Proliferation of Cardiac Technology in Canada
http://www.100md.com
《循环学杂志》
the Institute for Clinical Evaluative Sciences (D.A.A., A.N., T.A.S.)
the University of Toronto Clinical Epidemiology and Health Care Research Program, Sunnybrook & Women’s College site (D.A.A.)
the Divisions of Cardiology, Schulich Heart Centre (D.A.A.), Clinical Epidemiology Unit, Sunnybrook & Women’s College Health Sciences Centre
the Department of Health Policy, Management and Evaluation, University of Toronto (D.A.A., T.A.S.), Toronto, Canada.
Abstract
Background— Critics remain skeptical about the long-term sustainability of Medicare in Canada because of the proliferation of health technology and escalating expenditures. The objective of this study was to examine the temporal trends in the utilization and costs of cardiovascular technologies for the evaluation and/or management of patients with ischemic heart disease in Canada.
Methods and Results— This repeated cross-sectional population-based study of Ontario residents examined the temporal trends in the utilization and costs associated with echocardiography, stress (imaging and nonimaging) testing, coronary angiography, percutaneous coronary intervention (PCI), and bypass surgery between 1992 and 2001. Annual costs increased by nearly 2-fold over the 10-year study period and cumulatively accounted for more than $2.8 billion (Canadian) in expenditures. The proliferation in use of cardiac testing/interventions over time outstripped both demographic shifts and changes in the prevalence of coronary artery disease. Annual increases were widespread for all procedures (P<0.001) and ranged from 2% per year for nonimaging stress tests to 12% per year for PCI, after adjustment for age and sex. Generally, utilization rates were higher among the elderly, males, and those of low socioeconomic status. With few exceptions, annual increases in the utilization rates of cardiac tests and procedures were disproportionately higher among the elderly and women, but they were similar across socioeconomic subgroups. Increases in utilization appeared to reflect referrals toward higher-risk populations.
Conclusions— Although definitive conclusions about the appropriateness of temporal patterns cannot be ascertained, the proliferation of cardiac testing challenges the sustainability of Medicare in Canada, especially given uncertainty as to whether the accompanying incremental rise in total expenditures translates into significant outcome benefits in the population.
Key Words: tests cardiovascular diseases epidemiology
Introduction
Canada’s Health Act stipulates that Canadians are entitled to universal access for necessary medical services regardless of affordability because they are free to the patient at the point of service.1 Patient evaluation and management decisions are left to the discretion of healthcare providers, with no explicit decision-making criteria imposed by policymakers. Most physicians are reimbursed on a fee-for-service basis.
Editorial p 330
Clinical Perspective p 387
Although viewed by many as the country’s most valued and successful social program, critics remain skeptical about the long-term sustainability of Canadian Medicare.2–4 Healthcare expenditures in Canada continue to rise and in 2001 constituted 9.4% of Canada’s gross domestic product (GDP), fourth among Organisation for Economic Co-operation and Development (OECD) nations, ranking only behind Germany, Switzerland, and the United States.5 Although capacity for specialized medical services has been historically constrained compared with the United States,6 healthcare expenditures in Canada have risen significantly over the past decade.7–9 The extent to which such increases in healthcare expenditures are attributable to the proliferation of cardiac technology and reflect appropriate clinical decision-making behavior or "good value for money" remain unknown.
The objective of this study was to explore temporal trends in the utilization and costs of cardiovascular technologies for the evaluation or management of patients with ischemic heart disease. We focused on ischemic heart disease because it remains the leading cause of death in Canada,10 and because cardiovascular technology accounts for greater healthcare expenditures than technology for any other disease.11 We examined whether temporal increases in cardiovascular technologies corresponded to changes in disease burden and whether incremental growth rates were preferentially targeted to higher- versus lower-risk populations.12
Methods
Study Population
A repeated cross-sectional population-based study was undertaken, comprising the Ontario, Canada adult population aged 20 years and older.13 Ontario is the Canadian province with the largest population, consisting of 11.9 million people. Ontario is subdivided into 17 824 census dissemination areas (DAs) and 50 counties. The median population size per DA is 415 adults (aged 20 years and older) and ranges from 5 to 8365 adults; the corresponding population size per county ranges from 9462 to 1.98 million adults. Ontario has approximately 0.42 catheterization laboratories and 3.9 cardiologists per 100 000 residents, which is significantly lower than the cardiovascular service supply in the United States (ie, 1.2 catheterization suites and 6.4 cardiologists per 100 000).14,15
Cardiovascular services in Canada are provided without patient user fees. Ontario physicians receive fee-for-service payment administered by the Ministry’s Ontario Health Insurance Plan.
Data Sources
Patient healthcare utilization records were linked across multiple health administrative databases by a unique encrypted study number to protect patient confidentiality. Information about noninvasive testing was obtained through medical claims data from the Ontario Health Insurance Plan (OHIP).16 OHIP captures all outpatient claims throughout Ontario but does not capture services provided in alternate payment regions (5% of total claims for Ontario), nor does it identify echocardiograms, perfusion studies, and stress tests performed on hospitalized patients. Although physician claims data can be used to ascertain the performance of invasive cardiac procedures, the fee code structure resulted in the double counting of coronary angiography procedures compared with Ministry of Health and Long-Term Care annual volumes, as reported independently by the Cardiac Care Network of Ontario (www.ccn.on.ca). Accordingly, invasive cardiac services were identified with the Canadian Institutes for Health Information (CIHI) database, a data source well used and validated in other studies by our group.17,18 CIHI was also used to determine patients recently discharged from the hospital with acute myocardial infarction (AMI; see below).
Neighborhood median household income, our socioeconomic indicator,19 was derived at the DA level with official 1996 and 2001 Census data and linked to patient residence with the postal code conversion file.20 Census data were also used to derive DA age- and sex-specific population estimates. The Registered Persons Data Base was used to derive patient demographic information and mortality. The Ontario Case Costing Initiative provided direct expenditures for invasive cardiac procedures, including associated costs for hospitalization,21 whereas the Ontario Medical Association Economic Committee provided costing data for noninvasive testing; costs were indexed to 2000 to 2001 data.16 Physician remuneration for outpatient noninvasive testing includes professional (ie, diagnostic interpretation) and technical (ie, operating costs) fees, with cumulative reimbursement totals of $99.81 for graded exercise stress tests, $229.63 for 2D echocardiography, and $335.83 for nuclear perfusion imaging in 2001. The study received ethics approval at Sunnybrook and Women’s College Health Sciences Centre.
Cardiovascular Technologies
Echocardiography, graded exercise treadmill testing (hereafter termed "stress testing"), nuclear perfusion imaging, coronary angiography, percutaneous coronary intervention (PCI), and CABG surgery were measured on individual patients and served as the main outcome variables. Given differences in data availability between OHIP and CIHI data, annual rates of noninvasive tests were examined between January 1st and December 31st of each calendar year, whereas invasive procedures were examined between April 1st and March 31st (but annual rates were attributed to the calendar year that represented December 31st).
Noninvasive test utilization was identified with the professional component (ie, diagnostic interpretation) of physician claims data. We excluded multiple tests per day to the same patient for any diagnostic service. Codes for echocardiography included 1D or 2D studies with or without Doppler examination (G561, G562, G567, G568, G571, G572, and G575). Stress tests were identified by G319, whereas perfusion imaging tests included exercise or dipyridamole myocardial perfusion imaging with or without SPECT using either sestamibi or thallium as its radiotracer (J607, J608, J609, J666, J807, J808, J809, and J866). Such codes have been used previously to examine noninvasive cardiac testing in Ontario.16 Given that perfusion imaging tests may be conducted over 1 or more consecutive days, we applied a 2-day window on either side of the date of a nuclear imaging claim to avoid duplicate counting. A similar 2-day window was applied to stress testing to differentiate isolated stress tests from those with concomitant perfusion.16 We examined 2 surrogates of testing yield, the ratio of angiograms per noninvasive stress or imaging test and the ratio of revascularization procedures per angiogram.
Prevalence of Coronary Artery Disease
Annual age- and sex-adjusted AMI hospitalization rates were used as a surrogate for the underlying prevalence of coronary artery disease. Previous evidence has demonstrated that the prevalence of AMI hospitalization is strongly correlated with population-based illness rates22 and with variations in the burden of self-reported risk factors (R2=0.56) and hospitalizations for chest pain syndromes (R2=0.51), angina (R2=0.64), and congestive heart failure (R2=0.56) across 16 District Health Council regions throughout Ontario.23,24 AMI admissions were defined with the most responsible diagnostic code (International Classification of Diseases–9th Revision code 410) and a minimum length of stay of 3 days unless death occurred.
Subgroup Analyses
We examined the utilization rates of cardiac testing within 3 prespecified sociodemographic subgroups: age (65 versus <65 years), gender, and socioeconomic status (median household income above versus below $50 000 per year). These subgroups were chosen because each has previously been demonstrated to be an important determinant of cardiac testing, and evidence that suggests that temporal changes in cardiac service intensity among selected clinical subgroups (eg, AMI) has preferentially favored elderly and higher-income patients.12
To explore how utilization patterns have changed over time among selected clinical subgroups, we examined those hospitalized for AMI within the previous 3 months and those hospitalized within the last 6 months of life, for 2 reasons. First, both represented subgroups of patients with greater illness severity than the general population,25 yet they differ in the characteristics of illness severity they represent; recent AMI hospitalization serves as a better surrogate for cardiac severity, whereas end-of-life status better serves to identify patients with chronic or palliative disease. Second, both subgroups account for significant overall healthcare expenditures.26–28 In addition, some may argue that disproportional increases in cardiac testing in the end of life may represent a less appropriate or efficient use of cardiac resources.29,30
Statistical Analysis
Population-based rates of cardiac tests and procedures were computed by year, and according to 5-year age group, gender, and income groups, by dividing the number of procedures by the corresponding adult population of Ontario. They were directly standardized according to age and gender with the Ontario population.
For each cardiac test and procedure, we used Poisson regression to estimate the annual rate of increase over the 10-year period, adjusted for age and gender.31 The unit of analysis was the age group, gender, and DA stratum, both for computational efficiency and because we did not distinguish among multiple tests to the same person. The dependent variable was the stratum-specific annual count of tests/procedures, with an offset to incorporate the stratum-specific population. The annual trend over the 10-year study period was estimated with a continuous term for calendar year, adjusted for age group, gender, and their interactions. Point estimates and CIs for the relative annual rates were obtained by exponentiation of the regression parameter that corresponded to the calendar-year term. Trend models were also stratified by age group (age 20 to 64 years or 65 years), gender, socioeconomic status (median household income <$50 000; $50 000), recent AMI, and end-of-life subgroups. Annual trends were assessed for similarity across subgroups with a test for interaction between subgroup and calendar year. We used longitudinal data analysis methods (generalized estimating equation) for clustered count data to account for the correlation among outcomes over time.32 SAS version 8.2 statistical software using the procedure GENMOD was used for modeling. All tests were performed at the 5% level of significance and were 2-sided.
Results
Overall Temporal Trends
Figure 1 illustrates unadjusted relative changes in the use of cardiac technologies compared with 1992. Increases in cardiac testing rates outstripped both demographic shifts and prevalence of AMI, as measured through AMI hospitalizations, over the same time period (Figure 1). Table 1 illustrates that rates for all technologies demonstrated significant exponential growth over the study period. Age- and sex-adjusted population rates were highest for echocardiography and lowest for PCI (during earlier years) and coronary artery bypass surgery (during later years). The yield of stress tests to catheterization, calculated as the ratio of their respective counts, steadily rose from 11.6% in 1992 to 16.7% in 2001, whereas the ratio of revascularization to angiography remained relatively stable over time (46%).
Overall Expenditures
Table 2 illustrates the total expenditures associated with cardiac technology over the past 10 years. Annual expenditures for cardiac technology in the evaluation or management of ischemic heart disease increased by nearly 2-fold over the 10-year study period and exceeded (CAN) $2.8 billion over the 10-year period. They were highest for coronary artery bypass surgery (CAN $1.2 billion), followed by echocardiography (CAN$498 million) and coronary angiography (CAN $402 million). Increases in costs mirrored the temporal changes in the growth of cardiac technologies (Table 2).
Subgroup Analyses
Table 3 illustrates absolute and relative trends over time for cardiac technology according to age, gender, and socioeconomic subgroups. The elderly, males (except echocardiography), and low socioeconomic status groups generally had higher testing and procedure rates; however, annual increases in cardiac testing and procedure rates were consistently higher among the elderly and, with the exception of echocardiography, were significantly greater for women. With 2 exceptions, annual increases in the use of cardiac technology were similar across socioeconomic groups; however, echocardiography and coronary angiography rates increased disproportionately among higher-income groups.
The proportion of noninvasive tests to patients within 3 months of AMI steadily declined over time, whereas the converse was true for invasive procedures (Figure 2). The proportion of cardiac technology use to patients within their last 6 months of life was under 5%, regardless of year, test, or procedure, and with 1 exception (PCI), it steadily declined over time (Figure 2).
Discussion
The long-term sustainability of Medicare in Canada continues to receive intense debate among public, provider, and policy stakeholders, especially in view of the steady growth in healthcare expenditures throughout the past decade.7–9 As demonstrated in the present study, the proliferation of cardiac invasive and noninvasive technology accounts for a significant component of rising health-related expenditures. Specifically, in 2001, approximately CAN $400 000 000 was spent on the evaluation and management of ischemic heart disease in Ontario alone, a cost that doubled over 10 years.
Although the proliferation of cardiac technology has been widespread, temporal growth has been most marked for coronary angiography and PCI, a phenomenon only partially attributable to the more intensive use of such interventions among AMI populations. Although the proliferation of coronary angiography has outstripped corresponding increases in nonimaging and imaging stress testing, the use of noninvasive diagnostic services has increased in magnitudes ranging from 2% to 10% per year, after adjustment for age, gender, and socioeconomic status. Notwithstanding such temporal patterns, the population rates for perfusion imaging stress tests, coronary angiography, PCI, and coronary artery bypass surgery services in Canada in 2001 lag significantly behind corresponding 1992 population rates in the United States.14,33
Our findings are consistent with many studies that have also demonstrated accelerated growth in the use of cardiac technology.12,34–39 For example, our US colleagues have demonstrated similar dramatic relative increases in cardiac testing and treatment over time, although with absolute rates in 1993 that exceed those in Canada from 2001.40 Moreover, intercountry differences in the absolute rates of invasive cardiac testing appear to correlate with their corresponding variations in catheterization laboratory supply (eg, catheterization laboratory supply of 0.42 versus 1.2 per 100 000 in Ontario versus the United States, respectively). Temporal changes in growth are likely attributable to continued proliferation in cardiac technology and specialty physician supply, as well as to changes in physician referral behaviors.15,36 In short, the magnitude of annual increases in cardiac service in Ontario appears similar to the corresponding magnitudes observed in other healthcare systems.
The present study was not designed to address appropriateness; however, some may interpret our findings as evidence that referral patterns for cardiac testing in Ontario have been inappropriate, given that cardiac service utilization has been shown to be driven by resource supply and availability rather than by clear indicators of need.17,22,25,41 Indeed, the proliferation of cardiac technology use observed in the present study markedly outstripped demographic shifts and AMI disease prevalence in the population.
Others view such temporal trends as evidence of appropriate care. For example, our results demonstrated that annual increases in cardiac testing rates were significantly greater among higher-risk subgroups such as the elderly and women, which suggests that physician referral propensity appeared to reflect appropriately the underlying risk profile of their patients. Moreover, by virtue of the low testing rates among patients at or near the last 6 months of life, increased referral among elderly high-risk populations still appeared somewhat selective and geared to those who might benefit most and was possibly aligned with the expectations and preferences of the patients whom these physicians serve.29 At minimum, the present study provided no evidence that the accelerated growth of cardiac technologies was inappropriate.
Even if we were to assume that the prevalence of inappropriate utilization for cardiac testing was low, the tremendous costs associated with the rise in cardiac technology would necessitate marked survival and quality-of-life benefits for temporal utilization patterns to meet cost-effective thresholds. For example, in the present study, the $2.87 billion expenditure investment for cardiac technology would have needed to translate into 57 336 lives saved over 10 years for temporal utilization patterns to have met traditional cost-benefit benchmarks (eg, $50 000 per life-year gained), a survival yield which, if true, would exceed those achievable from established life-saving therapies, such as statins, when used in secondary prevention populations.42–44
Equally important is the absence of incentives, disincentives, or safeguards that might otherwise curtail the continued proliferation of cardiac technologies, especially when one considers that increases in resource expenditures do not appear to have been offset by potential cost savings associated with fewer cardiac hospitalizations or medications. For example, expenditures for cardiovascular pharmacotherapy in Canada have risen 94% between 1996 and 2001, an increase of $1.6 billion in annual drug expenditures, attributable to a 61% rise in the annual number of cardiovascular drug prescriptions.45 If left unchecked, such temporal utilization patterns raise serious concerns about the sustainability of Medicare in Canada, especially when one considers their associated annual expenditures. As a corollary, the disproportionate expenditures allocated to cardiac technology compared with cardiac rehabilitation46,47 underscore an imbalance in focus between "technology" and "prevention initiatives" in Canada.
There are several limitations associated with the present study. First, our analysis was cross-sectional in design and was conducted without available clinical detail or information about referral indication. We also used human ecology rather than individual indicators of socioeconomic status, which may have led to misclassification of individual-level socioeconomic status. Nonetheless, available evidence has demonstrated that such ecological indicators are valid proxies for socioeconomic status.19,48 Although the inclusion of more clinical detail and/or individual socioeconomic indicators might have better explained changes in referral patterns over time, the overall temporal patterns observed in the present study would not have changed. Second, limitations in data availability precluded us from identifying any noninvasive testing conducted on hospitalized patients, which would underestimate the temporal growth in noninvasive testing. Third, our study did not adjust for inflationary pressures; however, inflation rates were marginal throughout the study period and would have accounted for only a small proportion of the increase in cardiac technology expenditures. Finally, our data were limited to Ontario. It is possible that temporal trends in the use of cardiac services may have differed in other Canadian jurisdictions. Nonetheless, Ontario represents 40% of the Canadian population, and temporal growth in cardiac technology has occurred in similar magnitudes in other Canadian jurisdictions.49,50
In conclusion, the present study illustrates continued annual acceleration in cardiac technology utilization. Although changes in utilization patterns were generally geared more preferentially to higher-risk subgroups than to lower-risk subgroups, definitive conclusions about the appropriateness of such proliferation cannot be drawn. Notwithstanding this, the high costs associated with the rise in cardiac technology would necessitate marked benefits in outcomes to justify the cost-effectiveness of procedural growth rates in the population. Given the absence of safeguards that prevent or attenuate the proliferation in cardiac technology use, continued growth in the use of invasive and noninvasive cardiac services will challenge the sustainability of Medicare in Canada.
Acknowledgments
This work was supported in part by a grant by the Heart and Stroke Foundation of Canada. The Institute for Clinical Evaluative Sciences is supported in part by a grant from the Ontario Ministry of Health. Dr Alter is a New Investigator at the Canadian Institutes of Health Research. The results, conclusions, and opinions are those of the authors, and no endorsement by the Ministry of Health and Long Term Care or the Institute for Clinical Evaluative Sciences.
Disclosures
None.
References
Canada Health Act. C-6, Section 10. 1984.
Davidson A. Romanow, Kirby and Courchene: Canada’s health system: a moral or a business enterprise Healthc Manage Forum. 2003; 16: 26–28.
Globerman S, Hodges H, Vining A. Canadian and the United States’ health care systems performance and governance: elements of convergence. Appl Health Econ Health Policy. 2002; 1: 75–88.
Guyatt G, Yalnizyan A, Devereaux PJ. Solving the public health care sustainability puzzle. CMAJ. 2002; 167: 36–38.
Organisation for Economic Co-operation and Development Web site. Available at: http://www.oecd.org. Accessed May 4, 2005.
Cardiac Care Network of Ontario, Target Setting Working Group. Final Report and Recommendations. Available at: http://www.ccn.on.ca/pdfs/IC-2001-06.pdf. Accessed January 13, 2006.
Buske L. Health care spending rises 4.6% in 2003. CMAJ. 2004; 170: 325.
Buske L. Health care bill reaches 3245 dollars per Canadian. CMAJ. 2003; 169: 595.
Seshamani M, Gray A. Health care expenditures and ageing: an international comparison. Appl Health Econ Health Policy. 2003; 2: 9–16.
Brophy JM. The epidemiology of acute myocardial infarction and ischemic heart disease in Canada: data from 1976 to 1991. Can J Cardiol. 1997; 13: 474–478.
Chan B, Young W. Burden of cardiac disease. In: Cardiovascular Health and Services in Ontario: An ICES Atlas. Naylor CD SP, ed. Toronto, Canada: Institute for Clinical Evaluative Sciences; 2000.
Khaykin Y, Austin PC, Tu JV, Alter DA. Utilisation of coronary angiography after acute myocardial infarction in Ontario over time: have referral patterns changed Heart. 2002; 88: 460–466.
Feldman HA, McKinlay SM. Cohort versus cross-sectional design in large field trials: precision, sample size, and a unifying model. Stat Med. 1994; 13: 61–78.
Wennberg DE, Birkmeyer JD. Coronary artery disease. In: The Dartmouth Atlas of Cardiovascular Health Care. Hanover, NH: Center for the Evaluative Clinical Sciences, Dartmouth Medical School; 1999: 51–96.
Goodman DC. Twenty-year trends in regional variations in the U.S. physician workforce. Health Aff (Millwood). 2004; Suppl Web Exclusive: VAR90–VAR97.
Alter DA, Przybysz R, Iron K. Non-invasive Cardiac Testing in Ontario. Toronto, Canada: Institute for Clinical Evaluative Sciences; 2004. Available at: http://www.ices.on.ca/file/Cardiac_ Testing%20Report_Oct_2004.pdf. Accessed November 2, 2004.
Alter DA, Naylor CD, Austin PC, Tu JV. Long-term MI outcomes at hospitals with or without on-site revascularization. JAMA. 2001; 285: 2101–2108.
Alter DA, Iron K, Austin PC, Naylor CD. Socioeconomic status, service patterns, and perceptions of care among survivors of acute myocardial infarction in Canada. JAMA. 2004; 291: 1100–1107.
Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med. 1999; 341: 1359–1367.
Wilkins R. Use of postal codes and addresses in the analysis of health data. Health Rep. 1993; 5: 157–177.
Ontario Case Costing Initiative Web site. Available at: http://www.OCCP.com. Accessed November 2, 2004.
Wennberg JE, Barnes BA, Zubkoff M. Professional uncertainty and the problem of supplier-induced demand. Soc Sci Med. 1982; 16: 811–824.
Jaglal SB, Bondy SJ, Slaughter PM. Risk factors for cardiovascular disease. In: Naylor CD, Slaughter PM, eds. Cardiovascular Health and Services in Ontario: An ICES Atlas. Toronto, Canada: Institute for Clinical Evaluative Sciences; 1999: 63–82.
Basinski ASH. Hospitalization for cardiovascular medical diagnoses. In: Naylor CD, Slaughter PM, eds. Cardiovascular Health and Services in Ontario: An ICES Atlas. Toronto, Canada: Institute for Clinical Evaluative Sciences; 1999: 15–50.
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending: part 2: health outcomes and satisfaction with care. Ann Intern Med. 2003; 138: 288–298.
Hoover DR, Crystal S, Kumar R, Sambamoorthi U, Cantor JC. Medical expenditures during the last year of life: findings from the 1992–1996 Medicare current beneficiary survey. Health Serv Res. 2002; 37: 1625–1642.
Levinsky NG, Yu W, Ash A, Moskowitz M, Gazelle G, Saynina O, Emanuel EJ. Influence of age on Medicare expenditures and medical care in the last year of life. JAMA. 2001; 286: 1349–1355.
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending: part 1: the content, quality, and accessibility of care. Ann Intern Med. 2003; 138: 273–287.
Teno JM, Fisher ES, Hamel MB, Coppola K, Dawson NV. Medical care inconsistent with patients’ treatment goals: association with 1-year Medicare resource use and survival. J Am Geriatr Soc. 2002; 50: 496–500.
Covinsky KE, Eng C, Lui LY, Sands LP, Yaffe K. The last 2 years of life: functional trajectories of frail older people. J Am Geriatr Soc. 2003; 51: 492–498.
McCullagh P, Nelder JA. Generalized Linear Models. New York, NY: Chapman & Hall; 1989.
Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986; 42: 121–130.
Lucas FL, Wennberg DE, Malenka DJ. Variation in the use of echocardiography. Eff Clin Pract. 1999; 2: 71–75.
Chan B, Cox JL, Anderson G. Trends in the utilization of noninvasive cardiac diagnostic tests in Ontario from fiscal year 1989/90 to 1992/93. Can J Cardiol. 1996; 12: 237–248.
Gore JM, Goldberg RJ, Alpert JS, Dalen JE. The increased use of diagnostic procedures in patients with acute myocardial infarction: a community-wide perspective. Arch Intern Med. 1987; 147: 1729–1732.
Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996–1998 practice patterns. Radiology. 2002; 222: 144–148.
Togni M, Balmer F, Pfiffner D, Maier W, Zeiher AM, Meier B. Percutaneous coronary interventions in Europe 1992–2001. Eur Heart J. 2004; 25: 1208–1213.
Pate GE, Humphries KH, Izadnegahdar M, Gao M, Kiely M, Carere RG. Population rates of invasive cardiac procedures in British Columbia, 1995 to 2001. Can J Cardiol. 2004; 20: 712–716.
Gillespie ND, Struthers AD, Pringle SD. Changing echocardiography request patterns between 1988 and 1993. Health Bull (Edinb). 1996; 54: 395–401.
Lucas FL, DeLorenzo MA, Siewers AE, Wennberg DE. Temporal trends in the utilization of diagnostic testing and treatments for cardiovascular disease in the United States, 1993–2001. Circulation. 2006; 113: 374–379.
Wennberg D, Dickens J Jr, Soule D, Kellett M Jr, Malenka D, Robb J, Ryan T Jr, Bradley W, Vaitkus P, Hearne M, O’Connor G, Hillman R. The relationship between the supply of cardiac catheterization laboratories, cardiologists and the use of invasive cardiac procedures in northern New England. J Health Serv Res Policy. 1997; 2: 75–80.
Jackevicius CA, Mamdani M, Tu JV. Adherence with statin therapy in elderly patients with and without acute coronary syndromes. JAMA. 2002; 288: 462–467.
Ko DT, Mamdani M, Alter DA. Lipid-lowering therapy with statins in high-risk elderly patients: the treatment-risk paradox. JAMA. 2004; 291: 1864–1870.
Ross SD, Allen IE, Connelly JE, Korenblat BM, Smith ME, Bishop D, Luo D. Clinical outcomes in statin treatment trials: a meta-analysis. Arch Intern Med. 1999; 159: 1793–1802.
Jackevicius CA, Tu K, Filate WA, Brien SE, Tu JV. Trends in cardiovascular drug utilization and drug expenditures in Canada between 1996 and 2001. Can J Cardiol. 2003; 19: 1359–1366.
Stone J. Cardiac rehabilitation: cost and care effective. Can J Cardiol. 2004; 20: 1256–1257.
Taylor RS, Brown A, Ebrahim S, Jolliffe J, Noorani H, Rees K, Skidmore B, Stone JA, Thompson DR, Oldridge N. Exercise-based rehabilitation for patients with coronary heart disease: systematic review and meta-analysis of randomized controlled trials. Am J Med. 2004; 116: 682–692.
Finkelstein MM. Ecologic proxies for household income: how well do they work for the analysis of health and health care utilization Can J Public Health. 2004; 95: 90–94.
Pilote L, Lavoie F, Ho V, Eisenberg MJ. Changes in the treatment and outcomes of acute myocardial infarction in Quebec, 1988–1995. CMAJ. 2000; 163: 31–36.
Quan H, Cujec B, Jin Y, Johnson D. Acute myocardial infarction in Alberta: temporal changes in outcomes, 1994 to 1999. Can J Cardiol. 2004; 20: 213–219.(David A. Alter, MD, PhD, FRCPC; Therese )
the University of Toronto Clinical Epidemiology and Health Care Research Program, Sunnybrook & Women’s College site (D.A.A.)
the Divisions of Cardiology, Schulich Heart Centre (D.A.A.), Clinical Epidemiology Unit, Sunnybrook & Women’s College Health Sciences Centre
the Department of Health Policy, Management and Evaluation, University of Toronto (D.A.A., T.A.S.), Toronto, Canada.
Abstract
Background— Critics remain skeptical about the long-term sustainability of Medicare in Canada because of the proliferation of health technology and escalating expenditures. The objective of this study was to examine the temporal trends in the utilization and costs of cardiovascular technologies for the evaluation and/or management of patients with ischemic heart disease in Canada.
Methods and Results— This repeated cross-sectional population-based study of Ontario residents examined the temporal trends in the utilization and costs associated with echocardiography, stress (imaging and nonimaging) testing, coronary angiography, percutaneous coronary intervention (PCI), and bypass surgery between 1992 and 2001. Annual costs increased by nearly 2-fold over the 10-year study period and cumulatively accounted for more than $2.8 billion (Canadian) in expenditures. The proliferation in use of cardiac testing/interventions over time outstripped both demographic shifts and changes in the prevalence of coronary artery disease. Annual increases were widespread for all procedures (P<0.001) and ranged from 2% per year for nonimaging stress tests to 12% per year for PCI, after adjustment for age and sex. Generally, utilization rates were higher among the elderly, males, and those of low socioeconomic status. With few exceptions, annual increases in the utilization rates of cardiac tests and procedures were disproportionately higher among the elderly and women, but they were similar across socioeconomic subgroups. Increases in utilization appeared to reflect referrals toward higher-risk populations.
Conclusions— Although definitive conclusions about the appropriateness of temporal patterns cannot be ascertained, the proliferation of cardiac testing challenges the sustainability of Medicare in Canada, especially given uncertainty as to whether the accompanying incremental rise in total expenditures translates into significant outcome benefits in the population.
Key Words: tests cardiovascular diseases epidemiology
Introduction
Canada’s Health Act stipulates that Canadians are entitled to universal access for necessary medical services regardless of affordability because they are free to the patient at the point of service.1 Patient evaluation and management decisions are left to the discretion of healthcare providers, with no explicit decision-making criteria imposed by policymakers. Most physicians are reimbursed on a fee-for-service basis.
Editorial p 330
Clinical Perspective p 387
Although viewed by many as the country’s most valued and successful social program, critics remain skeptical about the long-term sustainability of Canadian Medicare.2–4 Healthcare expenditures in Canada continue to rise and in 2001 constituted 9.4% of Canada’s gross domestic product (GDP), fourth among Organisation for Economic Co-operation and Development (OECD) nations, ranking only behind Germany, Switzerland, and the United States.5 Although capacity for specialized medical services has been historically constrained compared with the United States,6 healthcare expenditures in Canada have risen significantly over the past decade.7–9 The extent to which such increases in healthcare expenditures are attributable to the proliferation of cardiac technology and reflect appropriate clinical decision-making behavior or "good value for money" remain unknown.
The objective of this study was to explore temporal trends in the utilization and costs of cardiovascular technologies for the evaluation or management of patients with ischemic heart disease. We focused on ischemic heart disease because it remains the leading cause of death in Canada,10 and because cardiovascular technology accounts for greater healthcare expenditures than technology for any other disease.11 We examined whether temporal increases in cardiovascular technologies corresponded to changes in disease burden and whether incremental growth rates were preferentially targeted to higher- versus lower-risk populations.12
Methods
Study Population
A repeated cross-sectional population-based study was undertaken, comprising the Ontario, Canada adult population aged 20 years and older.13 Ontario is the Canadian province with the largest population, consisting of 11.9 million people. Ontario is subdivided into 17 824 census dissemination areas (DAs) and 50 counties. The median population size per DA is 415 adults (aged 20 years and older) and ranges from 5 to 8365 adults; the corresponding population size per county ranges from 9462 to 1.98 million adults. Ontario has approximately 0.42 catheterization laboratories and 3.9 cardiologists per 100 000 residents, which is significantly lower than the cardiovascular service supply in the United States (ie, 1.2 catheterization suites and 6.4 cardiologists per 100 000).14,15
Cardiovascular services in Canada are provided without patient user fees. Ontario physicians receive fee-for-service payment administered by the Ministry’s Ontario Health Insurance Plan.
Data Sources
Patient healthcare utilization records were linked across multiple health administrative databases by a unique encrypted study number to protect patient confidentiality. Information about noninvasive testing was obtained through medical claims data from the Ontario Health Insurance Plan (OHIP).16 OHIP captures all outpatient claims throughout Ontario but does not capture services provided in alternate payment regions (5% of total claims for Ontario), nor does it identify echocardiograms, perfusion studies, and stress tests performed on hospitalized patients. Although physician claims data can be used to ascertain the performance of invasive cardiac procedures, the fee code structure resulted in the double counting of coronary angiography procedures compared with Ministry of Health and Long-Term Care annual volumes, as reported independently by the Cardiac Care Network of Ontario (www.ccn.on.ca). Accordingly, invasive cardiac services were identified with the Canadian Institutes for Health Information (CIHI) database, a data source well used and validated in other studies by our group.17,18 CIHI was also used to determine patients recently discharged from the hospital with acute myocardial infarction (AMI; see below).
Neighborhood median household income, our socioeconomic indicator,19 was derived at the DA level with official 1996 and 2001 Census data and linked to patient residence with the postal code conversion file.20 Census data were also used to derive DA age- and sex-specific population estimates. The Registered Persons Data Base was used to derive patient demographic information and mortality. The Ontario Case Costing Initiative provided direct expenditures for invasive cardiac procedures, including associated costs for hospitalization,21 whereas the Ontario Medical Association Economic Committee provided costing data for noninvasive testing; costs were indexed to 2000 to 2001 data.16 Physician remuneration for outpatient noninvasive testing includes professional (ie, diagnostic interpretation) and technical (ie, operating costs) fees, with cumulative reimbursement totals of $99.81 for graded exercise stress tests, $229.63 for 2D echocardiography, and $335.83 for nuclear perfusion imaging in 2001. The study received ethics approval at Sunnybrook and Women’s College Health Sciences Centre.
Cardiovascular Technologies
Echocardiography, graded exercise treadmill testing (hereafter termed "stress testing"), nuclear perfusion imaging, coronary angiography, percutaneous coronary intervention (PCI), and CABG surgery were measured on individual patients and served as the main outcome variables. Given differences in data availability between OHIP and CIHI data, annual rates of noninvasive tests were examined between January 1st and December 31st of each calendar year, whereas invasive procedures were examined between April 1st and March 31st (but annual rates were attributed to the calendar year that represented December 31st).
Noninvasive test utilization was identified with the professional component (ie, diagnostic interpretation) of physician claims data. We excluded multiple tests per day to the same patient for any diagnostic service. Codes for echocardiography included 1D or 2D studies with or without Doppler examination (G561, G562, G567, G568, G571, G572, and G575). Stress tests were identified by G319, whereas perfusion imaging tests included exercise or dipyridamole myocardial perfusion imaging with or without SPECT using either sestamibi or thallium as its radiotracer (J607, J608, J609, J666, J807, J808, J809, and J866). Such codes have been used previously to examine noninvasive cardiac testing in Ontario.16 Given that perfusion imaging tests may be conducted over 1 or more consecutive days, we applied a 2-day window on either side of the date of a nuclear imaging claim to avoid duplicate counting. A similar 2-day window was applied to stress testing to differentiate isolated stress tests from those with concomitant perfusion.16 We examined 2 surrogates of testing yield, the ratio of angiograms per noninvasive stress or imaging test and the ratio of revascularization procedures per angiogram.
Prevalence of Coronary Artery Disease
Annual age- and sex-adjusted AMI hospitalization rates were used as a surrogate for the underlying prevalence of coronary artery disease. Previous evidence has demonstrated that the prevalence of AMI hospitalization is strongly correlated with population-based illness rates22 and with variations in the burden of self-reported risk factors (R2=0.56) and hospitalizations for chest pain syndromes (R2=0.51), angina (R2=0.64), and congestive heart failure (R2=0.56) across 16 District Health Council regions throughout Ontario.23,24 AMI admissions were defined with the most responsible diagnostic code (International Classification of Diseases–9th Revision code 410) and a minimum length of stay of 3 days unless death occurred.
Subgroup Analyses
We examined the utilization rates of cardiac testing within 3 prespecified sociodemographic subgroups: age (65 versus <65 years), gender, and socioeconomic status (median household income above versus below $50 000 per year). These subgroups were chosen because each has previously been demonstrated to be an important determinant of cardiac testing, and evidence that suggests that temporal changes in cardiac service intensity among selected clinical subgroups (eg, AMI) has preferentially favored elderly and higher-income patients.12
To explore how utilization patterns have changed over time among selected clinical subgroups, we examined those hospitalized for AMI within the previous 3 months and those hospitalized within the last 6 months of life, for 2 reasons. First, both represented subgroups of patients with greater illness severity than the general population,25 yet they differ in the characteristics of illness severity they represent; recent AMI hospitalization serves as a better surrogate for cardiac severity, whereas end-of-life status better serves to identify patients with chronic or palliative disease. Second, both subgroups account for significant overall healthcare expenditures.26–28 In addition, some may argue that disproportional increases in cardiac testing in the end of life may represent a less appropriate or efficient use of cardiac resources.29,30
Statistical Analysis
Population-based rates of cardiac tests and procedures were computed by year, and according to 5-year age group, gender, and income groups, by dividing the number of procedures by the corresponding adult population of Ontario. They were directly standardized according to age and gender with the Ontario population.
For each cardiac test and procedure, we used Poisson regression to estimate the annual rate of increase over the 10-year period, adjusted for age and gender.31 The unit of analysis was the age group, gender, and DA stratum, both for computational efficiency and because we did not distinguish among multiple tests to the same person. The dependent variable was the stratum-specific annual count of tests/procedures, with an offset to incorporate the stratum-specific population. The annual trend over the 10-year study period was estimated with a continuous term for calendar year, adjusted for age group, gender, and their interactions. Point estimates and CIs for the relative annual rates were obtained by exponentiation of the regression parameter that corresponded to the calendar-year term. Trend models were also stratified by age group (age 20 to 64 years or 65 years), gender, socioeconomic status (median household income <$50 000; $50 000), recent AMI, and end-of-life subgroups. Annual trends were assessed for similarity across subgroups with a test for interaction between subgroup and calendar year. We used longitudinal data analysis methods (generalized estimating equation) for clustered count data to account for the correlation among outcomes over time.32 SAS version 8.2 statistical software using the procedure GENMOD was used for modeling. All tests were performed at the 5% level of significance and were 2-sided.
Results
Overall Temporal Trends
Figure 1 illustrates unadjusted relative changes in the use of cardiac technologies compared with 1992. Increases in cardiac testing rates outstripped both demographic shifts and prevalence of AMI, as measured through AMI hospitalizations, over the same time period (Figure 1). Table 1 illustrates that rates for all technologies demonstrated significant exponential growth over the study period. Age- and sex-adjusted population rates were highest for echocardiography and lowest for PCI (during earlier years) and coronary artery bypass surgery (during later years). The yield of stress tests to catheterization, calculated as the ratio of their respective counts, steadily rose from 11.6% in 1992 to 16.7% in 2001, whereas the ratio of revascularization to angiography remained relatively stable over time (46%).
Overall Expenditures
Table 2 illustrates the total expenditures associated with cardiac technology over the past 10 years. Annual expenditures for cardiac technology in the evaluation or management of ischemic heart disease increased by nearly 2-fold over the 10-year study period and exceeded (CAN) $2.8 billion over the 10-year period. They were highest for coronary artery bypass surgery (CAN $1.2 billion), followed by echocardiography (CAN$498 million) and coronary angiography (CAN $402 million). Increases in costs mirrored the temporal changes in the growth of cardiac technologies (Table 2).
Subgroup Analyses
Table 3 illustrates absolute and relative trends over time for cardiac technology according to age, gender, and socioeconomic subgroups. The elderly, males (except echocardiography), and low socioeconomic status groups generally had higher testing and procedure rates; however, annual increases in cardiac testing and procedure rates were consistently higher among the elderly and, with the exception of echocardiography, were significantly greater for women. With 2 exceptions, annual increases in the use of cardiac technology were similar across socioeconomic groups; however, echocardiography and coronary angiography rates increased disproportionately among higher-income groups.
The proportion of noninvasive tests to patients within 3 months of AMI steadily declined over time, whereas the converse was true for invasive procedures (Figure 2). The proportion of cardiac technology use to patients within their last 6 months of life was under 5%, regardless of year, test, or procedure, and with 1 exception (PCI), it steadily declined over time (Figure 2).
Discussion
The long-term sustainability of Medicare in Canada continues to receive intense debate among public, provider, and policy stakeholders, especially in view of the steady growth in healthcare expenditures throughout the past decade.7–9 As demonstrated in the present study, the proliferation of cardiac invasive and noninvasive technology accounts for a significant component of rising health-related expenditures. Specifically, in 2001, approximately CAN $400 000 000 was spent on the evaluation and management of ischemic heart disease in Ontario alone, a cost that doubled over 10 years.
Although the proliferation of cardiac technology has been widespread, temporal growth has been most marked for coronary angiography and PCI, a phenomenon only partially attributable to the more intensive use of such interventions among AMI populations. Although the proliferation of coronary angiography has outstripped corresponding increases in nonimaging and imaging stress testing, the use of noninvasive diagnostic services has increased in magnitudes ranging from 2% to 10% per year, after adjustment for age, gender, and socioeconomic status. Notwithstanding such temporal patterns, the population rates for perfusion imaging stress tests, coronary angiography, PCI, and coronary artery bypass surgery services in Canada in 2001 lag significantly behind corresponding 1992 population rates in the United States.14,33
Our findings are consistent with many studies that have also demonstrated accelerated growth in the use of cardiac technology.12,34–39 For example, our US colleagues have demonstrated similar dramatic relative increases in cardiac testing and treatment over time, although with absolute rates in 1993 that exceed those in Canada from 2001.40 Moreover, intercountry differences in the absolute rates of invasive cardiac testing appear to correlate with their corresponding variations in catheterization laboratory supply (eg, catheterization laboratory supply of 0.42 versus 1.2 per 100 000 in Ontario versus the United States, respectively). Temporal changes in growth are likely attributable to continued proliferation in cardiac technology and specialty physician supply, as well as to changes in physician referral behaviors.15,36 In short, the magnitude of annual increases in cardiac service in Ontario appears similar to the corresponding magnitudes observed in other healthcare systems.
The present study was not designed to address appropriateness; however, some may interpret our findings as evidence that referral patterns for cardiac testing in Ontario have been inappropriate, given that cardiac service utilization has been shown to be driven by resource supply and availability rather than by clear indicators of need.17,22,25,41 Indeed, the proliferation of cardiac technology use observed in the present study markedly outstripped demographic shifts and AMI disease prevalence in the population.
Others view such temporal trends as evidence of appropriate care. For example, our results demonstrated that annual increases in cardiac testing rates were significantly greater among higher-risk subgroups such as the elderly and women, which suggests that physician referral propensity appeared to reflect appropriately the underlying risk profile of their patients. Moreover, by virtue of the low testing rates among patients at or near the last 6 months of life, increased referral among elderly high-risk populations still appeared somewhat selective and geared to those who might benefit most and was possibly aligned with the expectations and preferences of the patients whom these physicians serve.29 At minimum, the present study provided no evidence that the accelerated growth of cardiac technologies was inappropriate.
Even if we were to assume that the prevalence of inappropriate utilization for cardiac testing was low, the tremendous costs associated with the rise in cardiac technology would necessitate marked survival and quality-of-life benefits for temporal utilization patterns to meet cost-effective thresholds. For example, in the present study, the $2.87 billion expenditure investment for cardiac technology would have needed to translate into 57 336 lives saved over 10 years for temporal utilization patterns to have met traditional cost-benefit benchmarks (eg, $50 000 per life-year gained), a survival yield which, if true, would exceed those achievable from established life-saving therapies, such as statins, when used in secondary prevention populations.42–44
Equally important is the absence of incentives, disincentives, or safeguards that might otherwise curtail the continued proliferation of cardiac technologies, especially when one considers that increases in resource expenditures do not appear to have been offset by potential cost savings associated with fewer cardiac hospitalizations or medications. For example, expenditures for cardiovascular pharmacotherapy in Canada have risen 94% between 1996 and 2001, an increase of $1.6 billion in annual drug expenditures, attributable to a 61% rise in the annual number of cardiovascular drug prescriptions.45 If left unchecked, such temporal utilization patterns raise serious concerns about the sustainability of Medicare in Canada, especially when one considers their associated annual expenditures. As a corollary, the disproportionate expenditures allocated to cardiac technology compared with cardiac rehabilitation46,47 underscore an imbalance in focus between "technology" and "prevention initiatives" in Canada.
There are several limitations associated with the present study. First, our analysis was cross-sectional in design and was conducted without available clinical detail or information about referral indication. We also used human ecology rather than individual indicators of socioeconomic status, which may have led to misclassification of individual-level socioeconomic status. Nonetheless, available evidence has demonstrated that such ecological indicators are valid proxies for socioeconomic status.19,48 Although the inclusion of more clinical detail and/or individual socioeconomic indicators might have better explained changes in referral patterns over time, the overall temporal patterns observed in the present study would not have changed. Second, limitations in data availability precluded us from identifying any noninvasive testing conducted on hospitalized patients, which would underestimate the temporal growth in noninvasive testing. Third, our study did not adjust for inflationary pressures; however, inflation rates were marginal throughout the study period and would have accounted for only a small proportion of the increase in cardiac technology expenditures. Finally, our data were limited to Ontario. It is possible that temporal trends in the use of cardiac services may have differed in other Canadian jurisdictions. Nonetheless, Ontario represents 40% of the Canadian population, and temporal growth in cardiac technology has occurred in similar magnitudes in other Canadian jurisdictions.49,50
In conclusion, the present study illustrates continued annual acceleration in cardiac technology utilization. Although changes in utilization patterns were generally geared more preferentially to higher-risk subgroups than to lower-risk subgroups, definitive conclusions about the appropriateness of such proliferation cannot be drawn. Notwithstanding this, the high costs associated with the rise in cardiac technology would necessitate marked benefits in outcomes to justify the cost-effectiveness of procedural growth rates in the population. Given the absence of safeguards that prevent or attenuate the proliferation in cardiac technology use, continued growth in the use of invasive and noninvasive cardiac services will challenge the sustainability of Medicare in Canada.
Acknowledgments
This work was supported in part by a grant by the Heart and Stroke Foundation of Canada. The Institute for Clinical Evaluative Sciences is supported in part by a grant from the Ontario Ministry of Health. Dr Alter is a New Investigator at the Canadian Institutes of Health Research. The results, conclusions, and opinions are those of the authors, and no endorsement by the Ministry of Health and Long Term Care or the Institute for Clinical Evaluative Sciences.
Disclosures
None.
References
Canada Health Act. C-6, Section 10. 1984.
Davidson A. Romanow, Kirby and Courchene: Canada’s health system: a moral or a business enterprise Healthc Manage Forum. 2003; 16: 26–28.
Globerman S, Hodges H, Vining A. Canadian and the United States’ health care systems performance and governance: elements of convergence. Appl Health Econ Health Policy. 2002; 1: 75–88.
Guyatt G, Yalnizyan A, Devereaux PJ. Solving the public health care sustainability puzzle. CMAJ. 2002; 167: 36–38.
Organisation for Economic Co-operation and Development Web site. Available at: http://www.oecd.org. Accessed May 4, 2005.
Cardiac Care Network of Ontario, Target Setting Working Group. Final Report and Recommendations. Available at: http://www.ccn.on.ca/pdfs/IC-2001-06.pdf. Accessed January 13, 2006.
Buske L. Health care spending rises 4.6% in 2003. CMAJ. 2004; 170: 325.
Buske L. Health care bill reaches 3245 dollars per Canadian. CMAJ. 2003; 169: 595.
Seshamani M, Gray A. Health care expenditures and ageing: an international comparison. Appl Health Econ Health Policy. 2003; 2: 9–16.
Brophy JM. The epidemiology of acute myocardial infarction and ischemic heart disease in Canada: data from 1976 to 1991. Can J Cardiol. 1997; 13: 474–478.
Chan B, Young W. Burden of cardiac disease. In: Cardiovascular Health and Services in Ontario: An ICES Atlas. Naylor CD SP, ed. Toronto, Canada: Institute for Clinical Evaluative Sciences; 2000.
Khaykin Y, Austin PC, Tu JV, Alter DA. Utilisation of coronary angiography after acute myocardial infarction in Ontario over time: have referral patterns changed Heart. 2002; 88: 460–466.
Feldman HA, McKinlay SM. Cohort versus cross-sectional design in large field trials: precision, sample size, and a unifying model. Stat Med. 1994; 13: 61–78.
Wennberg DE, Birkmeyer JD. Coronary artery disease. In: The Dartmouth Atlas of Cardiovascular Health Care. Hanover, NH: Center for the Evaluative Clinical Sciences, Dartmouth Medical School; 1999: 51–96.
Goodman DC. Twenty-year trends in regional variations in the U.S. physician workforce. Health Aff (Millwood). 2004; Suppl Web Exclusive: VAR90–VAR97.
Alter DA, Przybysz R, Iron K. Non-invasive Cardiac Testing in Ontario. Toronto, Canada: Institute for Clinical Evaluative Sciences; 2004. Available at: http://www.ices.on.ca/file/Cardiac_ Testing%20Report_Oct_2004.pdf. Accessed November 2, 2004.
Alter DA, Naylor CD, Austin PC, Tu JV. Long-term MI outcomes at hospitals with or without on-site revascularization. JAMA. 2001; 285: 2101–2108.
Alter DA, Iron K, Austin PC, Naylor CD. Socioeconomic status, service patterns, and perceptions of care among survivors of acute myocardial infarction in Canada. JAMA. 2004; 291: 1100–1107.
Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med. 1999; 341: 1359–1367.
Wilkins R. Use of postal codes and addresses in the analysis of health data. Health Rep. 1993; 5: 157–177.
Ontario Case Costing Initiative Web site. Available at: http://www.OCCP.com. Accessed November 2, 2004.
Wennberg JE, Barnes BA, Zubkoff M. Professional uncertainty and the problem of supplier-induced demand. Soc Sci Med. 1982; 16: 811–824.
Jaglal SB, Bondy SJ, Slaughter PM. Risk factors for cardiovascular disease. In: Naylor CD, Slaughter PM, eds. Cardiovascular Health and Services in Ontario: An ICES Atlas. Toronto, Canada: Institute for Clinical Evaluative Sciences; 1999: 63–82.
Basinski ASH. Hospitalization for cardiovascular medical diagnoses. In: Naylor CD, Slaughter PM, eds. Cardiovascular Health and Services in Ontario: An ICES Atlas. Toronto, Canada: Institute for Clinical Evaluative Sciences; 1999: 15–50.
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending: part 2: health outcomes and satisfaction with care. Ann Intern Med. 2003; 138: 288–298.
Hoover DR, Crystal S, Kumar R, Sambamoorthi U, Cantor JC. Medical expenditures during the last year of life: findings from the 1992–1996 Medicare current beneficiary survey. Health Serv Res. 2002; 37: 1625–1642.
Levinsky NG, Yu W, Ash A, Moskowitz M, Gazelle G, Saynina O, Emanuel EJ. Influence of age on Medicare expenditures and medical care in the last year of life. JAMA. 2001; 286: 1349–1355.
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending: part 1: the content, quality, and accessibility of care. Ann Intern Med. 2003; 138: 273–287.
Teno JM, Fisher ES, Hamel MB, Coppola K, Dawson NV. Medical care inconsistent with patients’ treatment goals: association with 1-year Medicare resource use and survival. J Am Geriatr Soc. 2002; 50: 496–500.
Covinsky KE, Eng C, Lui LY, Sands LP, Yaffe K. The last 2 years of life: functional trajectories of frail older people. J Am Geriatr Soc. 2003; 51: 492–498.
McCullagh P, Nelder JA. Generalized Linear Models. New York, NY: Chapman & Hall; 1989.
Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986; 42: 121–130.
Lucas FL, Wennberg DE, Malenka DJ. Variation in the use of echocardiography. Eff Clin Pract. 1999; 2: 71–75.
Chan B, Cox JL, Anderson G. Trends in the utilization of noninvasive cardiac diagnostic tests in Ontario from fiscal year 1989/90 to 1992/93. Can J Cardiol. 1996; 12: 237–248.
Gore JM, Goldberg RJ, Alpert JS, Dalen JE. The increased use of diagnostic procedures in patients with acute myocardial infarction: a community-wide perspective. Arch Intern Med. 1987; 147: 1729–1732.
Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996–1998 practice patterns. Radiology. 2002; 222: 144–148.
Togni M, Balmer F, Pfiffner D, Maier W, Zeiher AM, Meier B. Percutaneous coronary interventions in Europe 1992–2001. Eur Heart J. 2004; 25: 1208–1213.
Pate GE, Humphries KH, Izadnegahdar M, Gao M, Kiely M, Carere RG. Population rates of invasive cardiac procedures in British Columbia, 1995 to 2001. Can J Cardiol. 2004; 20: 712–716.
Gillespie ND, Struthers AD, Pringle SD. Changing echocardiography request patterns between 1988 and 1993. Health Bull (Edinb). 1996; 54: 395–401.
Lucas FL, DeLorenzo MA, Siewers AE, Wennberg DE. Temporal trends in the utilization of diagnostic testing and treatments for cardiovascular disease in the United States, 1993–2001. Circulation. 2006; 113: 374–379.
Wennberg D, Dickens J Jr, Soule D, Kellett M Jr, Malenka D, Robb J, Ryan T Jr, Bradley W, Vaitkus P, Hearne M, O’Connor G, Hillman R. The relationship between the supply of cardiac catheterization laboratories, cardiologists and the use of invasive cardiac procedures in northern New England. J Health Serv Res Policy. 1997; 2: 75–80.
Jackevicius CA, Mamdani M, Tu JV. Adherence with statin therapy in elderly patients with and without acute coronary syndromes. JAMA. 2002; 288: 462–467.
Ko DT, Mamdani M, Alter DA. Lipid-lowering therapy with statins in high-risk elderly patients: the treatment-risk paradox. JAMA. 2004; 291: 1864–1870.
Ross SD, Allen IE, Connelly JE, Korenblat BM, Smith ME, Bishop D, Luo D. Clinical outcomes in statin treatment trials: a meta-analysis. Arch Intern Med. 1999; 159: 1793–1802.
Jackevicius CA, Tu K, Filate WA, Brien SE, Tu JV. Trends in cardiovascular drug utilization and drug expenditures in Canada between 1996 and 2001. Can J Cardiol. 2003; 19: 1359–1366.
Stone J. Cardiac rehabilitation: cost and care effective. Can J Cardiol. 2004; 20: 1256–1257.
Taylor RS, Brown A, Ebrahim S, Jolliffe J, Noorani H, Rees K, Skidmore B, Stone JA, Thompson DR, Oldridge N. Exercise-based rehabilitation for patients with coronary heart disease: systematic review and meta-analysis of randomized controlled trials. Am J Med. 2004; 116: 682–692.
Finkelstein MM. Ecologic proxies for household income: how well do they work for the analysis of health and health care utilization Can J Public Health. 2004; 95: 90–94.
Pilote L, Lavoie F, Ho V, Eisenberg MJ. Changes in the treatment and outcomes of acute myocardial infarction in Quebec, 1988–1995. CMAJ. 2000; 163: 31–36.
Quan H, Cujec B, Jin Y, Johnson D. Acute myocardial infarction in Alberta: temporal changes in outcomes, 1994 to 1999. Can J Cardiol. 2004; 20: 213–219.(David A. Alter, MD, PhD, FRCPC; Therese )