The authors report on a conceptual "specialty ACO" design in cardiology in which cardiology-specific historic spending targets are created to capture cardiovascular disease—related reimbursement and a quality measurement system is introduced.
ABSTRACTObjectives: Engaging specialists in accountable care organizations (ACOs) may make them more responsive to pressures to lower costs and raise quality. This paper introduces a novel accountable care design in cardiology.
Study Design: Preliminary study using baseline data.
Methods: The Accelerating Clinical Transformation for Creating Value and Controlling Cost in Cardiology concept study involved providers employed by the Providence Medical Group, Oregon. First, using claims data from 2009 through 2011, we created a historic budget to capture cardiovascular disease (CVD)-related costs for attributed patients on a per patient per year basis. Second, we introduced a validated quality metric, the Clinical Performance and Value vignette, to a sample of cardiology providers to examine clinical practice variation in treating coronary heart disease (CHD), coronary heart failure (CHF), and atrial fibrillation (AF). Lastly, we analyzed reimbursement claims paid for CHD, CHF, and AF, and forecasted potential cost savings from reductions in clinical variation.
Results: Examining historic costs, we found they were stable over time, but variable by provider and disease. Quality scores, measured against evidence-based cardiology guidelines, ranged from 48.9% to 85.4% (mean = 66.8%; SD = 5.4%), and the prevalence of unnecessary testing was 46% in CHD, 71% in CHF, and 30% in AF. We project that reducing unnecessary care by 15% to 25% would yield $200,000 to $498,000 in savings ($50-$83 per patient visit) annually. And, if the top 10% of providers as determined by CVD-related costs reduced their costs by 25%, savings would be an additional $283,512 per year.
Conclusions: This accountable care design framework is timely for cardiology and could be applied for other specialty conditions, such as cancer.
Am J Manag Care. 2015;21(4):284-292
Specialty care accounts for the vast majority of healthcare spending and its use is a major driver of cost over time. The United States spends the most on cardiovascular disease (CVD)—$273 billion in total direct costs in 2010.1 Cardiologists are leaders in adopting evidence-based medicine to improve clinical quality and practice effectiveness, thereby improving clinical outcomes,2 patient satisfaction,3 mortality, and morbidity.4,5 However, despite this impressive, professional society-endorsed evidence base and the equally impressive clinical and economic utility of adhering to the cardiology evidence, a “translational gap” between evidence and practice remains.6,7 Notably, there is widespread practice variation in cardiology that presents an opportunity to improve quality and rein in costs. To improve care and slow cost growth, payers are increasingly turning to using shared savings models, which introduce more aggressive risk-sharing with incentives for quality.8,9 Shared savings involve an ex ante discussion with a payer and agreement on expected spending—usually historically determined—weighed against actual savings with some formula for sharing the difference if there is indeed any savings.10 Shared savings arrangements have been effective at lowering input costs such as drugs, implants, and imaging studies from the standpoint of hospitals and facilities,11,12 yet the overall impact of cost savings to the healthcare system has not been clearly demonstrated.13
Payments based upon a condition and then linked to shared savings for specialists and primary care providers, however, could encourage managing care across place and time so that profitability can be linked to both efficient delivery of services for an acute clinical episode, and a shift in utilization to lower-cost venues such as the clinic or even the home. Ultimately, shared savings may have their greatest impact if they are condition-focused and reduce utilization by limiting exacerbations requiring readmission, diagnostic testing, or recurrent treatment.
It is surprising that although specialists are at the forefront of evidenced-based practice initiatives, along with being important cost centers in merging payment models, they are not the basis for organizing accountable care organizations (ACOs); to date, efforts to control costs and manage care have been centered on primary care practice. However, as the number of complicated patients grows—particularly those with chronic conditions such as CVD who require specialty care and often multi-specialty care—arguably a shift of focus is required to organize care around the patient and condition rather than the provider and the facility. Yet, there are few examples of how specialists can be engaged with physician leadership and share accountability and savings with all other providers caring for a particular condition.
A specialty ACO requires that the risk level and payment for providers be set in a practical and equitable fashion that does not compromise quality. While at the system level a global budget may exist for an ACO, importantly, the responsibility for cost control is not totally borne by individual physicians. The portion of shared savings “at risk” for physicians, and thus the incentive, would ideally be for quality based upon adherence to guidelines describing which best practices are linked to better clinical outcomes.14 An opportunity to test this concept might exist in the treatment of CVD, because currently, more than 60% of US cardiologists are employed or are in negotiation for employment.15
We argue that the receipt of shared savings payments must be conditional upon meeting a series of practice-based quality standards. The challenge has been how to determine (and raise) the quality thresholds. A newer and increasingly adopted method for measuring and improving quality through serial measurement and feedback is the Clinical Performance and Value (CPV) vignette, an open-ended simulated case in which a provider is asked to care for a presenting patient and is then scored based on practice guidelines. CPV vignettes are the basis for qualifying for shared savings for cardiology care reported here.
The Accelerating Clinical Transformation for Creating Value and Controlling Cost in Cardiology (ACT-C3) Project is a novel accountable care design in cardiology, centered on CPV vignettes, in which our aim was to introduce a methodology to determine historic costs and identify areas of savings, all while maintaining a focus on evidence-based quality of care. The design requires a measurement of clinical practice and the establishment of a quality threshold in order to participate in any shared savings.
The Accelerating Clinical Transformation for Creating Value and Controlling Cost in Cardiology (ACT-C3) Project
Sponsored by Providence Health & Services—a nonprofit healthcare system consisting of a network of health plans, 27 nonprofit hospitals, 214 physician clinics, and a broad range of clinical programs and affiliated services—researchers at the University of California (San Francisco) and from the Providence Medical Group formed the Accelerating Clinical Transformation for Creating Value and Controlling Cost in Cardiology (ACT-C3) Project to design a new model for the care of patients with CVD.
The cardiologists, hospitalists, and nurse practitioners in the project were employed by the Providence Medical Group (PMG), the physician organization that centralizes operational infrastructure and promotes clinical best practices. PMG Oregon has the largest cardiovascular medicine group in Portland, with the PMG cardiovascular medicine division employing about 34 cardiovascular physicians and 20 advanced practice providers.
ACT-C3 completed pilot investigations in 2 general areas: business (costs) and clinical (quality). For the examination of costs to the payer, we created a historic budget for CVD care—named “CardioGrouper”—that would be used to determine the total spending threshold for a shared savings agreement between a payer and the medical group; savings would accrue to the whole medical group. Using the Clinical Classification Software (CCS) from the Agency for Healthcare Research & Quality (AHRQ), we pulled specific CVD diagnosis codes and then used patient-level claims data from several plans to capture all CVD-related claims for attributable patients on a per patient per year basis. We examined these data for a 3-year span from 2009 to 2011 (described in Part 1 below).
The clinical activities centered on the CPV vignettes, a measure used to examine quality (described in Part 2 below) by determining correct scores for clinical scenarios based on American College of Cardiology (ACC) guidelines. Our motivation was to make the CPV vignette score the basis for qualifying into the shared savings program via achieving a minimum threshold of quality performance. By incorporating this quality requirement into our design model, we were assuming that an increased use of evidence-based practice standards, as measured by CPV vignette scores, would lead to better outcomes and costs. We then analyzed clinical variation alongside historic spending identified by our CardioGrouper to model potential cost savings opportunities (described in Part 3 below). The ultimate goal of this care design project is to show how these opportunities can lead to new payer-provider contracts, wherein savings would be shared among the purchasers or payers, cardiology providers, and the delivery system.
Part 1: Establishing the historic budget—looking at reimbursement over time to establish a spending target for the CVD service line. Costs incurred to payers as claims paid were determined specifically for CVD on a per patient per year basis—we called this the CardioGrouper. We defined “attributable patients” as adult patients seen in the Portland service area with unique member identification for Providence Health Plan. Patients were stratified into those with and without a CVD diagnosis using the list of CVD-specific diagnosis codes from AHRQ CCS developed by the Healthcare Cost and Utilization Project for International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), CCS-7 (diseases of the circulatory system, excluding 7.5; diseases of veins and lymphatics). Patients thus identified were then stratified into low versus high risk, defined by whether the patient had been seen by a cardiologist with a new visit code during the claims year. We defined risk this way because a visit to a cardiologist for cardiovascular disease is a reasonable threshold for signaling that additional expert care was required.
We determined the historic budget (ie, historic reimbursed charges) by taking all attributed patients and aggregating all their CVD-related patient-level claims paid. To capture the total service line claims paid for each attributed patient, we took common CVD procedural lists from existing cardiology practice groups and reconciled them against CCS Services and Procedural categories relevant to CVD (13,820 Current Procedural Terminology codes in total, used to identify the patients). To capture CVD-specific inpatient claims paid, we used a select list of 59 Major Diagnostic Category 5, cardiology-specific diagnostic-related groups. In total, all CVD-related claims associated with inpatient facility, outpatient procedure, or professional services were included. Using 2011 Providence Medicare Advantage plan data, we then analyzed the CVD cost—defined as the claims paid to the provider—for patients with all CVD diagnoses and by disease for our 3 target clinical conditions: 1) coronary heart disease (CHD), 2) coronary heart failure (CHF), and 3) paroxysmal atrial fibrillation (AF).
We analyzed the cost to the payer (claims paid) by condition for the 2 main cardiology groups employed by the Providence Medical Group, which we refer to as Group A and Group B (the groups are comparable in number of clinical staff, patient load, and referral base), to ascertain if the costs were stable over time. In addition, we stratified claims paid by patient risk, where low risk is defined as having had no new cardiology visit in the year of analysis, and high risk as having had a cardiology new-patient visit in that year. We further examined test utilization by looking at the number of cardiovascular tests per patient per year. We compared utilization (15 different CVD tests) between Groups A and B for our 3 different target conditions. We used a Bonferroni approach to adjust for multiple testing and established a significance level of P <.0001.
Part 2: Measuring quality to capture details of practice variation. The ACT-C3 project used CPV vignettes, which are a case mix—adjusted method to measure performance, identify practice variation, and affect behavioral change that would ultimately lead to reducing unnecessary utilization and controlling cost of care.16-19 The CPV vignette measurement and feedback system consists of open-ended simulated cases in which a provider is asked to care for a patient presenting with 1 of the 3 conditions: CHD, CHF, or paroxysmal AF (see eAppendix, available at www.ajmc.com). We used clinical practice guidelines developed by the ACC to prepare the performance criteria for the 3 different cases. The CPV vignettes were completed by a group of 24 providers—employed by Providence Medical Group—consisting of advanced practice providers (nurse practitioners and physician assistants), physician hospitalists, and physician cardiologists sampled from the provider roster of the PMG. The 24 participating providers were approached to participate in the study and each one agreed. All 3 CPVs were taken at one time so there was no opportunity to improve one’s test-taking ability and thereby one's score just by getting better at testing. There was no subsequent follow-up testing in the findings reported here. Scores generated from the cases were reported as the percentage of criteria correct divided by the total number of criteria for each case. The scoring of the vignettes was completed by a team of trained physicians who were blinded to the identities of the providers taking the vignettes.
Part 3: Projecting cost savings. As in Part 1, using 2011 claims data from the Providence Medicare Advantage population, we analyzed the claims paid for the 3 clinical cases of interest: CHD, CHF, and AF. Costs (to the payer) were attributed to an individual cardiologist based on the evaluation and management new visit code with 1 of the 3 diagnoses and then aggregated on a per patient per year basis. We then identified providers who generated costs above the 90th percentile of charges and applied various cost reduction scenarios to determine the potential savings if these providers were to bring their costs down.
The scenarios used the quality and variation data from the CPV vignettes, such as the prevalence of unnecessary testing (defined above as ordering a test whose result would not change clinical management, as identified in the ACC guidelines). To assign costs for each ordered test for each clinician, we used local charge rates. For this analysis (in Part 3) of the potential savings from reduced clinical variation, we relied on the peer-reviewed literature, noting the emerging body of literature that links evidence-based practice to fewer unnecessary tests and treatments.20-23 Using this literature for our cost savings projection, we assumed a 15% to 25% reduction in unnecessary testing due to improvements in adherence to cardiology practice guidelines resulting from the measurement and feedback program. We then applied this value (15%-25%) to project the impact of reduced unnecessary testing on overall costs to the payer. Lastly, we used practice data to estimate a range of 4000 to 6000 new patients seen by both cardiology groups annually.
Part 1: Establishing the historic budget—looking at reimbursement over time to establish a spending target for the CVD service line. Using the CardioGrouper, we found that for the CVD service line, costs to the payer were stable over time (2-tailed independent t test comparing all costs in 2009 vs 2010, P = .46; 2010 vs 2011, P = .06; 2009 vs 2011, P = .95) but were variable by patient risk (2-tailed independent t test comparing low vs high risk for years 2009, 2010, and 2011 data all resulted in P <.0001) (reimbursements graphed in ).
When we looked at the per patient utilization of cardiovascular tests for each specific diagnosis by group, we noted no significant differences in the frequency of patient visits (new or follow-up visits per 100 patients per year) between the practice groups. We found test utilization between groups for the exercise treadmill test (2-tailed independent t test comparing Group A with Group B; P <.0001) differed significantly after adjusting for physician case load ().
Part 2: Measuring quality to capture details of practice variation. We measured the quality of care for 24 CVD providers. Eleven were cardiologists, 9 were hospitalists, and 4 were advanced practice providers who exclusively cared for patients with CVD. Of these providers, 58% were men; 75% had 6 or more years of experience practicing cardiology, 91% saw more than 20 CVD patients per week, and 68% had more than 80% of their practice in CVD. The overall quality of care, as measured by CPV scores for the 3 cases, revealed that conformity to ACC guidelines varied, ranging from 48.9% to 85.4% correct with a mean of 66.8% and a standard deviation of 5.4% (). Scores were highest for history and physical exam domains, whereas ordering labs/imaging, diagnosis, and treatment scores were lowest overall. Scores by case type also varied and are shown in . There was a substantial amount of unnecessary testing for each vignette: 46% of the CHD, 71% of the CHF, and 30% of the AF cases had unnecessary testing.
Part 3: Projecting cost savings. Looking at the utilization patterns of individual physicians for the 3 CVD diagnoses and the variation found in the case mix—adjusted CPV cases, we found significant inter-provider variability in cost and practice, respectively. We first calculated that if the top 10% of providers, as determined by CVDrelated charges reimbursed (CardioGrouper per patient per year), reduced their charges by 25%, savings would be $283,512 per year. Potential cost savings calculated using the variation determined by the CPV results yielded similar estimates. First, based on provider claims data and the prevalence of unnecessary testing established by the vignettes, we estimated that providers incur between $233 and $482 (average = $331) per patient per year in unnecessary testing across the 3 disease types. Reducing unnecessary care by 15% to 25% yields between $200,000 and $498,000 savings (or $50-$83 per patient visit) annually.
With the ACT-C3 project, we identified provider practice variation using the CPV vignettes as a measurement tool, then projected potential savings opportunity for the payer using condition-specific historic reimbursement. If referring providers and cardiovascular specialists are to share accountability in performance, utilization, and costs, the payment incentive must be shifted away fromfee-for-service. We propose here that setting a cardiology-specific spending target for a payer for a given patient population is feasible. This spending target could then be used for negotiation between payer and delivery system and providers toward a shared savings plan. The program envisions collaborative management of patients with shared savings conditioned upon quality measured by CPV vignette-standardized cases. Under this arrangement, the focus of care management is on the quality of care and on greater efficiency (such as the reduction in unnecessary testing) and provides the basis for savings to be shared between the payer and providers, including cardiologists.
We propose that a blended payment model in specialty care—one that combines an at-risk shared savings payment and is conditioned on quality—has the potential to reduce costs by improving clinical practice patterns. We think it is vital to the success of new payment models to have goal congruence between the reimbursement scheme (transaction between payer and provider groups) and the compensation model (how a provider group motivates its individual providers). Given the importance of cardiology, payers might find a “cardiology ACO” or “specialty ACO” to be a useful way to initiate a shared savings program. For an existing ACO or a delivery system under global contract, a specialty ACO might want to create a business line sub-ACO (or subcontracts) to accelerate care delivery transformation and control specialty cost.
A focus on quality in an ACO should require an emphasis on patient care and care processes, beyond organizational infrastructure.24 Current ACO models may be workable in the short term, but we think they are an unsustainable business model due to their lack of sufficient focus on specialty cost and its reimbursement. Without reengineering the care process and payment incentives, delivery systems are increasingly realizing that a large body of employed providers contributes to rising financial liabilities or incentives to increase expensive specialty service volume.
An estimated 227 provider organizations have established ACO contracts with payers, be it Medicare, Medicaid, or private.25 To our knowledge, ours is the first to propose a specialty-based ACO. With this pilot study, we argue that clinical variation in practice is what we ought to target, with the ultimate goal of improving quality, which thereby may have positive benefits on reducing costs. We do, however, recognize that the directional association between quality and cost cannot be taken for granted.26 In reality, higher quality can correct for underutilization and thus may raise costs, just as it decreases overutilization and eliminates unnecessary care thus decreasing costs.18 Cardiology patients are often the costliest ones, requiring long-term monitoring and close clinical management.
There are important challenges in our model. Practice groups and delivery systems in general contract with a myriad of payers—each with different contracts and their own conversion factors, and many of which are not costor quality-based or patient centered. Thus, novel programs such as this one—which require reorienting care delivery to focus on coordinating evidence-based care around clinical conditions rather than procedures and visits, gathering and monitoring clinical quality data—require commitment from a “lead payer,” which was the Providence Health Plan in this study. The challenge for the longer term is that other payers will benefit from this care transformation but are not obligated to share in the savings.
There were a number of other obstacles to implementation. In this preliminary concept study, there appeared to be a great deal of administrative uncertainty to changing care delivery processes in anticipation of the tipping point of payment models. Care transformation relies heavily on physician leadership and payer data. In some systems, this degree of innovation and the aspirations to raise quality and lower cost may prove impractical, in our experience, because leadership is either not adequately trained or not sufficiently aware of the interplay between payers, providers, and clinical quality. There are also contracting challenges for the payers who are being asked to layer a specialty-care shared saving program on top of existing contracts. There are also challenges on the provider side, as specialists and primary care providers must find an equitable and efficient way to share resulting savings according to the effort and contributions of each. This challenge is mitigated in a system that employs specialists and generalists, and in our opinion should not be adjudicated by the payer.
First, we relied on Providence Medicare Advantage (MA) claims data for our analysis of costs (Parts 1 and 3), and thus our generalizability may be limited to the extent that the conditions around an MA ACO are different from those of the Medicare Shared Savings Plans that are all in traditional fee-for-service Medicare. With only baseline data on quality (the CPVs) and costs, we are not able to exactly measure how costs change with increased adherence to guidelines; we assumed the link between evidenced-based care and improved outcomes based on the literature,27-34 and applied an assumed value for our cost projections in Part 3. While the results in Part 3 are just projections, we believe such projections remain an important way for any organization that seeks to create a shared savings program based on care standardization (ie, improving evidence-based practice) to understand the implications of its work, and offer a compelling and easy-to-understand view of the monetary opportunities when they take this on. Lastly, a technical challenge that all programs will have to overcome is the lag between historical cost data and, in this study, the CPV data. We propose further study to obtain serial measurement data. As we link costs to adherence, we will also want to include clinical outcomes, such as fewer hospital days, fewer readmissions, and lower mortality rates.
Our study is a first step in demonstrating that in the era of heightened accountability for costs and outcomes, a focus on the patient and quality is key and requires bringing all the players to the table, specialists and nonspecialists alike. We offer the potential of a new accountability model centered on particular clinical conditions where savings are shared across all providers treating the condition and predicated on a case mix—adjusted, validated measure of quality.Author Affiliations: QURE Healthcare (JWP, RS), San Rafael, CA; Global Health Sciences, University of California (JWP), San Francisco, CA; Providence Heart Clinic (XH), Portland, OR; Harvard T. H. Chan School of Public Health (MR), Boston, MA.
Source of Funding: The ACT-C3 study, reported in this paper, was funded by Providence Health & Services (PH&S) through a grant to the University of California, San Francisco and QURE Healthcare. The design and conduct of the study, data collection, and management were conducted under this grant. The analysis, interpretation of the data, and preparation of this manuscript were done by the authors outside of the funding. Dr Peabody had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Author Disclosures: Dr Peabody is the owner and president of QURE Healthcare, which holds the trademark for the CPV vignettes measurement tool referenced in this paper. Dr Shimkhada is also employed by QURE Healthcare. Dr Huang is an employee of Providence Health & Services. Dr Rosenthal reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (XH, JWP, MR, RS); acquisition of data (XH); analysis and interpretation of data (XH, JWP, MR, RS); drafting of the manuscript (XH, JWP, RS); critical revision of the manuscript for important intellectual content (XH, JWP, MR, RS); statistical analysis (XH); provision of patients or study materials (XH); obtaining funding (XH); administrative, technical, or logistic support (JWP, RS).
Address correspondence to: John W. Peabody, MD, PhD, QURE Healthcare, 1000 4th St, San Rafael, CA 94901. E-mail: email@example.com.
1. Heidenreich PA, Trogdon JG, Khavjou OA, et al. American Heart Association Advocacy Coordinating Committee; Stroke Council; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; Council on Epidemiology and Prevention; Council on Arteriosclerosis; Thrombosis and Vascular Biology; Council on Cardiopulmonary; Critical Care; Perioperative and Resuscitation; Council on Cardiovascular Nursing; Council on the Kidney in Cardiovascular Disease; Council on Cardiovascular Surgery and Anesthesia, and Interdisciplinary Council on Quality of Care and Outcomes Research. Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation. 2011;123(8):933-944.
2. Cacoub PP, Zeymer U, Limbourg T, et al; REACH Registry Investigators. Effects of adherence to guidelines for the control of major cardiovascular risk factors on outcomes in the Reduction of Atherothrombosis for Continued Health (REACH) Registry Europe. Heart. 2011;97(8):660-667.
3. Weingarten SR, Stone E, Green A, et al. A study of patient satisfaction and adherence to preventive care practice guidelines. Am J Med. 1995;99(6):590-596.
4. Fox KA, Steg PG, Eagle KA, et al; GRACE Investigators. Decline in rates of death and heart failure in acute coronary syndromes, 1999-2006. JAMA. 2007;297(17):1892-1900.
5. Peterson ED, Roe MT, Mulgund J, et al. Association between hospital process performance and outcomes among patients with acute coronary syndromes. JAMA. 2006;295(16):1912-1920.
6. Bates DW, Kuperman GJ, Wang S, et al. Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality. J Am Med Inform Assoc. 2003;10(6):523-530.
7. Kahn R, Robertson RM, Smith R, Eddy D. The impact of prevention on reducing the burden of cardiovascular disease. Circulation. 2008;118(5):576-585.
8. Weissman JS, Bailit M, D’Andrea G, Rosenthal MB. The design and application of shared savings programs: lessons from early adopters. Health Aff (Millwood). 2012;31(9):1959-1968.
9. Delbanco SF, Martin Anderson K, Eikel Major C, Kiser MB, Wammack Toner B; Catalyst for Payment Reform; Booz Allen Hamilton. Promising payment reform: risk-sharing with accountable care organizations. The Commonwealth Fund website. http://www.commonwealthfund.org/~/media/files/publications/fund-report/2011/jul/1530delbancopromisingpaymentreformrisksharing-2.pdf. Published July, 2011. Accessed March 17, 2015.
10. Newman D; Congressional Research Service. Accountable care organizations and the Medicare Shared Savings Program. CRS Report for Congress. http://www.hklaw.com/files/Uploads/Documents/HealthcareReformCenter/ACOandMSSMemorandumApr252011.pdf. Published April 25, 2011. Accessed March 17, 2015.
11. Bailit M, Hughes C; Bailit Health Purchasing. Key design elements of shared-savings payment arrangements [issue brief]. New York, NY: The Commonwealth Fund; August 2011.
12. Cote B. Implementation of pathways: payers to share savings with compliant groups. OBR [Oncology Business Review] On-conomy website. http://obroncology.com/documents/OBR_sept_Pathways.pdf. Published September 2010. Accessed March 17, 2015.
13. US Senate Finance Committee. Empiric Data Lacking to Support Claims of Savings With Group Purchasing Organizations, September 2010. http://www.grassley.senate.gov/sites/default/files/about/upload/2010-09-24-GPO-Report.pdf. Accessed March 17, 2015.
14. Tu K, Cauch-Dudek K, Chen Z. Comparison of primary care physician payment models in the management of hypertension. Can Fam Physician. 2009;55(7):719-727.
15. Kaiser C. Survey: are the days of independent cardiologists numbered? Cardiovascular Business website. http://www.cardiovascularbusiness.com/topics/practice-management/survey-are-days-independent-cardiologists-numbered. Published May 28, 2010. Accessed March 17, 2015.
16. Peabody JW, Luck J, Glassman P, Dresselhaus TR, Lee M. Comparison of vignettes, standardized patients, and chart abstraction: a prospective validation study of 3 methods for measuring quality. JAMA. 2000;283(13):1715-1722.
17. Peabody JW, Luck J, Glassman P, et al. Measuring the quality of physician practice by using clinical vignettes: a prospective validation study. Ann Intern Med. 2004;141(10):771-780.
18. Peabody J, Shimkhada R, Quimbo S, et al. Financial incentives and measurement improved physicians’ quality of care in the Philippines. Health Aff (Millwood). 2011;30(4):773-781.
19. Peabody JW, Florentino J, Shimkhada R, Solon O, Quimbo S. Quality variation and its impact on costs and satisfaction: evidence from the QIDS study. Med Care. 2010;48(1):25-30.
20. Frei CR, Bell AM, Traugott KA, et al. A clinical pathway for community-acquired pneumonia: an observational cohort study. BMC Infect Dis. 2011;11:188.
21. Huang Z, Lu X, Duan H. On mining clinical pathway patterns from medical behaviors. Artif Intell Med. 2012;56(1):35-50.
22. Aoki T, Nakajima T, Saito Y, et al. Assessment of the validity of the clinical pathway for colon endoscopic submucosal dissection. World J Gastroenterol. 2012;18(28):3721-3726.
23. Hauck LD, Adler LM, Mulla ZD. Clinical pathway care improves outcomes among patients hospitalized for community-acquired pneumonia. Ann Epidemiol. 2004;14(9):669-675.
24. Walker J, McKethan A. Achieving accountable care—”it’s not about the bike.” N Engl J Med. 2012;366(2):e4.
25. Fisher ES, Shortell SM, Kreindler SA, Van Citters AD, Larson BK. A framework for evaluating the formation, implementation, and performance of accountable care organizations. Health Aff (Millwood). 2012;31(11):2368-2378.
26. Hussey, PS, Wertheimer S, Mehrotra A. The association between health care quality and cost: a systematic review. Ann Intern Med. 2013;158(1):27-34.
27. Institute of Medicine. Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. Washington, DC: The National Academies Press; 2012.
28. Rotter T, Kinsman L, James E, et al. Clinical pathways: effects on professional practice, patient outcomes, length of stay and hospital costs. Cochrane Database Syst Rev. 2010;3:CD006632.
29. Schippits MG, Schippits KM. Clinical pathways leading healthcare reform: transformational strategies for oncology and beyond. J Med Pers. 2013;11(2): doi:10.1007/s12682-013-0151-4.
30. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness of evidence-based treatment guidelines for the treatment of non-small-cell lung cancer in the community setting. J Oncol Practice. 2010;6(1):12-18.
31. Loeb M, Carusone SC, Goeree R, et al. Effect of a clinical pathway to reduce hospitalizations in nursing home residents with pneumonia: a randomized controlled trial. JAMA. 2006;295(21):2503-2510.
32. Gesme DH, Wiseman M. Strategic use of clinical pathways. J Oncol Pract. 2011;7(1):54-56.
33. Carlson B. Controlling the cost of care through clinical pathways. Biotechnol Healthc. 2009;6(1):23-26.
34. Winn W, Wong J, Cooper T, Olsen RS, Trehan JA, Feinberg BA. Potential cost savings of a continuous quality improvement program as part of an oncology medical home. J Clin Oncol 31, 2013 (suppl; abstr e17579).