Article Text
PDF
Abstract
Objective: To compare the characteristics, management, and outcomes of patients with non-ST-segment elevation acute coronary syndromes (NSTE ACS) who would have been eligible for inclusion in clinical trials of glycoprotein (GP) IIb/IIIa inhibitors with those of ineligible patients.
Design: Multinational, prospective, observational study (GRACE, Global Registry of Acute Coronary Events).
Setting: Patients hospitalised for a suspected acute coronary syndrome and enrolled in GRACE between April 1999 and December 2004.
Patients: 29 039 patients with NSTE ACS.
Main outcome measures: Characteristics and outcomes were compared for trial-eligible (75.0%) and trial-ineligible (25.0%) patients.
Results: GP IIb/IIIa inhibitors were administered to 20.0% of eligible and 15.3% of ineligible patients. Compared with eligible patients, ineligible patients who received GP IIb/IIIa inhibitors had significantly higher rates of hospital death (6.8% vs 3.7%) and major bleeding (4.9% vs 2.2%). After adjustment for their higher baseline risk, ineligible patients still experienced higher hospital death rates (adjusted odds ratio (OR) 1.60; 95% confidence interval (CI) 1.01 to 2.39), but not higher bleeding rates, than the eligible group. Use of GP IIb/IIIa inhibitors was associated with a trend towards lower 6-month mortality in eligible (OR 0.86, 95% CI 0.72 to 1.02) and ineligible (OR 0.82, 95% CI 0.65 to 1.05) patients compared with those in whom this therapy was not used.
Conclusions: GP IIb/IIIa inhibitors were markedly underused in the real-world population, irrespective of whether patients were trial-eligible or not. Despite the higher risk of ineligible patients, the benefits of GP IIb/IIIa inhibitors appear to be no less than in eligible patients.
Statistics from Altmetric.com
Several large randomised clinical trials (RCTs) have highlighted the role of intravenous platelet glycoprotein (GP) IIb/IIIa inhibitors in the management of patients with acute coronary syndromes (ACS).1–4 Data from these trials have shown that the blockade of GP IIb/IIIa receptors in selected patients leads to improved clinical outcomes when given alongside other proven standard medical therapies and percutaneous coronary intervention (PCI). These findings have led to strong endorsement of the integration of GP IIb/IIIa antagonist therapy in the management of patients with ACS in the American College of Cardiology/American Heart Association guidelines, particularly in high-risk patients with non-ST-segment elevation (NSTE) ACS and in those in whom an invasive approach is contemplated.5 6 However, “real-life” use of these drugs is less well-known; specifically, the proportion of patients who reflect the characteristics of those enrolled in clinical trials that are used to demonstrate the benefit of GP IIb/IIIa inhibitors in ACS is not well-characterised. Conversely, the profile of patients who are ineligible for inclusion in present clinical trials, but who receive these drugs, is also unknown, as is the impact of GP IIb/IIIa inhibitor use on their subsequent short-term and long-term outcomes.
The Global Registry of Acute Coronary Events (GRACE)7–9 provides a unique opportunity to analyse the “real-life” practice-based use of GP IIb/IIIa inhibitors in the management of patients with unstable angina and non-ST-segment elevation myocardial infarction (NSTEMI) and the impact of this therapy on patients’ clinical outcomes. The objectives of this study were to evaluate the baseline characteristics, hospital management, and risk profiles of patients receiving GP IIb/IIIa inhibitors, specifically in patients who would be eligible for inclusion in major RCTs10–12 that have established the value of these drugs in this context, and of patients who were ineligible for inclusion in these trials.
METHODS
Full details of the GRACE study have been published.13 GRACE is a multinational, observational cohort study that has been designed to reflect an unbiased population of patients with ACS, irrespective of geographical region. Currently, 106 hospitals located in 14 countries are participating in this large observational study. Patients entered in the registry were at least 18 years old and alive at the time of hospital presentation, were admitted with ACS as a presumptive diagnosis, and had at least one of the following: electrocardiogram changes consistent with ACS, serial increases in serum biochemical markers of cardiac necrosis, and/or documentation of coronary artery disease. The qualifying ACS must not have been precipitated by significant non-cardiovascular comorbidity such as trauma or surgery. Approximately 6 months after hospital discharge, patients were followed up to ascertain the occurrence of selected long-term study outcomes. Where required, study investigators received approval from their local hospital ethics or institutional review board and patients gave informed consent.
To enrol an unselected population of patients with ACS, sites were encouraged to recruit the first 10 to 20 consecutive eligible patients each month. Data were collected by a trained coordinator using a standardised case report form. Demographic characteristics, medical history, symptoms at hospital presentation, duration of prehospital delay, biochemical and electrocardiographic findings, treatment practices, and a variety of hospital outcome data were collected.
All patients who were admitted with a diagnosis of NSTE ACS were included in the present study. NSTE ACS was defined at the time of hospital admission as not having new, or presumed new, ST-segment elevation on the index electrocardiogram or new left bundle-branch block. All other patients were defined as having ST-segment elevation myocardial infarction (STEMI) at admission.
Clinical end points
The end points of the present investigation were in-hospital death and episodes of major bleeding. Death was defined as all-cause mortality. Major bleeding was defined as the occurrence of life-threatening bleeding requiring transfusion of two or more units of packed red blood cells, or resulting in an absolute decrease in haematocrit of 10% or greater, or death, or development of haematoma. The 6-month death end point was also defined as all-cause mortality from treatment initiation.
Definition of patients excluded from clinical trials
We used the exclusion criteria from the Platelet Receptor Inhibition in Ischemic Syndrome Management (PRISM),10 Platelet Receptor Inhibition in Ischemic Syndrome Management in Patients Limited by Unstable Signs and Symptoms (PRISM PLUS),11 and Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy (PURSUIT)12 clinical trials to define our ineligible comparison patient group in the GRACE registry. The main eligibility criteria for these trials are summarised in the Appendix.
Statistical analysis
Categorical variables are summarised as frequencies and percentages and were tested using Χ2 tests. Continuous variables are expressed as medians and 25th and 75th percentiles, and were evaluated using the Wilcoxon rank-sum test. Multivariable logistic regression modelling, using the GRACE risk scores for in-hospital mortality14 and major bleeding,15 was carried out to calculate adjusted odds ratios predicting in-hospital and 6-month outcomes of death, in-hospital major bleeding, and stroke. Death rates at 7 and 30 days in trial-eligible and trial-ineligible patients were also compared. All tests were double-sided and considered significant at α<0.05. SAS 9.1 (SAS Institute, Cary, NC) was utilised for all analyses.
RESULTS
A total of 29 039 patients with NSTE ACS were enrolled in the GRACE study from 106 hospitals in 14 countries between April 1999 and December 2004. Of these patients, 21 824 (75.0%) met the inclusion and exclusion criteria for three major clinical trials, PRISM,10 PRISM PLUS,11 and PURSUIT;12 the remaining 7215 patients (25.0%) did not meet the criteria for enrolment in these trials. Overall, only 5479 patients (19.0%) received GP IIb/IIIa inhibitors during hospitalisation (fig 1).
The use of GP IIb/IIIa inhibitors was significantly higher in trial-eligible patients than in trial-ineligible patients (20.0% vs 15.3%; p<0.001). The overall use of GP IIb/IIIa inhibitors was highest in centres located in the United States (31.6%), followed by Europe (19.0%), and Canada / Australia / New Zealand (10.1%); the lowest reported use of this therapy was in centres located in Argentina and Brazil (7.4%).
Baseline characteristics of eligible and ineligible patients
Patients in the trial-ineligible group had a higher risk profile than those in the trial-eligible group (table 1). Ineligible patients were more likely to be older, and had more risk factors for coronary artery disease and comorbidities present than trial-eligible patients. In contrast, trial-eligible patients had lower rates of prior revascularisation than trial-ineligible patients (table 1). Ineligible patients had a higher average risk of dying during admission than the eligible group (119 vs 132 points using the GRACE risk score for in-hospital mortality,14 p<0.001). Trial-eligible patients were more likely than trial-ineligible patients to be treated with GP IIb/IIIa inhibitors across all geographical regions; the rate of use in trial-ineligible patients was approximately 75% of that in eligible patients (fig 2). The use of GP IIb/IIIa inhibitors was twofold to fourfold higher in the USA compared with countries outside the USA. Downstream use of GP IIb/IIIa receptor antagonists (during or after PCI) and upstream use (medical treatment or before PCI) was almost evenly distributed in both the US and non-US countries.
Hospital management
There were minor differences in the rates of treatment with aspirin, thienopyridines, and beta-blockers between trial-eligible and trial-ineligible patients (table 2). However, trial-ineligible patients were more likely to receive angiotensin-converting enzyme inhibitors, calcium-channel blockers, and warfarin than eligible patients. Low-molecular-weight heparins were used less often, whereas unfractionated heparin was used to a greater extent in the trial-ineligible group than in the trial-eligible group. Cardiac catheterisation, PCI, and coronary artery bypass grafting were performed in fewer ineligible than eligible patients. Rates of PCI during the index hospitalisation in patients treated with GP IIb/IIIa inhibitors were higher in eligible than ineligible patients (71% vs 68%, respectively). For those not treated with GP IIb/IIIa inhibitors, the rates of PCI were slightly higher for eligible than ineligible patients (16% vs 14%, respectively). For both patient groups, the use of PCI was significantly higher in patients treated with GP IIb/IIIa inhibitors (71% vs 16%, and 68% vs 14%; p<0.001). Finally, 53% of eligible patients undergoing PCI received GP IIb/IIIa inhibitors compared with 46% of ineligible patients. Those who did not have PCI during hospitalisation had much lower rates of use of GP IIb/IIIa inhibitors (7.9% in eligible patients vs 6.2% in the ineligible cohort). Similarly, the use of thienopyridines was substantially higher in patients undergoing PCI (89% vs 85% in eligible and ineligible groups, respectively) than in those who did not receive PCI (eligible 25% vs ineligible 29%). The rate of GP IIb/IIIa use was higher among patients who received thienopyridines than in those who did not (36.3% vs 8% for eligible patients; 29.4% vs 6.5% for ineligible patients, respectively).
Hospital and long-term outcomes
Among patients receiving GP IIb/IIIa inhibitors, short-term hospital outcomes were consistently worse in the ineligible than the eligible group, with higher rates of major bleeding (6.8% vs 3.7%, respectively) and death (4.9% vs 2.2%, respectively) (fig 3). Both groups had similar, and low, rates of stroke. An identical pattern was seen among patients not receiving GP IIb/IIIa inhibitors (major bleeding 1.8% vs 3.1%; stroke 0.4% vs 1.2%; death 2.9% vs 6.5%).
The overall rates of in-hospital death and stroke were higher amongst patients who did not receive a GP IIb/IIIa inhibitor than in those who did (table 3). The rate of major bleeding was higher in patients who received a GP IIb/IIIa inhibitor, and was highest amongst those who were ineligible for inclusion in the randomised clinical trials (table 3).
Following risk adjustment using the GRACE risk scores for in-hospital mortality14 and major bleeding,15 ineligible patients treated with GP IIb/IIIa inhibitors had significantly higher rates of dying during hospitalisation, but had similar rates of major bleeding and stroke, compared with eligible patients (table 4a). Use of GP IIb/IIIa inhibitors was associated with a non-significant trend for reduced 6-month mortality in both the eligible and ineligible patient groups (table 4b). The use of GP IIb/IIIa inhibitors had no impact on the rates of stroke following hospital discharge.
Comparison of GRACE patients who were eligible for enrolment and RCT patients
Overall, GRACE patients who were eligible for inclusion in the major clinical trials of GP IIb/IIIa therapy had similar baseline characteristics to patients actually enrolled in the RCTs (table 1). Unadjusted mortality rates at 7 and 30 days for eligible patients were in the same range as those in the clinical trial populations (fig 4).
DISCUSSION
In this large, multinational, observational study, only one-fifth of patients with NSTE ACS who fulfilled the inclusion criteria for RCTs involving GP IIb/IIIa inhibitors were actually treated with these drugs during their index hospitalisation. These findings show that current use of GP IIb/IIIa inhibitors in the management of patients with NSTE ACS is considerably lower than expected, given the fact that NSTE ACS patients in GRACE were, on average, at intermediate and high risk for adverse outcomes.16 Even more striking is the fact that fewer than half of the patients undergoing PCI, a subgroup that has the greatest potential for benefit from these drugs, received GP IIb/IIIa inhibitors (53% in eligible patients vs 46% in ineligible patients). This finding could be attributed to poor access to PCI centres or to their availability, because the use of PCI was not restricted to patients admitted to PCI centres. A substantial fraction of patients with ACS were admitted initially to centres without PCI facilities (23%). Of those, 21% were subsequently transferred to centres with such facilities for PCI or other cardiac procedures.17 It could also be related to lack of financial support for GP IIb/IIIa inhibitor therapy as reflected by the geographical variations in the use of this efficacious therapy in a similar population of patients to those who participated in the clinical trials. In addition, these findings could be due to uncertainty surrounding the integration of multiple antiplatelet therapies (particularly aspirin, high-loading-dose clopidogrel, and intravenous GP IIb/IIIa inhibitors) into routine medical practice, given that the trials that established the value of GP IIb/IIIa blockers all antedate the “clopidogrel era” (ie the CURE trial).18
This marked underuse of GP IIb/IIIa inhibitors demonstrates a gap between available published evidence and its application in contemporary clinical practice, depriving high-risk patients of the benefits of these drugs.19 It is unlikely that this gap is related simply to the lag between accumulation of evidence and changes in practice, because the trials referred to were published several years before the present data were acquired. The underutilisation of GP IIb/IIIa inhibitors highlights the need for greater dissemination of guidelines and for the development of strategies to raise physicians’ awareness of the benefits of these drugs. Although the rate of use of GP IIb/IIIa inhibitors in this study population was low, it was consistently higher in patients who received thienopyridines and among those undergoing PCI in both cohorts – eligible and ineligible patients. In addition, the low rate of use of GP IIb/IIIa inhibitors in GRACE may not be attributed to clinicians’ belief that the combination of oral drugs is insufficient to protect against ischaemic events, otherwise the rate of use of GP IIb/IIIa inhibitors would have been lower among patients who received thienopyridines than among those who did not.
We identified several patient-related factors that contributed to the infrequent use of GP IIb/IIIa inhibitors in our study. These were older age, history of myocardial infarction, congestive heart failure, renal dysfunction, increased initial serum creatinine, and elevated diastolic blood pressure at admission. These characteristics are associated with a markedly increased risk of adverse outcomes among patients with NSTE ACS, making them a subgroup with the greatest potential for risk reduction with use of these therapies. Additional unmeasured factors (eg physician preferences, concerns about bleeding, and cost) may also have played a role in the low observed rates of use of GP IIb/IIIa inhibitors in our study sample. Understanding and overcoming existing barriers to the use of these drugs is crucial if the benefits noted in RCTs are to be realised in “real-world” practice.
Only a quarter of the GRACE patients with NSTE ACS would not have been eligible for enrolment in the trials of GP IIb/IIIa inhibitors, drugs that are perceived by many physicians to have a very narrow therapeutic index. Not surprisingly, these patients were at a substantially higher risk for adverse outcomes than eligible patients and, indeed, experienced higher death rates that remained significant after multivariable adjustment. Nevertheless, the use of GP IIb/IIIa inhibitors in these patients was apparently associated with a similar potential for improving survival to that seen in eligible patients. Although the bleeding rates were higher in ineligible patients receiving GP IIb/IIIa inhibitors than in ineligible patients not receiving any of these drugs, this rate was no different from that observed in trial-eligible patients who were treated with GP IIb/IIIa inhibitors. Therefore the benefits of GP IIb/IIIa inhibitors appear to be generalisable to “real-life” patients with ACS. These findings are similar to those seen with many cardiovascular drugs, in which equal or greater benefits have been demonstrated in trial-ineligible patients. Clearly, more data are needed to identify factors associated with the risks and benefits of GP IIb/IIIa inhibitors in ineligible patients to enable better selection of patients. Until that time, physicians should rely on their knowledge and experience and use these drugs judiciously in selected ineligible patients in whom they think that the benefits outweigh the risks.
Study limitations
GRACE is an observational study, and is prone to the limitations inherent in such investigations, including the collection of non-randomised data, missing or incomplete information, and potential confounding by drug indication or other unmeasured covariates. We cannot guarantee that the doses and protocols used in the GRACE registry match recommendations in practice guidelines, and some drugs may have been given off-label. In addition, adjustment for known variables does not allow for unknown, or inadequately measured, factors to be balanced between comparison groups. Our findings should be regarded as hypothesis-generating, requiring confirmation in further studies. Nonetheless, we believe that our data provide a “real-world” insight into the underuse of these drugs. Moreover, our findings provide insights into factors that are associated with the use or non-use of GP IIb/IIIa inhibitors; these factors could certainly be targeted in quality-improvement strategies. Appropriate caution must be exercised in the interpretation of these results, however, given the potential for residual confounding in these patients. These findings should in no way be used to overrule the clearly and consistently positive outcomes of the three RCTs that have established the value of this therapy.
Conclusions and clinical implications
A major purpose of studies such as GRACE is to provide practitioners with feedback on patient care and to identify areas in which greater consistency between evidence and clinical practice will improve outcomes. Although sufficient data exist to support the use of GP IIb/IIIa inhibitors in patients with ACS, our study has shown significant underuse of this therapy in a real-world population, irrespective of whether they were trial-eligible or not. Increasing the use of these therapies in appropriate patients with NSTE ACS should remain an important goal for physicians and institutions caring for higher-risk patients with ACS. Finally, our data suggest no evidence that trial-ineligible patients with NSTE ACS have a lesser treatment effect than eligible patients given GP IIb/IIIa inhibitor therapy. Future studies are needed to validate this finding.
Acknowledgments
We thank the physicians and nurses participating in GRACE. Further information about the project, along with the complete list of participants, can be found at www.outcomes.org/grace. The authors are grateful to Sophie Rushton-Smith, PhD, who provided editorial assistance and was funded by sanofi-aventis. GRACE is supported by an unrestricted educational grant from sanofi-aventis to the Center for Outcomes Research, University of Massachusetts Medical School. Sanofi-aventis had no involvement in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Appendix
Main inclusion and exclusion criteria
GRACE
Inclusion criteria
Patients with an admission diagnosis of unstable angina or non-ST-segment elevation myocardial infarction (UA/NSTEMI) according to the criteria described in the methods section of this paper.
Patients were defined according to each of the three clinical trial populations (PRISM, PRISM-PLUS, and PURSUIT), using the criteria specified in these trials, as detailed below.
GRACE variables were used to mimic the exclusion criteria of these trials, as detailed below.
The three GRACE trial-like populations constituted the trial-eligible cohort. The remaining patients were defined as the trial-ineligible cohort.
PRISM10
Inclusion criteria
Coronary artery disease as manifest by one of the following:
Electrocardiographic evidence of myocardial ischaemia in two contiguous leads during an episode of chest pain with new, persistent, or transient ST-segment depression of ⩾0.1 mV (0.08 seconds after the J point); new, persistent, or transient T-wave inversion; or transient ST-segment elevation (lasting <20 minutes) of ⩾0.1 mV.
Elevated cardiac enzyme levels consistent with the occurrence of non-Q-wave myocardial infarction.
History of myocardial infarction, percutaneous revascularisation more than 6 months earlier, coronary surgery >1 month earlier, positive exercise stress test or dipyridamole (or adenosine) nuclear stress test, or narrowing of ⩾50% of the luminal diameter of a major coronary artery on a previous arteriogram.
Patients were retrospectively classified as having non-Q-wave infarction if the creatine kinase level exceeded twice the normal value or if the level of the creatine kinase MB fraction (CK-MB) was above normal within the first 24 h, in the absence of a new infarction.
Exclusion criteria
Thrombolytic therapy within the previous 48 h.
Allergy to, or intolerance of, heparin.
Serum creatinine concentration >2.5 mg/dl (221 μmol/l).
Active bleeding disorder; history of gastrointestinal bleeding; haematuria; positive faecal occult-blood test; known coagulopathy; platelet disorder or history of thrombocytopenia.
Persistent systolic blood pressure >180 mm Hg, diastolic blood pressure >110 mm Hg, or both, at the time of enrolment.
History of haemorrhagic cerebrovascular disease or an active intracranial pathologic process; history of cerebrovascular disease or transient ischaemic attack within the previous year.
Major surgical procedure within the previous month; active peptic ulceration within the previous 3 months.
Invasive procedure within 14 days before enrolment that would substantially increase the risk of haemorrhage.
GRACE exclusion criteria for the PRISM Study
Thrombolytic therapy within the previous 48 h or while in hospital.
Serum creatinine level >2.5 mg/dl (221 μmol/l).
History of bleeding*.
Systolic blood pressure >180 mm Hg, diastolic blood pressure >110 mm Hg, or both, at the time of enrolment.
History of transient ischaemic attack†.
History of major surgical procedure within 2 weeks of the cardiac event.
PRISM PLUS11
Inclusion criteria
Patients with unstable angina or non-Q-wave myocardial infarction with prolonged anginal pain or repetitive episodes of angina at rest or during minimal exercise in the previous 12 h and new transient or persistent ST-T ischaemic changes on the electrocardiogram (ST-segment elevation or depression of ⩾0.1 mV, T-wave inversion of ⩾0.3 mV in ⩾3 limb leads or ⩾4 precordial leads excluding V1, or pseudonormalisation of ⩾0.1 mV), or elevation of plasma levels of creatine kinase and of the creatine kinase MB fraction (CK-MB).
Exclusion criteria
ST-segment elevation lasting >20 minutes.
Thrombolysis within previous 48 h.
Coronary angioplasty within the previous 6 months or bypass surgery within the previous month.
Angina caused by identifiable factors.
History of a platelet disorder or thrombocytopenia.
Active bleeding or a high risk of bleeding.
Stroke within the previous year.
Serum creatinine values >2.5 mg/dl (220 μmol/l).
Platelet count <150 000/mm3.
GRACE exclusion criteria for the PRISM PLUS Study
Thrombolytic therapy within the previous 48 h or in hospital.
Serum creatinine level >2.5 mg/dl (221 μmol/l).
History of bleeding*.
History of percutaneous coronary angioplasty or bypass surgery.
History of transient ischaemic attack†.
PURSUIT12
Inclusion criteria
Symptoms of ischaemic chest pain at rest, lasting ⩾10 minutes, within the previous 24 h.
Transient ST-segment elevation of >0.5 mm.
Transient or persistent ST-segment depression of >0.5 mm.
T-wave inversion of >1 mm within 12 h before or after chest pain.
Serum creatine kinase MB isoenzyme (CK-MB) concentration above the upper limit of normal for the hospital where it was evaluated.
Exclusion criteria
Persistent ST-segment elevation >1 mm.
Active bleeding or a history of bleeding diathesis, gastrointestinal or genitourinary bleeding within 30 days before enrolment.
Systolic blood pressure >200 mm Hg or diastolic blood pressure >110 mm Hg.
History of major surgery within the previous 6 weeks.
History of non-haemorrhagic stroke within the previous 30 days or any history of haemorrhagic stroke, renal failure, or pregnancy.
Planned administration of a platelet glycoprotein IIb/IIIa receptor inhibitor or thrombolytic drug.
Thrombolytic therapy within the previous 24 h.
GRACE exclusion criteria for the PURSUIT Study
Thrombolytic therapy prior to hospitalisation or in hospital.
Serum creatinine level >2.5 mg/dl (221 μmol/l).
History of bleeding*.
Systolic blood pressure >180 mm Hg, diastolic blood pressure >110 mm Hg, or both, at the time of enrolment.
History of transient ischaemic attack†.
History of major surgical procedure within 2 weeks of the cardiac event.
*Documented history of significant blood loss from any site, for example ulcer, genitourinary tract, not related to trauma, and requiring medical treatment.
†History of loss of neurological function caused by ischaemia that was abrupt in onset but with complete return of function within 24 h. Stroke or cerebral vascular accident defined as loss of neurological function caused by an ischaemic event with residual symptoms.
REFERENCES
Footnotes
Competing interests: O H Dabbous: none; F A Anderson Jr: sanofi-aventis; J M Gore: sanofi-aventis; K A Eagle: Biosite, Bristol-Myers Squibb, Cardiac Sciences, Blue Cross Blue Shield of Michigan, Hewlett Foundation, Mardigian Fund, Pfizer, sanofi-aventis, Varbedian fund, National Heart, Lung & Blood NIH; K A A Fox: British Heart Foundation, Medical Research Council, The Wellcome Trust, Aventis, Sanofi-Synthelabo, GlaxoSmithKline, Bristol-Myers Squibb; R H Mehta: none; R J Goldberg: none; G Agnelli: sanofi-aventis; P G Steg: Boehringer Ingelheim, Bristol-Myers Squibb, GlaxaSmithKline, Merck Sharp & Dohme, Novartis, Nycomed, Sanofi-Aventis, Sankyo, Servier, ZLB-Behring, Pfizer, Takeda.
Funding: GRACE is funded by an unrestricted educational grant from sanofi-aventis (Paris, France) to the Center for Outcomes Research, University of Massachusetts Medical School (Worcester, MA, USA).