Author + information
- Received June 21, 2004
- Accepted July 2, 2004
- Published online September 1, 2004.
- Harlan M. Krumholz, MD, SM⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Harlan M. Krumholz, Yale University School of Medicine, 333 Cedar Street, P.O. Box 208088, New Haven, Connecticut 06520-8088
Outcomes researchis applied clinical and population-based research that seeks to study and optimize the end results of health care—that is, the experience of the patient and the impact on society. All studies have outcomes, but this research is distinctive because of its focus on the end result of what health care seeks to achieve for patients and society. Outcomes research is not defined by a particular methodology, but rather by its objective. The research can focus on understanding current patterns of care, evaluating interventions to improve care, or generating insights to improve decision making or health-care delivery.
This type of research is growing in importance. As advances in cardiovascular research over the last several decades have augmented our ability to prevent, diagnose, and treat heart disease, questions are increasingly being raised about the dissemination of these strategies, their effectiveness in clinical practice, and their value to society. Clinicians and patients are asking how various strategies affect people's lives—not just how long they live but how well they live. Moreover, there is increasing concern about the current quality of health care and how it may be improved.
Health-care delivery and outcomes research have been increasingly represented in the published literature.This overview highlights information from prominent health-care delivery and outcomes research publications in 2003 and early 2004. It would not be possible to include all the noteworthy studies on this topic in a single review; thus, I have selected from among those with the most impact. The National Academy of Sciences Institute of Medicine (IOM) identified core goals for the future of American health care that include safety, effectiveness, equity, efficiency, timeliness, and patient-centeredness. This review is aligned with those themes, substituting the broader category of quality of care for safety and effectiveness.
Several studies have addressed the issue of disparities in our health-care system. Although equity in treatment should be a property of every health-care system, many experts have raised concerns about whether gender, race/ethnicity, and socioeconomic status influence the provision of care and the outcomes of patients. The findings of studies on this topic are occasionally inconsistent, indicating a need for further investigation. The highest impact studies addressed equity issues in the treatment of patients with ischemic heart disease and heart failure. They also examined the use of cardiac procedures.
Canada, with its universal health care, provides an opportunity to assess equity in a system in which cost is not a barrier to care. Two studies found that Canadian referral patterns are associated with socioeconomic status. The Socio-Economic Status and Acute Myocardial Infarction (SESAMI) study, a prospective, longitudinal, observational study of patients hospitalized with acute myocardial infarction (AMI) in Ontario, Canada, found that affluent patients were more likely to be referred for coronary angiography, cardiac rehabilitation, and to a cardiologist (1). Moreover, education was a stronger determinant of specialty cardiac care than was income. However, neither income nor education was significantly associated with adjusted risk of mortality at one year. In a study from Quebec, Canada, investigators also documented differences in cardiac referral rates for procedures after AMI by socioeconomic status (2). The findings raise questions about the underlying cause of the different referral patterns and whether they have any impact on long-term patient outcomes.
Several studies have addressed whether care and outcomes vary by race/ethnicity. The findings seem to depend on the population studied. Jha et al. (3) from the Heart and Estrogen/progestin Replacement Study (HERS), a randomized trial, investigated medical care and outcomes among black and white women with established coronary artery disease. Black women were less likely to receive aspirin after AMI (80% vs. 74%; p = 0.1), beta-blockers after AMI (34% vs. 32%), and statins if low-density lipoprotein (LDL) was >100 mg/dl (35% vs. 26%), but those with a history of heart failure were more likely to receive an angiotensin-converting enzyme (ACE) inhibitor (36% vs. 46%). Black women were also more likely to have poorer blood pressure and lipid control than were white women. Over more than four years of follow-up, the black women had an adjusted 60% greater risk of coronary heart disease (CHD) events. Interestingly, the differences in treatment did not explain the differences in outcome. After considering medication therapy and risk-factor control, the excess hazard was still 52% higher.
Investigators from the Atherosclerosis Risk In Communities (ARIC) study investigated racial/ethnic differences in mortality after AMI in 471 white and 171 black patients without a prior history of AMI. The crude mortality rate was higher for blacks compared with whites (21% vs. 14%). In contrast to the HERS, adjustment for demographics, vascular risk factors, socioeconomic factors, and severity of AMI and treatment eliminated the difference by race/ethnicity.
In heart failure, it appears that black patients and women have a survival advantage. In a large, nationally representative study, Rathore et al. (4) evaluated the care and outcomes of 29,732 black and white Medicare beneficiaries hospitalized with heart failure from 1998 to 1999. This study, based on detailed medical record review and conducted with data from the Centers for Medicare and Medicaid Services, found that blacks were not disadvantaged in two key processes of care: 1) the evaluation of left ventricular ejection fraction (67.8% black vs. 66.6% white), and 2) ACE inhibitor use (81.0% black vs. 73.8% white). Black patients had a lower risk of mortality at 30 days (6.3% vs. 10.7%) and 1 year (31.5% vs. 40.1%). After adjusting for potential confounders, blacks had a 22% lower risk of death at 30 days and a 7% lower risk at 1 year than whites, but their risk of readmission was 9% higher. The investigators concluded that the black patients did not appear to have worse heart failure care. However, their higher readmission rates raise concern.
Ghali et al. (5) from the Beta-blocker Evaluation of Survival Trial (BEST) investigated gender differences among 593 men and 2,115 women with advanced heart failure. After adjusting for major factors related to mortality, women had a significant survival advantage over men. Further analysis suggested that this advantage was restricted to those who had a nonischemic etiology of heart failure. The differences are provocative, but whether they are the result of biology, environment, health care, or chance has yet to be determined.
The use of cardiac procedures by race/ethnicity and gender continues to receive attention in the published reports. Studies consistently find differences, but fall short of providing insight into the mechanism. Gillum (6) examined trends during the 1990s in the use of coronary revascularization by gender and race/ethnicity. Bypass surgery and angioplasty increased over the decade in all groups. Differences by race, as a percentage, were greater in the older age groups and in women compared with men. The absolute differences per 100,000 in the population remained relatively constant. The underlying cause of these differences has yet to be demonstrated—and there remains debate about whether this pattern represents underuse of one, overuse of the other, or some combination.
The use of implantable cardioverter-defibrillators (ICDs) has appropriately drawn some scrutiny. Groeneveld et al. (7) evaluated the rates of cardiac procedures among 5,429 white and 519 black patients who experienced a cardiac arrest. The information was derived from Medicare billing data from 1990 to 1999 and used an admission diagnosis code of ventricular fibrillation or cardiac arrest. Although ICDs were equally effective in blacks and whites, blacks age 66 to 74 years had a 42% lower odds of receiving them. In addition, blacks in that age group had a 30% higher risk of death than whites.
There is increasing interest in assessing outcomes from patients' perspective, with particular interest in health status and quality of life. These studies have also sought to understand some of the key determinants of health-status outcomes. The advent of tools to measure health status has rapidly advanced this field of investigation.
Several studies have examined the impact of particular factors on health-status outcomes. Two studies provided insight about the association of age with the health status of patients undergoing cardiac surgery. Conaway et al. (8) evaluated the impact of age on recovery from bypass surgery among 690 consecutive patients, including 138 who were older than 75 years, admitted to the Mid America Heart Institute in Kansas City, Missouri.The improvement in angina frequency and physical functioning was similar in younger and older patients by six months.
Sedrakyan et al. (9) sought to determine the association of age with the change in quality of life after valve surgery. They prospectively enrolled 126 patients with aortic and 60 patients with mitral valve surgery who completed baseline and follow-up questionnaires. All age groups experienced improvements in health status, demonstrating that age is not an important predictor of the change in health status after surgery.
A study by Vaccarino et al. (10,11) compared health-status outcomes after bypass surgery in 804 men and 309 women who underwent surgery at Yale-New Haven Hospital. The mortality rate was similar in men and women. By eight weeks after surgery, the readmission rate in women was almost twice that in men (20.5% vs. 11.0%). Although men had no change in their physical function by eight weeks, women had, on average, a marked decline. By six months, although both men and women, on average, had improved physical functioning and mental health since surgery, the men had improved much more than the women. Women did worse than men in rates of readmission (32.6% vs. 21.2%), declining functional status (25.7% vs. 11.1%), and declining mental health (17.5% vs. 12.6%). This study demonstrates the importance of health-status measurement, because an exclusive focus on mortality would have given a very different impression of the comparison between men and women.
Rumsfeld et al. (12) from the Department of Veterans Affairs (VA) compared health-related quality of life for high-risk patients randomized either to percutaneous coronary intervention or bypass surgery. The main trial reported no mortality difference between the groups, and this analysis added that the type of revascularization was not associated with any of the health-status measurements. Significant predictors of worse six-month physical health status were chronic obstructive pulmonary disease, diabetes, and an elevated creatinine. This type of analysis is critically important in providing more information about the impact of these strategies on the lives of patients.
Psychosocial factors, such as depressive symptoms, are being increasingly recognized as having an important impact on patients with cardiac disease. Rumsfeld et al. (13) assessed the impact of depressive symptoms on changes in health status in heart-failure patients. A striking finding was that 30% of the 460 outpatients in their study had significant depressive symptoms at baseline. Over approximately six weeks of follow-up, the patients with depressive symptoms were much more likely to experience a worsening of their heart-failure–specific health status. In another study evaluating the association of depressive symptoms in veterans with angina, Rumsfeld et al. (14) found a high prevalence of depressive symptoms and a strong association with worse health status after a hospitalization for AMI or unstable angina. Strik et al. (15) examined depression and anxiety as predictors of cardiac events after an AMI. Studying 318 men over a periodof 3.4 years, they found that both anxiety and depression were strongly associated with the risk of events, and anxiety appeared to mediate some of the risk associated with depressive symptoms. Although past interventions have been disappointing, these studies support further attempts to modify psychosocial factors as a means of improving outcomes in cardiac patients.
Another important goal of the health-care system is timeliness and access to care. Few studies have focused on the timeliness of care, particularly the prevalence and impact of delays that are inherent in the health-care system. Reports on access have specifically addressed specialty care and its impact on outcomes.
Many patients admitted with cardiac conditions do not receive care from a cardiologist. Foody et al. (16), using a nationally representative sample of Medicare patients hospitalized with heart failure, found that only about half of the patients received treatment from a cardiologist, and only one quarter had a cardiologist as their attending physician. Older patients and women were less likely to have a cardiologist as an attending physician, but race was not a significant factor.
The value of specialty care, however, remains controversial. Several studies in the past year have suggested that cardiology care is associated with better treatment and outcomes. Ansari et al. (17), using data from the California Kaiser Permanente Medical Care Program, found that cardiology participation in the care of patients with heart failure was associated with a higher likelihood of having had an ejection fraction measurement and having been prescribed ACE inhibitors and beta-blockers. Cardiology care was also associated with a substantially lower adjusted risk of death or cardiovascular hospitalization (hazard ratio [HR] 0.65; 95% confidence interval [CI] 0.45 to 0.96). Indridason et al. (18), using a national Veterans Health Administration database, found that cardiology care, or the combination of cardiology care and general practitioner care, was associated with better survival than was general practitioner care alone. How this effect is mediated has yet to be determined.
Future studies in this domain may focus on the timeliness of care and the impact on patient outcomes of delays in diagnosis and treatment. Moreover, there is a need to understand how the organization of care affects these delays.
With increasing health-care costs, growing interest has arisen in studies that provide information about patterns of health-care spending and how best to allocate resources. More studies are needed that provide insight about how to balance each patient's needs with what society is willing to provide.
Perhaps the most important contributions in this area were offered by Fisher et al. (19,20) in a pair of classic studies examining the implications of regional variation in Medicare spending. In the first study, they investigated several conditions, including AMI, to determine whether regions with higher spending delivered better care. They found marked differences in health-care spending across the U. S. In the last six months of life, spending by hospital referral region ranged from $9,074 in the lowest quintile to $14,644 in the highest quintile. Regions that spent more money provided more care, but they did not have lower mortality rates. Regions with higher expenditures did not provide higher-quality care. Functional status and satisfaction with care also did not improve with higher spending. These studies fail to show any benefit associated with the higher-spending regions. The researchers report that substantial sums could be saved if the nation at largecould reach the levels achieved by the regions with the lowest decile of spending.
Many of the notable studies provided economic analyses of specific clinical strategies. These studies used cost-effectiveness analysis to compare alternative strategies.
The value of B-type natriuretic peptide (BNP) as a screening strategy has received attention. Heidenreich et al. (21) investigated the cost-effectiveness of screening patients with BNP to identify those with depressed left-ventricular systolic function. They reported that the test was economically attractive for many populations. Among men age 60 years with no evidence of heart failure, screening (followed by echocardiography if the BNP test was abnormal) cost $22,300 per quality-adjusted life-year gained compared with no screening. The BNP strategy required 44 people to be screened to detect one person with a depressed ejection fraction and 127 to gain one quality-adjusted life-year. For women age 60 years,it cost $77,700 to gain a quality-adjusted life-year. In sensitivity analyses, the investigators found that screening was attractive for patient groups with at least a 1% prevalence of left ventricular systolic dysfunction.
The success of public defibrillation has led investigators to consider the value of various strategies of employing this approach. Nichol et al. (22) evaluated the cost-effectiveness of defibrillation by targeted nontraditional responders in public settings (using casinos as the base case). Compared with emergency medical services, targeted nontraditional responders cost $56,700 per quality-adjusted life-year. Public-access defibrillation was not economically attractive if the incidence of arrest was rare, if the time to defibrillate by nontraditional responders was long, or if responders were compensated while being trained. Walker et al. (23) also addressed this issue using data from the Scottish Ambulance Service. They considered the placement of defibrillators in major airports, railway stations, and bus depots. The public defibrillator strategy cost £41,146per quality-adjustedlife-year saved. These reports suggest that the judicious placement of these devices in high-risk areas is an attractive investment from society's perspective.
With the expansion of options for the diagnosis and treatment of ischemic heart disease, many payers are seeking economic analyses to justify clinical strategies. Recently, several notable studies addressed this issue, but there is a clear need for more work in this area.
Bakhai et al. (24), from the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial, assessed the cost-effectiveness of stenting and glycoprotein (GP) IIb/IIIa inhibition for patients undergoing primary angioplasty. Stenting cost an additional $11,237 for each quality-adjusted life-year. In contrast, anticoagulation was an economically dominant strategy (lower long-term costs and greater quality-adjusted life expectancy) compared with GP IIb/IIIa inhibition. Thus, stenting seemed worth the investment, but the value of GP IIb/IIIa inhibition could not be justified.
Lamy et al. (25) investigated the cost implications of the Heart Outcomes Prevention Evaluation (HOPE) study, a randomized trial of ramipril versus placebo among patients who were at least 55 years of age and had high-risk characteristics, but who were not known to have heart failure or a depressed ejection fraction. They found that the difference in costs between the two treatment groups ranged from $−734 to$+512 in the U. S. and $−269 to $+497 in Canada. Thus, the increased cost of ramipril was offset by the reduction in cardiac events.
The measurement of fractional flow reserve is growing, but whether its cost is worth the benefit was studied last year by Fearon et al. (26), who compared the following strategies: deferring the decision for percutaneous coronary intervention to obtain a nuclear stress test; measuring fractional flow reserve to guide the decision for percutaneous coronary intervention; and stenting all intermediate lesions. Quality-adjusted survival was the same for all three strategies. The fractional flow reserve strategy saved $1,795 per patient compared with the nuclear testing strategy and $3,830 compared with the stenting strategy.
Efficiency is also measured in other ways. Kaul et al. (27) made use of data from several international randomized trials to assess trends in length of hospital stay and the frequency of early discharge of low-risk patients. Based on data from the Global Utilization of Streptokinase and Tissue plasminogen activator for Occluded coronary arteries (GUSTO-I), the Global Use of Strategies To Open occluded coronary arteries (GUSTO-III), and the Assessment of the Safety and Efficacy of a New Thrombolytic (ASSENT-2) trials, they documented marked international variation in length of hospital stay. Overall, <40% of patients eligible for early discharge were sent home on or before the fourth hospital day. In particular, the European countries had longer length of stay and fewer early discharges of low-risk patients. Based on the extended stay of low-risk patients, Kaul et al. (27) calculated the number of hospital days lost per 100 patients enrolled ranged from 91 in the U.S. to 738 days in Poland. This study highlights potential waste that exists in all systems.
The hospital utilization issue was also examined with theVA medical-care system, where several initiatives to reduce hospital care were implemented in the mid-1990s. For heart failure, adjusted bed-days per patient-year decreased from 18.7 in 1994 to 10.6 in 1998. The average length of hospital stay decreased from 11.0 to 8.4 days.For angina, adjusted bed-days per patient-year decreased from 8.7 in 1994 to 4.4 in 1998. The average length of stay decreased from 8.3 to 6.3 days.The one-year survival rates for these patients showed statistically significant annual improvement between 1994 and 1998. Among Medicare-eligible patients, the drop in the use of VA hospitals was not associated with an increase in the use of non-VA hospitals.
Some studies addressed burden of disease and priorities for the future. Ezzati and Lopez (28) estimated that smoking caused almost five million premature deaths in 2000. In industrialized countries, smoking accounts for 19% of total adult mortality; in developing countries it accounts for 9% of total adult mortality. In another study, Fontaine et al. (29) estimated that severe obesity may reduce life expectancy by as much as 22%. Peeters et al. (30), in a separate analysis, also estimated large decrements in life expectancy that are associated with obesity. These studies identify where preventive strategies could have the most benefit. Murray et al. (31) evaluated the value of many preventive interventions. They found that nonpersonal health interventions, such as government action to limit salt content of processed food, are a very economically attractive approach in reducing cardiovascular disease. Personal intervention was found to be best guided by the patients' absolute risk, rather than their measured blood pressure or cholesterol concentration.
Quality of care
Quality of care remains an important topic in health-care delivery. Some important studies provided perspective on the gaps in health-care performance, whereas others tested strategies to improve care.
One of the most important studies documented the gap between optimal care and what is being provided to patients around the country. McGlynn et al. (32) from RAND reported on the quality of care delivered to adults in the U. S. based on telephone interviews and medical record review. Guideline-recommended treatment was commonly not provided overall or specifically for cardiac conditions. The percentage of recommended care received was 68% for coronary artery disease, 65% for hypertension, and 64% for heart failure.
Several studies specifically examined preventive cardiology practices. Ford et al. (33) used data from the National Health And Nutrition Examination Survey (NHANES), conducted in 1999 and 2000, to evaluate the awareness, treatment, and control of hypercholesterolemia. Among participants who had an elevated cholesterol level or were using cholesterol-lowering medication, only 35% were aware that they had hypercholesterolemia, 12% were undergoing treatment, and 5.4% had a total cholesterol level under 200 mg/dl. Hajjar and Kotchen (34) performed a similar analysis with hypertension, and they found that between 1999 and 2000, 69% of the patients studied were aware of their hypertension, 58% were treated, and 31% were controlled.
On a more optimistic note, Jha et al. (35) investigated the quality of care within the VA health-care system during a time of system reengineering to improve treatment. From 1994 to 2000 there were marked improvements in treatment. Blood pressure control increased from 25% to 46%; AMI indicators also improved, including aspirin (89% to 98%) and beta-blockers (70% to 98%). In 2000, ACE inhibitor use was 90% for AMI and 93% for heart-failure patients. This study demonstrates the marked improvements that can be achieved.
The issue of appropriateness of procedures is difficult to approach, but Petersen et al. (36) sought to determine the prevalence and impact of underuse of angiography in the VA health-care system in comparison with the Medicare population. They found underuse, based on recommendations from clinical-practice guidelines, among the 1,665 veterans from 81 VA hospitals and the 19,305 Medicare patients from non-VA hospitals. Among those with strong indications for angiography, it was performed on 44% of the veterans and 51% of the Medicare patients. This difference persisted even after adjustment for potential confounders. Mortality was similar in both groups, which raises questions about the definition of appropriateness.
Several studies tested interventions to improve care. In an innovative trial, Ferguson et al. (37) from the Society of Thoracic Surgeons randomized 359 institutions to a continuous quality-improvement intervention to increase the use of preoperative beta-blockers and internal mammary artery grafting. The intervention group received a call to action to a physician leader, educational products, and periodic nationally benchmarked, site-specific data. The intervention was associated with a modestly significant increase in the beta-blocker group (a 7.3% increase in the intervention compared with 3.6% in the control) and a trend in the internal mammary artery group (increase of 8.7% in the intervention compared with 5.4% in the control group).
Ansari et al. (38) evaluated computer reminders and nurse-facilitators to improve the use of beta-blockers for patients with heart failure and depressed systolic function. They randomized 74 providers and found that the nurse-facilitator group performed much better, initiating or titrating beta-blockers in more than 60% of the subjects compared with one-third of those in the computer-reminder and control groups. The study is notable for showing the ineffectiveness of the computer reminders and the gaps in quality even with the involvement of nurse-facilitators.
Murchie et al. (39) evaluated the effect of nurse-based clinics in primary care on secondary prevention of cardiovascular disease. Over almost four years of follow-up, the intervention group fared better with regard to aspirin use, blood pressure and lipid management, the use of moderate exercise, and a low-fat diet. The intervention was associated with a significant reduction in total mortality (HR 0.75; 95% CI 0.58 to 0.98).
In contrast to the studies that found nurse intervention to be beneficial, Lichtman and colleagues discovered that an ongoing nurse-based telephonic educational intervention was not effective in improving compliance with lipid guidelines (40). They randomized 375 patients to the intervention and 381 patients to usual care. After one year, no difference was seen in the proportion of patients who reached target LDL levels.
Volume relationships with outcome were a contentious issue. Several organizations have used volume as a proxy for quality, and a number of groups tested that assumption. Hannan et al. (41) emphasized the continuing relationship between volume and outcome for bypass surgery in New York State. Peterson et al. (42), however, found that volume was only weakly related to outcome for bypass surgery across the country, and was not a good proxy for outcomes. Volume does not discriminate well between institutions, and having high volume does not guarantee good performance. Rathore et al. (43), using a different national database, reached the same conclusion. Birkmeyer et al. (44) further undermined the value of institution volume by suggesting that surgeon volume accounted for 100% of the effect for aortic-valve replacement, 57% for elective repair of an abdominal aortic aneurysm, and 49% for coronary artery bypass grafting.
Other studies investigated key success factors for improving quality in hospitals. Holmboe et al. (45) focused on the role of physicians as leaders of quality improvement teams. They highlighted personal commitment; professional credibility, quality improvement behaviors and skills, and institutional linkages as important to the success of these physicians. Bradley et al. (46) identified the important aspects of data feedback, determining that it is most effective if it is perceived as valid, if it is timely, if it can be compared with other institutions, and if it is provided in a regular and consistent manner.
The importance of outcomes research is growing. The profession is challenged to provide care that is safe, effective, efficient, equitable, timely, and patient-centered. There is growing recognition of the need for a broad base of evidence to guide clinical decision making in the real world and to guide policies that will enhance, not degrade, the care as viewed from our patients' perspective. Recently, more than 250 abstracts were presented at the 5th Annual American Heart Association/American College of Cardiology Scientific Forum on Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke. A Working Group of the National Heart, Lung, and Blood Institute will soon publish recommendations about research priorities in this field. In the past year many important studies in outcome research informed clinicians and policymakers. The expectation is that this area of investigation, so closely tied to clinical practice and public health, is just beginning a rapid growth phase that will coincide with the needs of both the profession and society at large for information about how best to deliver care.
- Received June 21, 2004.
- Accepted July 2, 2004.
- American College of Cardiology Foundation
- Jha A.K.,
- Varosy P.D.,
- Kanaya A.M,
- et al.
- Ghali J.K.,
- Krause-Steinrauf H.J.,
- Adams J,
- et al.
- Groeneveld P.W.,
- Heidenreich P.A.,
- Garber A.M.
- Conaway D.G.,
- House J.,
- Bandt K.,
- Hayden L.,
- Borkon A.M.,
- Spertus J.A.
- Sedrakyan A.,
- Vaccarino V.,
- Paltiel A.D,
- et al.
- Vaccarino V.,
- Lin Z.Q.,
- Kasl S.V,
- et al.
- Vaccarino V.,
- Lin Z.Q.,
- Kasl S.V,
- et al.
- Rumsfeld J.S.,
- Magid D.J.,
- Plomondon M.E,
- et al.
- Rumsfeld J.S.,
- Havranek E.,
- Masoudi F.A,
- et al.
- Strik J.J.,
- Denollet J.,
- Lousberg R.,
- Honig A.
- Ansari M.,
- Alexander M.,
- Tutar A.,
- Bello D.,
- Massie B.M.
- Heidenreich P.A.,
- Gubens M.A.,
- Fonarow G.C.,
- Konstam M.A.,
- Stevenson L.W.,
- Shekelle P.G.
- Nichol G.,
- Valenzuela T.,
- Roe D.,
- Clark L.,
- Huszti E.,
- Wells G.A.
- Walker A.,
- Sirel J.M.,
- Marsden A.K.,
- Cobbe S.M.,
- Pell J.P.
- Bakhai A.,
- Stone G.W.,
- Grines C.L,
- et al.
- Lamy A.,
- Yusuf S.,
- Pogue J.,
- Gafni A.
- Ford E.S.,
- Mokdad A.H.,
- Giles W.H.,
- Mensah G.A.
- Ansari M.,
- Shlipak M.G.,
- Heidenreich P.A,
- et al.
- Murchie P.,
- Campbell N.C.,
- Ritchie L.D.,
- Simpson J.A.,
- Thain J.
- Hannan E.L.,
- Wu C.,
- Ryan T.J,
- et al.
- Bradley E.H.,
- Holmboe E.S.,
- Mattera J.A.,
- Roumanis S.A.,
- Radford M.J.,
- Krumholz H.M.