Radiological Alliance for Health Services Research
Comparative Effectiveness Research in Radiology: Patients, Physicians and Policy Makers James V. Rawson, MD, FACR Health care costs continue to rise, with significant geographic differences in spending on health care and outcomes within the United States. The goal of comparative effectiveness research is to reduce health care spending without adversely effecting overall health according to the Congressional Budget Office. There are unique challenges and barriers to applying comparative effectiveness research to radiology, including rapidly changing technology, complex multistep care processes, and the burden of proving the impact of a diagnostic exam on patient outcome. Radiology shares other challenges of acceptance of comparative effectiveness research results (diffusion of new knowledge and successful implementation of changes in clinical practice) with all of health care, but with the added complication that radiologists do not order radiology exams. Key Words: Comparative effectiveness research; health carer outcomes; health care reform. ªAUR, 2011
O
ver the past three decades, there has been great innovation and increased volume in medical imaging (1). Entire invasive procedures have been replaced by noninvasive imaging (eg, pneumoencephalography and head computed tomography [CT]). When 235 leading internists were surveyed to rank the innovations whose loss would have the most adverse effect on patients (2), five radiology procedures ranked in the top 12, with magnetic resonance imaging (MRI) and CT listed number one. Others listed in the top 12 included balloon angioplasty, mammography, and ultrasound. The growth in imaging technology was paralleled by the growth in imaging utilization and costs. Annual spending on diagnostic imaging increased from $220 to $419 per Medicare beneficiary between 2000 and 2006 (3). Despite the belief in the benefit of medical imaging to patients, evidence-based research is needed to show that that a specific procedure has a benefit to a patient or population in a specific clinical setting. The American Reinvestment and Recovery Act budgeted $1.1 billion for comparative effectiveness research (CER). It established the Federal Coordinating Council for Comparative Effectiveness Research to direct the vision and priorities for CER conducted under all federal agencies. The council sees the purpose of CER research (4–6) to improve health care decisions by reducing Acad Radiol 2011; 18:1067–1071 From the Department of Diagnostic, Therapeutic and Interventional Radiology, Georgia Health Science University, 1120 15th Street, Augusta, GA 30912. Received September 29, 2010; accepted January 14, 2011. Address correspondence to: J.V.R. e-mail:
[email protected] ªAUR, 2011 doi:10.1016/j.acra.2011.01.021
uncertainty when multiple treatment options are available. In 2009, the council defined CER as ‘‘the conduct and synthesis of research comparing the benefits and harms of different interventions and strategies to prevent, diagnose, treat and monitor health conditions in ‘real world’ settings. The purpose of this research is to improve health outcomes by developing and disseminating evidence-based information to patients, clinicians, and other decision makers, responding to their expressed needs, about which interventions are most effective for which patients under specific circumstances’’ (5). The Institute of Medicine has established the top 100 priorities for CER and defined CER as ‘‘the generation and synthesis of evidence that compares the benefits and harms of alternative methods to prevent, diagnose, treat, and monitor a clinical condition or to improve the delivery of care. The purpose of CER is to assist consumers, clinicians, purchasers, and policy makers to make informed decisions that will improve health care at both the individual and population levels’’ (4). These definitions acknowledge the multiple stakeholders in health care that would be expected to use the data of these studies as well as specifying real-world settings. They also establish the need to gather evidence-based medicine and disseminate it with the objective of improving health outcomes and the delivery of care. Also included in the definition of CER is the impact on individual patients as well as populations, which acknowledges personalized medicine. However, in a 2007 report on CER, the Congressional Budget Office (CBO) (7) described CER as an opportunity to constrain costs in both sectors (federal spending and private sector) without adverse health consequences. The rationale 1067
RAWSON
for this belief is based in part on the significant geographic variations in spending and clinical practices without clear improvement in life expectancy or other health statistics. The CBO suggested ‘‘that generating better information about the costs and benefits of different treatment options— through research on the comparative effectiveness of those options—could help reduce health care spending without adversely affecting health overall.’’ Variations in practice patterns are not a new phenomenon in medicine. Dr J. Allison Glover described his observations on the variations in tonsillectomy rates in the 1930s. In the 1980s, Wennberg and Gittelsohn (8) published observations on the variations in hysterectomy rates and cancer treatment options without demonstrable changes in outcomes, attracting attention to the variations in practice patterns. The CBO (7) has cited the wide range of Medicare spending per capita in the United States. Medicare spending per capita ranged from $4,500 to $11,600 in 2003. The Centers for Disease Control and Prevention (9) reported the variation in the rate that Medicare decedents aged $ 65 years used intensive care unit or coronary care unit care in the last 6 months of life. Intensive care unit stays ranged from a low of 23% in the states of Vermont and North Dakota to a high of 49% in Florida in 2005. Recently, PricewaterhouseCoopers (10) analyzed the >$2 trillion spent on US health care and categorized costs that are not directly providing benefit to patients. An estimated $210 billion annually is due to defensive medicine and unnecessary tests. Up to $210 billion annually is also spent on inefficient claims processing. An additional $200 billion is spent annually for care related to preventable conditions due to obesity. Although the exact value of any of these figures can be debated, this identifies a significant portion of American health care as potential waste and noncontributory to the improved outcomes of patients. If addressed, this would put our national health care expenditures more in line with those of other industrialized nations as a percentage of gross domestic product. Of the top 100 priority items designated by the Institute of Medicine as areas of priority for CER (4), nine directly affect radiology (11). Themes include cancer imaging, the role of specialists in imaging, the effectiveness of studies ordered by specialist compared to nonspecialists, and comparison of the effectiveness of imaging performed by radiologists and nonradiologists. This represents an opportunity to identify practices that do not improve patient outcomes as well as standardize care using practices that do improve patient outcomes. USE OF COMPARATIVE EFFECTIVENESS RESEARCH DATA CER data will likely be used by many stakeholders, from patients and physicians to policy makers. Unlike surgical or pharmacologic intervention, there is a gap between diagnostic tests and outcomes, with many steps in between (12). Lee and Foster (13) showed the lower mortality associated with 1068
Academic Radiology, Vol 18, No 9, September 2011
imaging services during an admission. Another study (14) showed lower lengths of stay for patients who received crosssectional imaging earlier in their admissions. Yet the direct linkage to patient outcomes for a diagnostic test is a challenge. Currently, the Centers for Medicare and Medicaid Services (CMS) 24 do not include any radiology metrics. It will not be enough to show that diagnostic imaging resulted in a change in management, as in the National Oncology PET Registry study; a change in patient outcomes must be demonstrated. Outcomes must be significant to patients, their families, and society. This could be a paradigm shift for the radiologist-patient relationship. Advocates for patient-centered radiology (15) would see this as an opportunity for radiologists to take a more active role in patient care and to educate patients about their options. This would be consistent with the American College of Radiology’s Face of Radiology campaign (16). CER studies must inform not only patients and physicians but also policy makers. The role of cost in CER is controversial (11). Some argue that CER without cost analysis is of limited value to policy makers. Others argue that introducing economics would inappropriately introduce cost to coverage decisions. The Patient Protection and Affordable Care Act placed limitations on the use of CER data by Medicare (17). CER cannot include a mandate for coverage or payment, and CER evidence can be used only as part of a larger process. CMS cannot use evidence that assigns a lower value to life if a patient is elderly, disabled, or terminally ill. CMS cannot use quality-adjusted life-years or similar costeffectiveness thresholds to determine coverage or payment. Because part of the CBO’s intent of CER is to lower costs without adversely affecting care, costs will need to enter the process at some point. It could be argued that a better understanding of the effectiveness of diagnostic radiology exams is needed before their cost can be addressed. Given the rate of growth of national health care expenditures with respect to the gross domestic product, there may not be enough time to examine effectiveness and cost in series. Health care reform includes the creation of a board to expand Medicare solvency and reduce spending growth. The Independent Payment Advisory Board’s first recommendations are due in 2014. If Congress fails to pass legislation that obtains the required savings through other policy changes, the Independent Payment Advisory Board’s recommendations will automatically take effect. This is similar to the National Institute for Health and Clinical Excellence, which is part of the National Health Service in Great Britain (18). Clinical outcomes in specific clinical settings and for defined patients and populations will help policy makers. PATIENT POPULATION One element of any CER study will need to be the definition of the population. Recent studies to evaluate the efficacy of virtual colonography (19) were criticized (20) for not being conducted specifically in Medicare populations; this resulted
Academic Radiology, Vol 18, No 9, September 2011
in a negative coverage decision for virtual colonoscopy for the Medicare populations. The Medicare population is not the same as the commercially insured population. The Medicare population differs from the commercially insured population by age, most common inpatient diagnosis-related group, and comorbidities. The Medicare population has obesity, diabetes, hypertension, and chronic obstructive pulmonary disease as comorbidities, representing a burden of chronic disease on this population (21). This has contributed to the spending on the Medicare population. In the non-Medicare population, under health care reform, a significant portion of the population that has not previously been insured would become insured. This population would also need to be educated as to their own responsibilities in maintenance of their health. In an era of personalized medicine, CER studies will need to identify small subpopulations that benefit from clinical services, even if those services do not improve care for the majority of the population. OBSTACLES TO COMPARATIVE EFFECTIVENESS RESEARCH IN RADIOLOGY Several obstacles exist to the successful application of CER to radiology exams. One challenge for CER in radiology is the rapidly changing technology (1). Large-scale, multicenter clinical trials try to compare mature diagnostic imaging techniques, but imaging technologies are rapidly evolving. While studies were being conducted to evaluate 16-detector computed tomographic scanners, 64-detector scanners were being developed. CER choices of diagnostic imaging modalities have been compared to the Coke versus Pepsi scenario (22). Radiology exams are part of a complex chain of events, from the ordering physician to the interpreting radiologist to the treating physician. How does one isolate a single chest x-ray during an admission or an outpatient workup and show the survival benefit to the patient? How does one demonstrate the benefit of one step in a multistep process? Once a CER study is completed, it will need to be adopted by stakeholders and implemented. In this arena, radiology shares many implementation challenges with the rest of the health care community. Consider the case of the recent US Preventive Services Task Force recommendations on indications for breast cancer screening (23) and the response (24). How would a CER study be viewed differently? Although CER research in radiology would focus on the impact of radiology exams, it is not radiologists who order these exams. Thus, referring physicians’ ordering behavior would need to change. Some success has been obtained using the American College of Radiology’s Appropriateness Criteria incorporated into electronic order entry systems. Decision support software has the potential to assist in the radiology exam being ordered that is most appropriate for the clinical setting, providing lower costs and improving the appropriateness of imaging studies (25). In studying the diffusion of knowledge, a common case study is scurvy (26). In 1497, 100 of Vasco de Gama’s crew
COMPARATIVE EFFECTIVENESS RESEARCH
of 160 sailors died of scurvy. More sailors died of scurvy than from injuries in warfare. In 1601, Captain Lancaster’s data showed the survival benefit of ascorbic acid in lemon juice in the British navy. Yet no change in practice occurred. Similar improved survival was again demonstrated in the British navy in 1747, yet the navy did not adopt citrus fruit for sailors until 1795. The British Board of Trade adopted a similar policy in 1865 for the merchant marines, 265 years after the initial data of Captain Lancaster. A CER study would need to get buy-in from stakeholders and be more rapidly and widely implemented than the Ottawa Ankle Rules (27). A recent study compared the use of routine daily chest x-rays with an on-demand strategy in the intensive care setting (28). Although it demonstrated a potential decrease in chest x-ray utilization of 32% without an adverse effect on length of stay, days on a ventilator, or mortality, the study occurred in closed intensive care units in France. Opponents of changing this clinical practice would question whether this translates to intensive care units in the United States, intensive care units, without full-time intensivists, or surgical or neurologic intensive care units. Although radiology has the advantage of more sophisticated communication techniques (eg, e-mail, the Internet, and social networking), implementation challenges remain. Getting information to physicians is not enough to change their behavior (29). The Institute of Medicine (30) estimated that it takes 17 years for new knowledge to be incorporated into clinical practice, and even then, the penetration of the new knowledge is heterogeneous. The successful implementation of CER will require using effective techniques for changing physicians’ behavior (31). Some of these potential obstacles to implementation could be tackled with new tools (eg, patient-radiologist relationships, decision support software). Policy-level changes that change patient and physician incentives may also help with implementation. Estimates of the health care costs of defensive medicine range widely (10,32,33). A Massachusetts Medical Society physician survey (34) showed that 28% of computed tomographic and MRI exams were ordered for defensive purposes. In the absence of tort reform, changing ordering physicians’ behavior is not a trivial task. Even if tort reform were passed and evidence-based indications for CT and MRI ordering indications were widely disseminated, it is unlikely that there would be a 28% reduction in the utilization of CT and MRI. Prospective payment systems such as diagnosis-related groups and bundling of services are likely to increase and to encourage the efficient use of care. An alternative proposal suggests that Medicare pay equally for services with equal patient outcomes, regardless of the cost to provide the services (17). Other changes in financial incentives (eg, accountability measures, value-based purchasing, accountable care organizations, pay for performance) could change the relationships between patients, physicians, and hospitals. Recent data on the CMS 24 have shown that 85.9% of US hospitals were >90% complaint with the CMS 24 accountability measures (35), an increase from 20.4% in 2002. Financial incentives 1069
RAWSON
Academic Radiology, Vol 18, No 9, September 2011
are not limited to physicians and hospitals. High-deductible health insurance may turn patients into bargain shoppers and consumers of health care. Many innovations and experimentations in financial incentives are likely to be tried and may change the patient-physician relationship. In addition to the multicenter, clinical trials that the American College of Radiology Imaging Network (ACRIN) conducts, the radiology community needs new tools to be able to successfully implement CER results. These include decision support at order entry, a more visible role in patients’ eyes and data mining of electronic medical records to demonstrate the link between radiology exams and patient outcomes.
strate a change in patient outcomes. Outcomes must be significant to patients, their families, and society. The CBO clearly states that the intent of CER is to change physician and patient behavior, resulting in reduced health care spending without affecting health outcomes. It may be hard to do that without financial incentives and without putting some of the responsibility on patients to lead healthy lifestyles. Although the silver bullet of aligned incentives between patients, physicians, hospitals, and policy makers that results in healthier lifestyles and appropriate utilization of imaging has not been found yet, CER may be a good step in that direction.
CHANGING PATIENT BEHAVIOR
REFERENCES
While changing the physician decision-making process is one objective of CER, another objective involves patients’ choices and lifestyles. A recent analysis of waste in health care identified approximately 10%, or $200 billion, spent annually on preventable conditions related to obesity (10). Cigarette smoking remains the single most important cause of preventable disease in the United States. The annual direct costs to the economy attributable to smoking were in excess of $298 billion, including workplace productivity losses of approximately $67.5 billion, premature death losses of $117 billion, and direct medical expenditures of $116 billion (36). With the economic impact of lifestyle choices made by Americans exceeding 20% of the annual US health care expenditures, it is understandable that this is part of CER. Possible mechanisms include increased patient responsibility for cost associated with unhealthy lifestyles or habits (eg, differential health insurance rates). Possible research implications for radiology are to study the utilization and diagnostic yields in different populations by body mass index and smoking history. The US Preventive Services Task Force has recommended that physicians start screening for obesity at 6 years old and refer obese children to moderate-intensity to high-intensity weight reduction programs (37). In the absence of changes in lifestyle or payment systems for obese patients and smokers, savings will need to be found in other aspects of health delivery. Beyond policy, free-market forces may introduce additional changes to employment-based health systems. Massachusetts police and fire departments put a hiring ban on smokers in place in 1997. The Massachusetts Hospital Association announced that it would no longer hire smokers in 2010 (38). CONCLUSIONS CER represents an opportunity for radiology to demonstrate the value of specific exams to specific populations in a clinical setting. Implementation of these insights will likely require changes in patient behavior, physician behavior, and policy. This could result in both a new business model that is driven by better aligned incentives and a new paradigm for radiologists with more patient contact. CER will need to demon1070
1. Hillman BJ, Goldsmith JC. The Sorcerer’s Apprentice. New York: Oxford University Press, 2011. 2. Fuchs VR, Sox HC Jr. Physician’s views of the relative importance of thirty medical innovations. Health Aff (Millwood) 2008; 20:30–42. 3. US Government Accountability Office. Medicare Part B imaging services: rapid spending growth and shift to physician offices indicate need for CMS to consider additional management practices. Available at: http://www. gao.gov/new.items/d08452.pdf. Accessed November 14, 2010. 4. Institute of Medicine. Report brief: initial national priorities for comparative effectiveness research. Available at: http://www.iom.edu/cerpriorities. Accessed November 7, 2010. 5. Federal Coordinating Council for Comparative Effectiveness Research. Report to the president and Congress. Available at: http://www.hhs.gov/ recovery/programs/cer/cerannualrpt.pdf. Accessed November 14, 2010. 6. Society for Medical Decision Making. Background paper on comparative effectiveness research. Available at: http://www.smdm.org/documents/ SMDMCERStatementMay2009_002.pdf. Accessed November 14, 2010. 7. Congressional Budget Office. CBO research on the comparative effectiveness of medical treatments: issues and options for an expanded federal role. Available at: http://www.cbo.gov/ftpdocs/88xx/doc8891/12-18ComparativeEffectiveness.pdf. Accessed November 14, 2010. 8. Wennberg J, Gittelsohn A. Variations in medical care in small areas. Sci Am 1982; 246:120–134. 9. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. Health, United States, 2009 with special feature on medical technology. Available at: http://www.cdc.gov/nchs/data/hus/hus09.pdf. Accessed November 7, 2010. 10. PricewaterhouseCoopers Health Research Institute. The price of excess: identifying waste in healthcare spending. Available at: http://www.pwc. com/us/en/healthcare/publications/the-price-of-excess.jhtml. Accessed March 1, 2010. 11. Pandharipande PV, Gazelle GS. Comparative effectiveness research: what it means for radiology. Radiology 2009; 253:600–605. 12. Hollingworth W. Radiology cost and outcomes studies: standard practice and emerging methods. AJR Am J Roentgenol 2005; 185:833–839. 13. Lee DW, Foster DA. The association between hospital outcomes and diagnostic imaging: early findings. J Am Coll Radiol 2009; 6:780–785. 14. Batlle JC, Hahn PF, Thrall JH, et al. Patients imaged early during admission demonstrate reduced length of hospital stay: a retrospective cohort study of patients undergoing cross-sectional imaging. J Am Coll Radiol 2010; 7: 269–276. 15. Brandt-Zawadski M, Kerlan RK. Patient-centered radiology: use it or lose it! Acad Radiol 2009; 16:521–523. 16. Neiman HL. Face of Radiology campaign. Acad Radiol 2009; 16:517–520. 17. Pearson SD, Bach PB. How Medicare could use comparative effectiveness research in deciding on new coverage and reimbursement. Health Aff (Millwood) 2010; 29:1796–1804. 18. Daschle T, Greenberger SS, Lambrew JM. Critical: What We Can Do About the Health-Care Crisis. New York: Thomas Dune, 2008. 19. Johnson CD, Chen M-H, Toledano AY, et al. Accuracy of CT colonography for detection of large adenomas and cancers. N Engl J Med 2008; 359: 1207–1217. 20. Dhruva SD, Phurrough SE, Salive ME, et al. CMS’s landmark decision on CT colonography—examining the relevant data. N Engl J Med 2009; 360:2699–2701.
Academic Radiology, Vol 18, No 9, September 2011
21. Thorpe KE, Howard DH. The rise in spending among Medicare beneficiaries: the role of chronic disease prevalence and changes in treatment intensity. Health Aff (Millwood) 2006; 25:w378–w388. 22. Hillman BJ. Coke vs Pepsi. J Am Coll Radiol 2009; 6:666. 23. US Preventive Services Task Force. Screening for breast cancer: recommendation statement. Available at: http://www.uspreventiveservicestaskforce. org/uspstf09/breastcancer/brcanrs.htm. Accessed November 8, 2010. 24. Hillman BJ. The problem with CER. J Am Coll Radiol 2010; 7:241–242. 25. Sistrom CL, Dang PA, Weilburg JB, et al. Effect of Computerized order entry with integrated decision support on the growth of outpatient procedure volumes: seven-year time series analysis. Radiology 2009; 251:147–155. 26. Rogers EM. Diffusion of Innovation. New York: Free Press, 1995. 27. Stiell IG, Greenberg GH, McKnight RD, et al. Decision rules for the use of radiography in acute ankle injuries: refinement and prospective validation. JAMA 1993; 269:1127–1132. 28. Hejblum G, Chalumeau-Lemoine L, Ioos V, et al. Comparison of routine and on-demand prescription of chest radiography in mechanically ventilated adults: a multicentre, cluster-randomized, two-period crossover study. Lancet 2009; 374:1687–1693. 29. Grimshaw JM, Eccles MP, Walker AE. Changing physicians’ behavior: what works and thoughts on getting more things to work. J Contin Educ Health Prof 2002; 22:237–243. 30. Institute of Medicine, Committee on Quality of Health Care in America. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press, 2001.
COMPARATIVE EFFECTIVENESS RESEARCH
31. Smith WR. Evidence for the effectiveness of techniques to change physician behavior. Chest 2000; 118:8S–17S. 32. Kessler D, McClellan M. Do doctors practice defense medicine? Q J Econ 1996; 111:353–390. 33. Congressional Budget Office. Key issues in analyzing major health insurance proposals. Available at: http://www.cbo.gov/ftpdocs/99xx/ doc9924/12-18-KeyIssues.pdf. Accessed November 7, 2010. 34. Massachusetts Medical Society. Investigation of defensive medicine in Massachusetts. Available at: http://www.massmed.org/AM/Template. cfm?Section=Professional_Liability&CONTENTID=23557&TEMPLATE=/ CM/ContentDisplay.cfm. Accessed November 7, 2010. 35. Chassin MR, Loeb JM, Schmaltz SP, et al. Accountability measures— using measurement to promote quality. N Engl J Med 2010; 363:683–688. 36. Rumberger JS, Hollenbeak CS, Kline D. Potential costs and benefits of smoking cessation in the United States. Available at: http://www.lungusa. org/stop-smoking/tobacco-control-advocacy/reports-resources/cessationeconomic-Benefits/reports/SmokingCessationTheEconomicBenefits.pdf. Accessed October 7, 2010. 37. US Preventive Services Task Force. Screening for obesity in children and adolescents. Available at: http://www.uspreventiveservicestaskforce.org/ uspstf10/childobes/chobesrs.htm. Accessed November 7, 2010. 38. Massachusetts Hospital Association. MHA policy to not hire tobaccousers. Available at: http://www.mhalink.org/AM/Template.cfm?Section= Newsroom&template=/CM/ContentDisplay.cfm&contentid=13254. Accessed November 10, 2010.
1071
