A systematic review of medical practice variation in OECD countries
Introduction
Health policy makers, researchers, and clinicians have long been aware of large variations in the use of medical care across regions and medical providers. Variation in health care resources and utilization raise questions of the quality, equity, and efficiency of resource allocation and use, and have important implications for health care and health policy. If populations in regions with higher spending and more resources have better care or improved health, then other regions might benefit from similar care. Alternatively, if residents of regions with more resources do not benefit from the additional spending, there are opportunities for reorganization of existing services or for re-direction of resources to meet other health care or social objectives.
Much of the conceptual, methodological and expository work in variations research has emanated from the Dartmouth research group over the past 40 years. Beginning with John Wennberg's seminal Science paper [1], the Dartmouth Atlas of Health Care series [2], [3], [4] has documented large regional variations in rates of hospital admission, surgical procedures and resource supply across US regions, with strong associations found between regional admission rates and acute care bed supply. Early investigations uncovered practice variation across England, Norway and New England [5]. Similar variations in hospital admissions, physician visits, diagnostic testing, prescription drug use, and surgical procedures have been observed within Canada and documented in a series of Ontario atlas publications (http://www.ices.on.ca) [6], [7], [8], [9], the Ontario POWER Study (Project for an Ontario Women's Health Evidence-Based Report) [10], [11], and variations in hospital admissions and pharmaceutical prescriptions across Canada [12], [13], [14]. National Atlases from England, Australia and Spain have document similar medical practice variations (www.atlasvpm.org) [15], [16], [17], [18]. Recently, the European Collaboration for Healthcare Optimization (ECHO) was initiated to understand medical practice variations across a number of European countries (http://www.echo-health.eu/).
Previous research has demonstrated that the magnitude of variations in hospitalization rates for medical and surgical conditions for adults and children depends on many factors besides illness [1], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. For some conditions, such as diabetes or chronic obstructive pulmonary disease (COPD) (deemed “high variation conditions”), regional variations in admission rates are striking; for others, such as incident acute myocardial infarction (AMI), hip fracture and colectomy for early stage colon cancer (deemed “low variation conditions”), there is relatively little variation in admission rates across regions [21], [22]. Low variation conditions are those where hospital admission rates appear to reflect population illness rates (need) since there is little difficulty making the diagnosis, patients always seek care, and hospital admission is mandatory [1], [20], [21], [22]. Only about 10% of hospital admissions are for categories that demonstrated low variation across U.S. regions. High variation conditions are those where admission rates reflect non-medical factors in addition to illness rates (need). For these conditions, associations have been established between health care use and health care resource capacity, physician practice style, specialty and training, leading to substantial physician discretion in decision-making as well as supplier-induced demand [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37].
Dartmouth researchers have identified three categories of services that exhibit the phenomenon known as unwarranted variation [38]. Unwarranted variation is medical practice variation across regions or provider groups (hospitals or physicians) that is not explained on the basis of illness, patient risk factors or patient preferences. These categories provide a theoretical framework for understanding sources of medical practice variations, and remedies for improving health care and reducing variation. Effective care consists of evidence-based interventions for which the benefits exceed the harms so that almost all patients in need should receive the service. Examples are childhood immunizations or beta-blockers following heart attack. The correct rate among eligible patients for most effective care is close to 100%. Variations in such treatments reflect a failure to deliver needed care, or underuse of effective care. Preference-sensitive care encompasses treatment decisions where different choices carry different benefits and risks, and where patients’ attitudes toward the outcomes may vary. The right rate reflects the choices of patients who have been adequately informed. An example would be the use of bypass surgery for heart disease, where surgery is likely to improve patients’ chest pain but carries a small but real risk of causing memory loss. Unwarranted variations in preference-sensitive care reflect both the limitations of current scientific evidence and the failure to ensure informed patient choice. Supply-sensitive care refers to services where the supply of a specific resource (e.g., number of specialists per capita) has a major influence on utilization rates in the absence of evidence for these additional services. Physician visits, hospitalizations, intensive care unit (ICU) admissions, and diagnostic imaging services are examples where local supply influences the frequency of use. If more is not better, there are large implications in terms of increased costs with no health benefits. In that case, these resources might be better deployed and/or reallocated to other, more effective uses.
An earlier literature review of small area analysis studies in major medical journals from 1952 to 1985 uncovered 59 papers of interest [39]. The objective of this study was to update the systematic review of the peer-reviewed literature on medical practice variations in OECD countries and make it available to researchers, and to provide a few examples of the extent of variation within and between countries for select high impact conditions. The systematic review was undertaken at the behest of OECD and resulted in an OECD report on medical practice variations [40].
An examination of medical practice variations within countries can lead to important insights into underuse, overuse and misuse of services. Combined with engagement from clinicians and policy makers, the study of the causes and consequences of variation has the potential to enhance the equity and efficiency of health care systems.
Section snippets
Search strategy
We undertook a literature search using Medline to find publications on medical practice variations in OECD countries. We developed a search strategy on medical practice variations in treatments and procedures based on region of residence, hospital or physician provider. We supplemented it with an independent search culled from papers published by the Wennberg International Consortium (WIC) (http://wennbergcollaborative.org/index.php).
We focused on studies of variations in physician practice
Results
The final search was conducted on March 23, 2011, and retrieved 6396 papers from North America and 1170 from other OECD countries. After applying the exclusions, 779 papers were included. The WIC search retrieved an additional 57 articles, resulting in a total of 836 studies. We have included the references for all 1114 studies including those from 1990 to 1999 (N = 278 studies) in an online Appendix, classified by condition and country.
There were 430 North American studies (51%), and 406 (49%)
Discussion
There were large medical practice variations across regions, hospitals and physician practices for almost every condition and procedure studied, as expected. Studies that involved multiple countries showed variations across these countries. Conditions and procedures studied were quite similar across OECD countries. Many studies focused on high-impact conditions such as AMI, heart failure and stroke, breast and lung cancer, COPD and asthma, obstetric and musculoskeletal conditions, as well as
Funding sources
This study was supported by an Emerging Team Grant (ETC92248) in Applied Health Services and Policy Research from the Canadian Institutes of Health Research (CIHR) and the Organization for Economic Cooperation and Development (OECD). The funding agencies had no role in the study design; in the collection, analysis, or interpretation of data; in the writing of the report; or in the decision to submit the report for publication. The opinions, results and conclusions reported in this paper are
Acknowledgments
We would like to thank Valerie Paris and Gaetan Lafortune, Organization for Economic Cooperation and Development, Paris, for critically reviewing the manuscript.
References (99)
- et al.
Professional uncertainty and the problem of supplier-induced demand
Social Science and Medicine
(1982) - et al.
Cervical cancer screening policies and coverage in Europe
European Journal of Cancer
(2009) - et al.
Contextual analysis of breast and cervical cancer screening and factors associated with health care access among United States women, 2002
Social Science and Medicine
(2008) - et al.
Variations in use of imaging in a national sample of men with early-stage prostate cancer
Urology
(2002) - et al.
Medical practice variations in hospital care; time trends of a spatial phenomenon
Health Place
(2004) - et al.
A multistate comparison of patient characteristics, outcomes, and treatment practices in acute myocardial infarction
American Journal of Cardiology
(2005) - et al.
International differences in evolution of early discharge after acute myocardial infarction
Lancet
(2004) - et al.
Small area variations in health care delivery
Science
(1973) - et al.
The Dartmouth atlas of health care
(1996) - et al.
The Dartmouth atlas of health care 1998
(1998)