1% | 33.2% | 0.062 | 35.1% | 31.5% | 0.30 | 35.5% | 29.0% | 0.002 |
| Previous heart failure | 11.5% | 5.9% | <0.001 | 13.1% | 9.4% | 0.037 | 9.7% | 4.6% | 0.015 | 11.3% | 3.8% | <0.001 |
| Previous percutaneous intervention | 20.1% | 16.6% | 0.005 | 17.2% | 13.2% | 0.049 | 22.1% | 16.2% | 0.046 | 20.6% | 19.2% | 0.46 |
| Previous coronary artery bypass surgery | 14.5% | 13.4% | 0.32 | 15.2% | 14.4% | 0.70 | 14.9% | 10.6% | 0.091 | 14.2% | 13.8% | 0.82 |
| Previous stroke or transient ischemic attack | 10.1% | 5.0% | <0.001 | 9.3% | 5.7% | 0.019 | 10.3% | 5.6% | 0.027 | 10.3% | 4.2% | <0.001 |
| Heart rate at presentation (beats/min) ⁎ | 76 (65,91) | 76 (64,88) | 0.007 | 72 (62,88) | 72 (62,84) | 0.27 | 76 (65,90) | 78 (64,92) | 0.38 | 78 (66,93) | 77 (66,89) | 0.090 |
| Systolic blood pressure at presentation (mm Hg) ⁎ | 145 (128,165) | 148 (128,166) | 0.096 | 148 (130,169) | 150 (130,166) | 0.83 | 148 (130,168) | 145 (123,167) | 0.30 | 144 (126,163) | 148 (129,166) | 0.025 |
| Diastolic blood pressure at presentation (mm Hg) ⁎ | 80 (70,91) | 82 (72,93) | <0.001 | 80 (70,92) | 85 (74,94) | 0.003 | 80 (70,92) | 82 (72,93) | 0.23 | 80 (69,90) | 81 (71,92) | <0.001 |
| Killip class I | 83.6% | 87.4% | 82.4% | 82.5% | 84.4% | 80.6% | 83.8% | 93.0% | ||||
| Killip class II | 11.4% | 8.9% | 0.005 | 14.1% | 14.6% | 0.84 | 11.5% | 10.3% | 0.021 | 10.6% | 4.3% | <0.001 |
| Killip class III/IV | 4.9% | 3.7% | 3.5% | 2.9% | 4.2% | 9.0% | 5.6% | 2.7% | ||||
| ST-segment deviation | 28.1% | 28.8% | 0.62 | 17.6% | 24.7% | 0.001 | 23.9% | 34.3% | 0.001 | 32.6% | 29.5% | 0.14 |
| Abnormal initial biomarker | 44.5% | 50.8% | <0.001 | 38.5% | 46.9% | 0.002 | 53.2% | 54.6% | 0.68 | 44.6% | 52.0% | 0.001 |
| Initial creatinine (μmol/L) ⁎ | 92 (78,112) | 90 (77,106) | 0.001 | 90 (78,109) | 91 (79,108) | 0.89 | 91 (78,110) | 92 (76,107) | 0.63 | 92 (78,114) | 89 (76,104) | <0.001 |
| GRACE risk score ⁎ | 120 (97,149) | 115 (95,139) | <0.001 | 113 (92,140) | 114 (94,138) | 0.94 | 117 (96,154) | 124 (100,156) | 0.018 | 123 (99,153) | 113 (94,136) | <0.001 |
⁎ Median (interquartile range).
Patients enrolled in clinical trials were more likely to be administered acetylsalicylic acid, unfractionated heparin, and glycoprotein IIb/IIIa inhibitor within 24 hours of admission ( Table 2 ). They were also more likely to undergo coronary angiography, percutaneous coronary intervention (PCI), coronary artery bypass grafting, and any revascularization (PCI or coronary artery bypass grafting; Table 2 ). Patients enrolled in clinical trials were more likely to be discharged on acetylsalicylic acid and thienopyridine ( Table 3 ).
| Variable | All Registries | ACS I | ACS II | GRACE/CANRACE | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 13,556) | (n = 3,125) | (n = 1,956) | (n = 8,475) | |||||||||
| Not Enrolled (n = 12,430) | Enrolled (n = 1,126) | p Value | Not Enrolled (n = 2,737) | Enrolled (n = 388) | p Value | Not Enrolled (n = 1,740) | Enrolled (n = 216) | p Value | Not Enrolled (n = 7,953) | Enrolled (n = 522) | p Value | |
| Medication use within 24 hours of admission | ||||||||||||
| Aspirin | 91.5% | 96.2% | <0.001 | 90.1% | 97.2% | <0.001 | 92.9% | 94.0% | 0.55 | 91.7% | 96.4% | <0.001 |
| Ticlopidine/clopidogrel | 52.0% | 51.1% | 0.57 | 8.7% | 9.6% | 0.58 | 55.3% | 62.5% | 0.046 | 66.3% | 77.2% | <0.001 |
| Aspirin + thienopyridine | 49.4% | 49.2% | 0.91 | 7.0% | 8.6% | 0.25 | 52.1% | 58.3% | 0.082 | 63.6% | 75.6% | <0.001 |
| Heparin (any) | 87.3% | 91.4% | <0.001 | 88.7% | 95.9% | <0.001 | 90.4% | 89.8% | 0.78 | 86.2% | 88.8% | 0.090 |
| Unfractionated heparin | 33.0% | 38.4% | <0.001 | 51.2% | 54.1% | 0.27 | 26.5% | 32.9% | 0.047 | 28.0% | 28.9% | 0.66 |
| Low-molecular-weight heparin | 56.1% | 54.3% | 0.24 | 37.5% | 41.8% | 0.10 | 63.9% | 56.9% | 0.045 | 60.9% | 62.5% | 0.45 |
| Glycoprotein IIb/IIIa inhibitor | 5.6% | 19.3% | <0.001 | 2.7% | 24.2% | <0.001 | 11.2% | 14.8% | 0.12 | 5.4% | 17.5% | <0.001 |
| In-hospital procedures | ||||||||||||
| Coronary angiography | 54.5% | 67.8% | <0.001 | 38.9% | 43.6% | 0.078 | 63.9% | 69.9% | 0.080 | 57.9% | 85.0% | <0.001 |
| Percutaneous coronary intervention | 26.4% | 36.7% | <0.001 | 14.2% | 19.4% | 0.007 | 30.1% | 37.6% | 0.026 | 29.9% | 49.2% | <0.001 |
| Coronary artery bypass surgery | 4.6% | 7.4% | <0.001 | 3.9% | 7.8% | 0.001 | 10.0% | 16.4% | 0.004 | 3.7% | 3.3% | 0.66 |
| Percutaneous coronary intervention or coronary artery bypass surgery | 30.6% | 43.9% | <0.001 | 18.1% | 26.9% | <0.001 | 39.7% | 54.0% | <0.001 | 33.0% | 52.4% | <0.001 |
| Variable | All Registries ⁎ | ACS I | ACS II | GRACE/CANRACE ⁎ | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 10,963) | (n = 3,059) | (n = 1,921) | (n = 5,983) | |||||||||
| Not Enrolled (n = 9,958) | Enrolled (n = 1,005) | p Value | Not Enrolled (n = 2,675) | Enrolled (n = 384) | p Value | Not Enrolled (n = 1,706) | Enrolled (n = 215) | p Value | Not Enrolled (n = 5,577) | Enrolled (n = 406) | p Value | |
| Medications at time of discharge | ||||||||||||
| Aspirin | 85.9% | 92.0% | <0.001 | 85.5% | 93.2% | <0.001 | 89.3% | 85.1% | 0.060 | 85.0% | 94.6% | <0.001 |
| Ticlopidine/clopidogrel | 48.2% | 53.9% | 0.001 | 20.6% | 24.7% | 0.070 | 59.4% | 58.6% | 0.81 | 58.0% | 78.9% | <0.001 |
| Oral anticoagulant | 9.5% | 7.4% | 0.029 | 7.9% | 9.6% | 0.24 | 8.1% | 7.0% | 0.55 | 10.7% | 5.5% | 0.001 |
| β Blocker | 77.3% | 77.7% | 0.77 | 75.2% | 71.4% | 0.11 | 85.3% | 82.3% | 0.25 | 75.9% | 81.4% | 0.013 |
| Angiotensin-converting enzyme inhibitor | 57.3% | 59.7% | 0.14 | 52.1% | 47.0% | 0.063 | 66.3% | 65.1% | 0.73 | 57.0% | 68.9% | <0.001 |
| Angiotensin receptor blocker | 11.3% | 8.4% | 0.006 | 4.0% | 4.2% | 0.85 | 9.8% | 9.8% | 0.97 | 15.3% | 11.7% | 0.053 |
| Angiotensin-converting enzyme inhibitor and/or angiotensin receptor blocker | 66.3% | 66.3% | 0.98 | 55.7% | 49.3% | 0.020 | 74.5% | 73.5% | 0.75 | 68.9% | 78.6% | <0.001 |
| Lipid-lowering agent | 74.4% | 76.7% | 0.10 | 54.1% | 60.6% | 0.016 | 82.9% | 82.3% | 0.84 | 81.5% | 88.9% | <0.001 |
⁎ Data not available for patients who were transferred to another site.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
