Overuse of Cardiovascular Services
2015; Lippincott Williams & Wilkins; Volume: 132; Issue: 3 Linguagem: Inglês
10.1161/circulationaha.114.012668
ISSN1524-4539
AutoresXiaoyan Huang, Meredith B. Rosenthal,
Tópico(s)Cardiac, Anesthesia and Surgical Outcomes
ResumoHomeCirculationVol. 132, No. 3Overuse of Cardiovascular Services Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBOveruse of Cardiovascular ServicesEvidence, Causes, and Opportunities for Reform Xiaoyan Huang, MD, MHCM and Meredith B. Rosenthal, PhD Xiaoyan HuangXiaoyan Huang From Providence Heart Clinic, Portland, OR (X.H.); and Department of Health Policy and Management, Harvard School of Public Health, Boston, MA (M.B.R.). and Meredith B. RosenthalMeredith B. Rosenthal From Providence Heart Clinic, Portland, OR (X.H.); and Department of Health Policy and Management, Harvard School of Public Health, Boston, MA (M.B.R.). Originally published21 Jul 2015https://doi.org/10.1161/CIRCULATIONAHA.114.012668Circulation. 2015;132:205–214Cardiovascular disease is the leading cause of death in the United States and constitutes 17% of overall national health expenditures,1 with noninvasive testing contributing much of the cost growth in recent years.2,3 Cardiovascular medicine has long been at the forefront of evidence-based practice, with arguably the longest-standing and most comprehensive array of efforts to measure risk factors and outcomes, to analyze the clinical effectiveness and cost-effectiveness of treatment options, and to formulate practice recommendations.4,5 Professional societies for cardiovascular medicine have established health policy statements,6 national databases on quality,7 clinical guidelines,8 and appropriate use criteria9 (AUC) for high-cost cardiovascular services (Table 1). More recently, the profession has taken steps to right-size cardiovascular services by defining the problem of overuse and underuse and working to shape clinical and policy interventions that improve quality and access and reduce unnecessary cost.10 For example, the American College of Cardiology Wisconsin and Florida chapters recently received a $15.8 million Centers for Medicare & Medicaid Innovation grant to pilot a comprehensive cardiovascular program to improve quality and to reduce cost of cardiovascular care through better data transparency, patient engagement, and shared decision making (SDM).Table 1. Summary of Initiatives to Identify and Reduce Overuse in Cardiovascular MedicineInitiativeDescriptionAUCJointly developed by ACC Foundation AUC Task Force since 2005 to promote evidence-based, effective use of cardiovascular technologyBased on RAND appropriateness methodologyAppropriateness is defined by weighing clinical benefit against risks, downstream procedures, and follow-up careCan be accessed at www.cardiosource.orgClinical guidelinesJointly developed by ACC/AHA Task Force on Practice Guidelines around discrete clinical conditions since 1980Expert analysis of evidence of benefit and harm for existing procedures and therapy to optimize patient outcome, to improve quality, and to control costCan be accessed at www.cardiosource.orgPerformance measuresDeveloped by the ACC/AHA Task Force on Performance Measures to establish discrete and composite performance measurement standards since 2005The performance measures are evidence based and focus on the process, outcome, composite, and efficiency of careCan be accessed at www.cardiosource.orgGet With The GuidelinesA suite of in-hospital programs developed by the AHA to promote guideline adherence for common clinical conditionsIncludes evidence-based guidelines, clinical tools, and resources for providers; patient educational resources; hospital certification program; clinical data registry; and performance feedback for continued quality improvementCan be accessed at www.heart.orgNational Cardiovascular Data RegistryDeveloped by the ACC in 1997 as a quality data repository for acute and outpatient cardiovascular care and cardiovascular proceduresUsed to collect clinical and outcomes data, to benchmark performance, to identify quality gaps for continued quality improvement, and to measure appropriate use of cardiovascular proceduresCan be accessed at www.ncdr.comChoosing WiselyAn American Board of Internal Medicine Foundation initiative since 2012 to engage providers and patients in conversations to ensure that the right care is delivered at the right time>60 Specialty societies have identified 5 low-value services to promote care through better diagnostic and treatment choices to reduce risk and wasteConsumer Reports is working with providers to help patients avoid potentially harmful and unnecessary careCan be accessed at www.choosingwisely.orgHealth policy statementsOverseen by the ACC Clinical Quality Committee since 1992 to promote or advocate the ACC position, to be informational in nature, and to offer guidance of the ACC stance on pressing health policy issues;Review current evidence around policy issues such as overuse, underuse, and misuse of servicesReconcile conceptual framework with clinical efficacy and economic efficiency of careProvide overarching policy recommendationsCan be accessed at www.cardiosource.orgACC indicates American College of Cardiology; AHA, American Heart Association; and AUC, appropriate use criteria.Despite efforts to improve the value of cardiovascular care, there is a concern that significant overuse of cardiovascular services persists in the United States. The published evidence of overuse is necessarily incomplete for 2 key reasons. First, for many tests and treatments, there is a lack of agreement on the basic definition of overuse. The Institute of Medicine originally defined overuse as occurring when "a health care service is provided under circumstances in which its potential for harm exceeds the possible benefit".11 Others have since expanded the notion of overuse to describe services that maybe unscientific, redundant, excessive, duplicative, unnecessary, or harmful, leading to the concern that the vague conceptual definition of overuse has likely contributed to the lack of specific measures of overuse and initiatives to address overuse.12 An alternative, the economist's definition of overuse, relates to whether the service in question has lower benefits than its economic costs (ie, the opportunity cost of the resources used to produce it). Second, clinical notions of overuse almost always require consideration of bedside nuance that can be obtained only from rich clinical databases, which are often unavailable for research. It is thus not surprising to discover that the issue of overuse, despite its economic and policy importance, remains understudied. A literature review of >100 000 articles over 2 decades only yielded 172 articles that addressed the issue of overuse in health care in general.13 Of these, the most commonly studied services highlight 3 cardiovascular procedures: coronary angiography, coronary artery bypass grafting (CABG), and carotid endarterectomy. To reduce overuse, it is therefore necessary to establish sound measures that reliably identify overuse, evaluate its clinical and economic impact, and identify the causes and opportunities for reform and implement effective interventions.In this article, we review what is known about the overuse of cardiovascular services based on specific definitions (Table 2), describe major causal influences, and identify emerging efforts to address overuse.Table 2. Examples of Overuse in Cardiovascular Disease by Categories of EvidenceCategories of Evidence for OveruseExamplesUse deemed uncertain or inappropriate based on published AUCChan et al14 (2011) and Ko et al15 (2012): 12%–19% of elective PCIs performed were found to be of inappropriate indication, and another 20%–38% were of uncertain indication.Gibbons and Miller16 (2013): 11%–23% of nuclear myocardial perfusion imaging tests done were found to be inappropriate.Bhatia et al17 (2013) and Willens et al18 (2013): 13%–30% of stress echocardiography studies were shown to be of inappropriate indication.Use deviating from clinical guidelinesScheffield et al19 (2013): 3.75% patients without existing cardiac conditions or risk factors underwent cardiac stress testing before elective noncardiac, nonvascular surgery against ACC/AHA practice guidelines.Al-Khatib et al20 (2011): Non–evidence-based ICD implantations in the US.Indirect evidence of overuseWennberg and Birkmeyer21 (1999): The Dartmouth Atlas of Cardiovascular Health Care showed unwarranted geographic variation of cardiovascular procedure use: >10-fold imaging stress test, 8.5-fold for echocardiography, 8-fold for PCI, and 3-fold for CABG.Chen et al22 (2011) and Huffman and van Geertruyden23 (2011): Very low downstream coronary angiography and revascularization after stress myocardial perfusion imaging and stress echocardiography suggest very low pretest likelihood of disease in patient selection and potential overuse of noninvasive stress testing.Ladapo et al24 (2014): National survey data indicated significant trend of growth in cardiac stress testing with imaging; 34.6% of testing was found to be probably inappropriate based on patient risk profile, presenting symptoms, and visit diagnosis, with associated annual costs estimated to be $501 million.Evidence of overuse based on recommendations from the Choosing Wisely CampaignSchwartz et al25 (2014): Applying claims-based measures of 26 low-value care showed that 25%–42% of beneficiaries are affected by low-value care.Colla et al26 (2014): Overall 13% of Medicare low-risk patients received nonindicated cardiac testing, including ECGs (12.4%), stress tests (0.3%), and echocardiograms (0.25%).ACC indicates American College of Cardiology; AHA, American Heart Association; AUC appropriate use criteria; CABG, coronary artery bypass graft; ICD, implantable cardioverter-defibrillator; and PCI, percutaneous coronary intervention.Evidence of OveruseThe most commonly used standard to identify overuse of cardiovascular services is the RAND appropriateness method.27 Over the last 9 years, the American College of Cardiology and American Heart Association have developed 13 AUC documents on imaging, diagnosis, and treatment for cardiovascular disease.28 These AUC documents have been valuable as serving as standards for program accreditation of echocardiography, nuclear imaging or cardiac catheterization laboratories, and pay-for-performance contracts from payers in an effort to measure overuse. Applying AUC for percutaneous coronary intervention (PCI)29 in large clinical databases14-15 showed that 12% to 19% of elective PCIs performed were found to be of inappropriate indication and another 20% to 38% were of uncertain indication. Revascularization in patients with uncertain or inappropriate indications was shown not to result in an improvement in clinical outcomes such as subsequent death or acute coronary syndrome. When AUC were applied for nuclear myocardial perfusion imaging, the rate of inappropriate studies was found to range from 11% to 23%.16 Thirteen percent to 30% of stress echocardiography has been shown to be inappropriate, depending on whether the studies were done in the inpatient or outpatient setting.17-18 Evidence of cardiac stress testing overuse has been shown with significant temporal growth and a shift toward stress test with imaging. Using data from the National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey, Ladapo et al24 recently found 34.6% of cardiac stress tests with imaging to be probably inappropriate, with associated annual costs and harms of $501 million and 491 future cases of cancer. In this study, clinical survey data from the National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey, including patient cardiac risk profile, clinical symptoms, and physician diagnosis information, were used to designate appropriateness of testing. The economic and health impacts of potential overuse were estimated from average national Medicare reimbursement and a previous study of the effect of ionizing radiation.These and other AUC documents have been widely accepted as the gold standard to evaluate for appropriateness of use, providing evidence of underuse and overuse. However, it is important to recognize that the appropriateness method was not intended for practice or physician profiling; the clinical scenarios described in these documents are by no means exhaustive to capture the myriad case-specific clinical details and exceptions in real-world practice. The limitation of AUC for profiling overuse is underscored by the substantial plurality of cases that are classified as uncertain. More recently, the terminology for AUC has shifted from "appropriate, uncertain, or inappropriate" criteria to "appropriate, maybe appropriate, or rarely appropriate."The second major category of overuse evidence comes from examination of the deviation of clinical practice from published guidelines or clinical trial results. For example, the 2007 American College of Cardiology/American Heart Association guidelines on perioperative evaluation for patients undergoing noncardiovascular surgery recommend that patients without existing cardiac conditions or risk factors should not undergo cardiac stress testing before elective noncardiac, nonvascular surgery.30 However, a Medicare inpatient claims analysis showed that 3.75% of such patients received nuclear stress testing, an expensive imaging test with significant radiation, cost, and potential for a downstream invasive procedure.19 In this analysis, there was a significant temporal trend of increase of overuse, and hospital size and high Medicare use region were also associated with higher use, confirming the theory of supply-induced demand described by the Dartmouth Atlas researchers nearly 2 decades ago.31 Likewise, non–evidence-based use of implantable cardioverter-defibrillators has been shown to have significant geographic variation and persistent overuse despite the publication of clinical guidelines.20Besides inappropriate use of diagnostic tests and interventions, other indirect evidence also exists to suggest overuse of cardiovascular services. An analysis of a large private insurance company database showed very low downstream coronary angiography and revascularization after stress myocardial perfusion imaging and stress echocardiography, suggesting very low pretest likelihood of disease in patient selection and implying potential overuse of noninvasive stress testing.22-23 Additionally, the >10-fold regional variation for imaging stress test use, 8.5-fold variation for echocardiography, 8-fold variation for PCI, and 3-fold variation for CABG shown in The Dartmouth Atlas of Cardiovascular Health Care21 is also felt to be reflective of unwarranted variation and suggest potential overuse of supply-sensitive care (Figure). On the other hand, cross-national studies of cardiac service use show that overuse is not linked to overall service volume,32 suggesting that blunt policies to simply restrict service volume may cause underuse even while reducing overuse. Although the geographic variation data on cardiac imaging studies, PCI, and CABG cannot be used to specifically measure the extent of overuse, the American College of Cardiology and American Heart Association have subsequently developed evidence-based practice guidelines to standardize care and to reduce variation.33Download figureDownload PowerPointFigure. Rates of imaging stress testing among hospital referral regions (1996). Rates of imaging stress testing ranged from ≈10 per 1000 enrollees to almost 120 after adjustment for differences in population age, sex, and race. Each point represents 1 of the 306 hospital referral regions in the United States. Reprinted from The Dartmouth Atlas of Cardiovascular Healthcare with permission from the publisher.21In addition, researchers have begun to develop direct measures of overuse of healthcare services, in part flowing from the recommendations from the American Board of Internal Medicine Foundation Choosing Wisely Campaign. For example, Schwartz et al25 developed claims-based measures of 26 recommendations related to avoiding low-value care and applied them to 2009 Medicare claims. Using either a more sensitive or more specific algorithm, they estimated that 42% or 25% of beneficiaries are affected by low-value care, accounting for 2.7% or 0.6% total claims, respectively.25 Colla et al26 stratified Medicare patients using claims data into low- versus high-cardiovascular-risk categories and identified that 13% (driven predominantly by ECGs) of low-risk patients received nonindicated, low-value cardiovascular testing. Significant regional variation of overuse of cardiovascular testing was also shown.Although research on overuse based on claims data clearly represents an important step toward more quantitative analysis of overuse, translating clinical guidelines into claims analysis algorithms could result in vital clinical information being "lost in translation." Further research is needed to develop and implement measures that can be applied to clinical data captured in registries and electronic medical records. Moreover, such clinical measures could be compared with claims-based estimates for the same populations to ascertain the validity (or lack thereof) of claims-based measurement of overuse.Causes of OveruseResearchers have identified a wide array of forces that contribute to the use of low-value services, all of which are likely implicated in cardiology. Conceptually, these forces can be categorized into clinical, financial, cultural, or legal drivers of overuse (Table 3). Clinical drivers include a traditional professional training style that promotes the development of a broad and exhaustive differential diagnosis, paucity of information about the appropriate use of alternative tests and treatments, lack of systematic comparative effectiveness research (CER) to support evidence-based treatment protocols, and insufficient patient education about the potential harm from overuse. Even when evidence exists, it is slow to disseminate among physicians and may not lead to change in practice patterns because of status quo bias and individual, team and organizational barriers to learning.34Table 3. Overuse in Cardiovascular Medicine: Causes, Evidence, and Opportunities for ReformCauses of OveruseSummary of EvidenceOpportunities for ReformClinical driversProviders:Lack of adequate evidenceDelay in knowledge disseminationIneffective provider behavior changePatients:Insufficient data for potential harm and side effectsLack of SDMComparative effectiveness researchMultipronged knowledge disseminationPerformance measurement, feedback, and continuous quality improvementBuilt-in clinical decision support in electronic medical record systemPatient education and SDMFinancial driversProviders:Fee-for-service reimbursementCompensation structure to reward productionPatients:Generous insurance coverageIndustry:Technological and pharmaceutical advances with high profit marginDirect-to-consumer marketingPayment incentive for providers to reward value: bundled payment, risk/cost sharing, pay for performance, global paymentPrice transparency for both providers and patientsValue-based insurance design to promote patient choice of high-value servicesPrior authorization, case reviewCultural driversProviders:Procedural and intervention focusedPatients:Action over inactionBetter clinical evidence and educationLegal driversProviders:Malpractice concerns and defensive medicineTort reformCER indicates comparative effectiveness research; and SDM, shared decision making.Financial drivers are likely to be an important impetus for overuse. Of particular concern is fee-for-service payment, which provides a financial incentive to deliver tests and treatments regardless of patient benefit, rewarding service intensity and volume instead of value.35 Similarly, generous insurance coverage encourages patients to seek more healthcare services and may reduce physician concerns about financial burden from overtreatment. Direct financial incentives are clearly not the only factors that drive overuse, however, as evidenced by the geographic variation in overuse as measured by Medicare claims despite a single benefit design and only modest geographic differences in reimbursement rates.Commentators have pointed to a culture in medicine that promotes action over inaction, which may be reinforced by physician training style, malpractice concerns, and local norms.36 A Dartmouth Atlas analysis of 23 academic training programs around the country showed a correlation between the supply of cardiologists and visits to cardiologists per 1000 Medicare enrollees among hospital referral regions, a 3.2-fold difference in CABG rate, and a 2.6-fold difference in PCI rate across the 23 top training programs.37 A similar culture appears to be prevalent among American consumers, who appear to be more willing than their European counterparts to support investments in technological innovation in health care.38 Direct-to-consumer marketing of pharmaceuticals, devices, and other technologies almost surely contributes to excessive use of these products, in keeping with an inflated impression of their value. The counterargument in support of a more conservative management strategy, that sometimes less is better and safer, often remains ignored or suspected of alternative motives, especially when endorsed by insurance companies.Despite the fact that performance data such as acute myocardial infarction and CABG outcomes have been made available to the public, there is no evidence of a consistent effect on consumer behavior.39 Even motivated consumers of health care may be constrained by "bounded rationality" because individuals make decisions limited by the cost of the information (Do they have the time and access to reliable quality information at the time of an acute myocardial infarction to choose the best hospital?) and cognitive limitations (Do they have the knowledge and ability to understand complex performance data presented to make the right choice?).Finally, physicians report (and researchers have observed in practice) that the specter of medical liability causes them to order additional tests and treatments. Although the magnitude of increased use that results from malpractice concerns has been estimated as modest overall (≈2% without differentiating low-value from high-value care), it appears to play a disproportionate role in physicians' perceptions of cost drivers.40,41Given that systematic evidence on overuse is lacking, inferences about the relative contribution of each of these causes of overuse are necessarily speculative. Moreover, even at a conceptual level, it is challenging to disentangle the causes of overuse. For example, is the culture of medicine that when in doubt action is preferable to inaction an independent driver of overuse or a consequence of the reimbursement model? As large database studies of overuse that leverage advances in data and measurement grow in number, more sophisticated causal analyses will be possible, and these may set the stage for identifying policy priorities.Opportunities for ReformEfforts to address overuse of cardiovascular services may take a variety of forms and address 1 or more of the underlying causes of overuse, as outlined in Table 3. We highlight a few areas in which reforms appear to be gaining traction. In some cases, these initiatives specifically target cardiovascular care, but payers and policy makers have generally approached overuse in an undifferentiated manner with the goal of cost control.42 Conceptually, interventions to reduce overuse of cardiovascular services can be organized in terms of whether they operate through patient behavior, through physician behavior, or by direct regulation of health system capacity.Although physicians make the majority of complex healthcare decisions, patients are increasingly willing and able to share in decision making for their care. In the case of complex choices with significant health consequences, patients need accessible information and support to play a meaningful role in selecting high-value care. Decision aids and a framework for SDM have been developed and used effectively in a number of conditions, including prostate and breast cancer treatment and treatment of low back pain. SDM is becoming a major priority of research for cardiovascular medicine in terms of its conceptual design, its data collection, and the measurement of its effectiveness.43 Thus far, there is modest evidence that SDM reduces the intensity and cost of care.44 Limitations of this approach include lack of time or motivation to engage in the process on the part of physicians, the limited set of clinical decisions for which decision aids have been developed, and the plausible set of comparatively effective, preference-sensitive choices for which SDM might enhance patient compliance and clinical outcome.In addition to engaging patients individually in SDM, patient information could be a lever for reducing overuse at a population level. Recent efforts by Consumer Reports in partnership with the American Board of Internal Medicine Foundation have produced accessible information for patients to inform them about services of questionable value, including routine cardiac stress testing.45Patient financial incentives, particularly in the form of higher cost sharing, may also affect overuse. Copayments and deductibles have been shown to reduce the use of all types of healthcare services, but evidence shows that patients do not have discriminative ability to ration their own care according to value; that is, they eliminate use without differentiating appropriate from inappropriate services.46 Targeted cost sharing, known as value-based insurance design, has been proposed as a response to the finding that blunt patient cost sharing eliminates both necessary and unnecessary care. However, value-based insurance design has been used largely to promote high-value services by lowering or eliminating copayments and to the best of our knowledge has yet to be used as a tool for reducing overuse, particularly for cardiac care, because doing so might be considered ethically contentious.47Given the role of physicians in healthcare decision making, it is not surprising that the focus of policy to reduce overuse of low-value services has been in the realm of evidence base, performance measurement, decision support, and payment incentives. Providing practicing physicians with valid information about the comparative effectiveness of screening, diagnosis, and treatment is essential for improving the value of care. As noted earlier, cardiovascular medicine has been fertile ground for such evidence development. However, adoption of new information about treatment effectiveness can be slow. Likewise, cardiology has been at the forefront of guideline development with evidence- and consensus-based best practices identified not only for treating specific conditions but also for avoiding overuse. Dissemination of guidelines, however, is no guarantee that the guidelines will be followed and alone is a weak tool for changing behavior, particularly when countervailing financial incentives are strong, as is typically the case for overuse. Although efforts at refining the AUC have resulted in decreasing the amount of unclassifiable studies based on AUC criteria, the percentage of inappropriate studies has not been shown to improve spontaneously over time or simply with traditional strategies of physician education.48,49A number of additional tactics to reduce overuse could be implemented on the basis of the opportunities for reform identified. First, when possible, the electronic medical record should alert providers when circumstances of potential overuse are noted. Decision support that flags a possible overuse situation and makes preferred, appropriate care the default option for order entry may be effective.50 In addition, didactic teaching, when combined with an electronic pocket card with AUC tips, individualized e-mail performance feedback, and smartphone application use, may help to improve appropriate referral to cardiac imaging tests.51,52Second, price transparency for both providers and patients may help foster a shared sense of stewardship and awareness of at least the immediate costs of testing and treatment. Despite compelling rationale for price transparency to control unsustainable healthcare cost growth and rare partisan agreement on the necessity for price transparency, barriers and challenges remain. With increasing cost sharing for both providers and patients of health care, there is increasing demand for price transparency from both the supplier and consumer sides of health care. Furthermore, the generational shift of healthcare consumers and technological innovations are making price transparency and consumer choice increasingly possible. Insurance companies such as Aetna, Health Care Service Corp, and UnitedHealthcare are making price and out-of-pocket cost information available to their members. Private firms such as Castlight, Change Healthcare, and Cost of Care are using social media, Web, and smartphone-based applications to provide price information to both providers and consumers of health care.53,54Finally, payment incentives for the ordering physician may be altered to address overuse. This can be done within the incremental changes to the current payment system by cutting the level of fee-for-service reimbursement, eliminating coverage for services in certain conditions, or introducing bonuses or penalties related to rates of testing or treatment. More significant payment reform appears to be taking hold, however, and overuse may be affected by a new paradigm in which providers are held accountable for the total cost o
Referência(s)