REPORTS FROM THE FIELD
VALUE-BASED PURCHASING
Bridges to Excellence: Building a Business Case for Quality Care Francois de Brantes, Robert S. Galvin, MD, and Thomas H. Lee, MD
Abstract • Objective: To describe how the Six Sigma methodology was used to develop a performance-based incentive program in which physicians who deliver high-quality care will receive bonuses paid by participating employers. • Methods: Description of process. • Results: The work team agreed on a set of core principles to define the program’s design and defined the critical needs of the “customer” (ie, physicians and patients). Using a tool called a Quality Functional Deployment, they identified and ranked processes with respect to their level of impact on customer needs. Three key process categories emerged: clinical information systems, care management, and patient education and support. Further refinement of these process categories led to the selection of a set of performance measures in the areas of diabetes care and patient care management systems. A patient incentive program was developed to encourage patients to take a more active role in managing their condition. The program was designed to be administratively simple and does not disrupt current network arrangements. • Conclusion: Using the Six Sigma methodology, a diverse coalition of employers, health plans, and physicians worked collaboratively to develop a reimbursement model to improve quality and provide a return on investment to both health care purchasers and providers.
T
he Institute of Medicine (IOM) 2001 report Crossing the Quality Chasm [1] documented the quality shortfalls in the U.S. health care system and provided a roadmap for change. Subsequent reports have substantiated the quality issue, including a recent paper in which 20% of physicians and 25% of the public reported having had a personal experience involving serious harm due to avoidable error [2]. A key point made by the IOM is that the current reimbursement system does not encourage, and frequently discourages, quality improvement. Health care purchasers, www.turner-white.com
concerned about the rising costs of health care, believe that improving quality will mitigate unnecessary cost increases; the Leapfrog Group, a coalition of health care purchasers [3], has established rewarding quality as a fundamental purchasing principle. Leading providers and provider organizations also recognize the instability of the status quo. Although several studies have demonstrated that quality can reduce overall costs, there is no consensus that this is true. Part of the problem in establishing a business case for quality is that results vary, depending on the type of quality improvement (reduction in overuse, misuse, or underuse), type of reimbursement system (fee-for-service or prepaid), and recipient of the reward (payers or providers) [4]. Nonetheless, there is a growing consensus in health care that better quality should be rewarded, and various pay-for-quality initatives are underway. General Electric (GE), through its adoption of a quality improvement methodology called Six Sigma, has demonstrated billions of dollars of savings over the past few years. Working together with a diverse coalition of physicians, health plans, consultants, and large employers, GE applied the same Six Sigma methodology (called Design for Six Sigma [DFSS]) to develop the Bridges to Excellence program. The program defined a clear mission: to create an adaptable health care model that simultaneously rewards providers (or provider organizations) and purchasers for quality performance. In this paper, we will describe how the DFSS methodology was used to bring this program to life. Design for Six Sigma The Six Sigma process lays out a series of steps, grouped in “tollgates,” and provides statistical tools to guide the development of a new product or service. Two unique concepts in Six Sigma are CTQs—program attributes that are Critical to Quality and define what the customer needs—and CTPs— design attributes of the product or service that are Critical to Process and will ensure that CTQs are met.
From General Electric, Inc., Fairfield, CT (Mr. de Brantes and Dr. Galvin) and Partners HealthCare, Boston, MA (Dr. Lee).
Vol. 10, No. 8 August 2003 JCOM 439
BRIDGES TO EXCELLENCE Defining the CTQs Defining the CTQs requires that all the stakeholders agree on a core set of principles that will define the program’s design. Given the nature of the Bridges to Excellence program—a performance-based incentive program with performance measures to be made public—both physicians and patients (consumers) are considered customers; all other parties, including purchasers, are stakeholders. This distinction is critical because it is the customers’ needs that drive most of the product’s design. Identifying, sorting, and ranking customer needs, or CTQs, was accomplished through a combination of interviews, focus groups, and literature search. In prior work, GE had defined key attributes that consumers require of health care–related information [5]. Physician needs were collected through focus groups and later validated by work done by Bailit et al on incentive programs [6]. In our focus groups, physicians identified 3 types of incentives and rated their importance: direct financial incentives (most important), indirect financial incentives, and nonfinancial incentives (least important). It was clear that physicians felt rewards and incentives have to be (1) meaningful enough to more than compensate for the added cost associated with data collection and measurement, (2) perceived as fair and equitable, (3) attainable, (4) periodically reviewed, and (5) incremental, with small-step increments as opposed to a “cliff.” In addition, they should be based on measures that are standard, well-accepted by experts, and actionable by the physician or provider being measured. Other key attributes of successful physician-based incentive programs that emerged from the focus groups include simplicity and standardization of processes, no added burden on staff or office, low intensity of data requirements, increased income while giving high-quality care, ability to educate and motivate patients to seek out high-quality providers, ability to educate staff and enable them to be better teachers, and avoid putting the physician at odds with the patient. As the design of the program evolved from highlevel to detailed design, providers were regularly interviewed and consulted to make sure that the incentives and rewards would meet their needs and stimulate their desire to achieve the performance measures. Defining the Program Specifications—The “What” Performance Measures Designing performance measures requires identifying key processes that meet all customer needs and expectations. Customer (physicians and patients) expectations include care that is safe, effective, patient-centered, timely, efficient, and equitable— attributes defined in the IOM report. Care processes in support of these attributes needed to be measurable, actionable, and under the control of the provider being measured. 440 JCOM August 2003 Vol. 10, No. 8
With these attributes of care defined, the team identified 16 processes that could impact these attributes. The Six Sigma tool used in this phase of the design is called a Quality Functional Deployment (Figure 1). As seen in the figure, each attribute was given an importance score (1 to 5) relative to the other attributes. Then each process was assigned a high, medium, or low impact on each attribute. The corresponding score in each intersecting cell is obtained by multiplying the relative importance rank by either 9, 3, or 1 (the scores assigned respectively to high, medium, and low). It took the team several weeks and then a full day of debate to agree on the ranking of each of the key processes of care with respect to how they impacted the critical attributes. The ranked processes were then grouped into 3 tiers, and the top 11-ranked processes were grouped into 3 categories, shown in Figure 2: clinical information systems with evidence-based decision support, care management, and patient education and support. These 3 categories of processes are consistent with the areas of focus identified in the IOM report [1] and are the focus of like initiatives [7–9]. Turning these 3 categories of care processes into meaningful measures led to a canvassing of existing performance measures to determine whether any set would map back to the 3 groups of processes. One program emerged as an excellent candidate: the American Diabetes Association (ADA)– National Committee for Quality Assurance Diabetes Physician Recognition Program (DPRP) [10], in which physicians measure their performance on a number of criteria related to care of diabetic patients, such as patient HbA1c and lipid levels. Data indicate that DPRP-recognized physicians have improved care delivery [11]. While the DPRP criteria are primarily outcomes based, achieving these outcomes requires a certain reengineering of a practice. We elected to use DPRP measures for our Diabetes Care Link program, which includes rewarding physicians for meeting DPRP standards and patients for better self-care. Given that most patients are not diabetic, measuring provider performance for diabetic care alone could not achieve the broad impact specified in the original Bridges to Excellence mission statement. Turning again to the 3 categories of key care processes that had been identified, the Bridges to Excellence team refined these processes into a set of performance indicators for another program, Physician Office Link (Figure 3). Rewards The existing data on cost savings due to improvements in treating diabetes or managing information flow in a physician’s office are not definitive. However, purchasers, with their bias towards action, believe that there is sufficient evidence to move forward. Actuarial models indicate that potential short- and long-term savings of about 7% of total costs www.turner-white.com
Value-based decision support focused on an integrated PBM program
Information and resources enabling the patient to make fully informed decisions for treatment of condition
Information and resources enabling the patient to make fully informed decisions for managing health
Proactive management of patient care—effective systems of communication across providers
Proactive management of patient care—effective systems of communication between patients and providers
Proactive management of patient care—risk factor screening
Proactive management of patient care—support systems for patient lifestyle changes
Information and resources for both clinicians and patients in managing specific, high-intensity conditions typically, but not always, after an acute episode or hospitalization
Total
L
M H
H
M
L
L
H
370
H
M
H
M H
H
M
H H
M
M
H
M
M
480
L
H M
5 M H
Patient-centered—focused on the patient’s values, physiology, respect- 5 ing the patient’s needs for information and support
Data capture and management of provider compliance with standard of care
5
Effective—care be evidence-based and designed to help the patient get better as soon as appropriate
Proactive management of patient care—coordination of multispecialty teams
Value-based decision support focused on an integrated radiology program
M
Safe—care not harm the patient
Proactive management of patient care—disease management programs
Evidence-based clinical treatment decision support focused on guidelines for care
H
Data capture and management of patient compliance with standards of care
L
Data capture and management for patient tracking
M
Importance
H
Customer expectation
Figure 1. Quality functional deployment tool.
VALUE-BASED PURCHASING
Evidence-based clinical treatment decision support focused on error prevention
REPORTS FROM THE FIELD
L
M
L
L
M
L
M
H
H
H H
H
H
H
H
H
470
L
M M
L
M
M M
M
L
M H
H
H
M
L
H
256
H
M
H
H
H
M
M
M M
H
M
M
L
H
440
Equitable—appropriate standards of care be applied to all, irrespective of 4 H M H gender, color, creed, socio-economic background, culture, etc
H
M
H
H
M
M
H M
M
L
H
L
H
368
Measurable—processes used be measurable
4 H
H
H
M
H
H
H
L
L
M L
L
M
H
M
M
328
Portable—processes defined be adaptable from one care system/ setting to another
3 H
H M
L
H
L
M
H
H
M L
M
H
H
H
H
288
Actionable—processes defined be implemented by providers
5 H
H
H
H
H
H
H
H
H
H M
H
H
H
H
H
690
Impactful—processes defined favorably impact the largest number of 2 H patients possible
H M
M
H
M H
M
M
M M
H
H
H
M
L
176
“Customer”-centric—convenient to the patient and delivered with a high level of “customer service”
M
L
L
L
H
H
M H
H
H
M
H
H
258
189
273
205
249 253
301
195
335
Timely—care be given when needed, with minimum delays, efficient flow, as appropriate
4
Efficient—care given represent the best use of resources to get the best value for dollars spent
5 M H
Total
www.turner-white.com
3
L
L
209 309 257
H
211 333 253
259 293
Vol. 10, No. 8 August 2003 JCOM 441
BRIDGES TO EXCELLENCE
Evidence-Based Clinical Information/Decision Support Systems • Presence of a reliable system that identifies patients and assesses their treatment plan against guidelines • Presence of framework to trigger action and help doctors follow guidelines
Care Management • PCP transfers care for highintensity chronic condition to appropriate specialist • Appropriate level of contact by care manager with patient based on clinical condition and compliance with followup care to assess clinical needs • Composition and actions of multispecialty team appropriate to acute care diagnosis and patient’s needs
Figure 2. Process groupings. PCP = primary care physician. Patient Education and Support • Patient educational assessment • Patient involved in decision around treatment plan • Provider/patient communications, post-visit and between visits • Available self-management tools for patients
can be achieved by improving outcomes for diabetic patients [12–15]. In addition, potential short-term savings of 4% of total cost of care (or overall premiums) can be achieved by a more thorough reengineering of physician practices [16–23]. Since purchasers will require a return on any additional monies paid to physicians, it is reasonable for them to keep 50% of the expected savings and to share the other 50%, setting those funds aside as the incentive pool available to those who meet the performance standards. This analysis and parsing of the savings pool led to establishing a yearly bonus of $100 per diabetic patient for physicians meeting the Diabetes Care Link performance measures, and $55 per patient per year for physicians meeting the Physician Office Link performance measures. The Diabetes Care Link performance bonus is a “cliff”— a physician either gets the bonus or doesn’t—which was not the preferred bonus method expressed by physicians. However, the Physician Office Link bonus is structured to be gradual and to provide the physician with seed money to invest in better systems. As seen in Figure 3, there are 9 separate modules of measures. The physician bonuses are structured to encourage physicians to meet an increasing number of modules over 3 years. In the first year they can qualify for the full bonus ($55) by meeting any one of the modules in each column, for a total of 3 modules. In the second year they 442 JCOM August 2003 Vol. 10, No. 8
have to meet 2 out of 3 in each column, for a total of 6 modules. In the third year they need to meet all 9 modules. If a physician does not improve his or her performance from one year to another, they still qualify for a bonus, albeit lower than the previous year. This system of providing a graduated increase in performance while still providing an opportunity to qualify for the maximum bonus seems to resonate with physicians more than the “cliff”-type bonuses. While nonfinancial rewards—in particular, public recognition programs via some form of rating system—are not of uniform importance to providers (in fact, some providers have expressed strong antipathy for public data dissemination [24]), purchasers and consumers have demonstrated a strong demand for comparative provider performance data [25]. Until now, however, there has not been widespread adoption of these data sets, and research has shown that most provider report cards are not understandable to consumers [26]. To better meet the needs of consumers in this domain, a research project was launched to gather critical input from consumers, enabling the design of an enhanced provider directory that could incorporate all of the data elements that are important for consumers to make informed decisions. An initial series of focus groups was conducted during which consumers delineated all the data elements that they wanted, and categorized those data into intuitive www.turner-white.com
REPORTS FROM THE FIELD CLINICAL INFORMATION SYSTEMS/ EVIDENCE-BASED MEDICINE Basic Registries and Follow-up 1. Type of registry used for chronic conditions 2. Percentage of patients in registry 3. Use of registry to identify patient populations 4. Use of paper or electronic system to track and follow up on referrals and test results
Electronic Registries, Rx, and Test Ordering 1. Types of patient information in registry 2. Capabilities of an electronic system for prescriptions and tests 3. Use of electronic system for ordering prescriptions and checking for safety and efficiency 4. Use of electronic system to order and retrieve tests 5. Use of electronic system to track missing test results, distinguish abnormal results, and prompt follow-up on test results
Electronic Medical Records (EMR) 1. Types of patient information in an EMR 2. Percentage of patients who have information in the EMR 3. EMR’s capability to report across practice on multiple fields 4. EMR’s capability to use decision support to prompt physician interventions 5. EMR’s capability to capture services ordered, delivered, or paid 6. Use of EMR to track referrals and test results
PATIENT EDUCATION AND SUPPORT
Pts Educational Resources 10 10 40 40
Pts
1. Assessment of patient language 30 preferences and risk factors 2. Identification of preferred lan35 guages in patient population 3. Provision of educational resources 35 in preferred languages for risk factors and chronic conditions
100
100
Pts Referrals for Risk Factors and Chronic Conditions
Pts
10 20 40
50 1. Percent of patient who have specific risk factors 2. Provision of referrals for education 50 and support to patients with risk factors and chronic conditions
VALUE-BASED PURCHASING
CARE MANAGEMENT Care of Chronic Conditions 1. Identification of the practice’s top three chronic conditions 2. Structured process for disease management for patients with the top three conditions 3. Use of resources to assist with medication compliance, appointments, and barriers to care
Pts 10 45
45
100 Preventable Admissions 1. Using data to identify patients who are at risk for emergency admissions 2. Identification of the reasons and prevalence of emergency admissions 3. Structured systems to prevent emergency admissions
Pts 20
20
60
10 20
100
100
Pts Quality measurement and Improvement
Pts
10 20 30 30
1. Identification of opportunities for improving outcomes or processes 2. Setting goals for performance for identified opportunities of improvement 3. Measurement of performance and identification of goals not met 4. Implementation of improvement activities
20 20
20 40
5 5
100
100
100 Care of High-Risk Medical Conditions 1. Resources for managing patients with high-risk conditions 2. Number and percent of patients who receive high-risk care management 3. Contents of the high-risk care management program 4. Qualifications of the high-risk care manager 5. Types of information in database of patients with high-risk conditions 6. Frequency of communication between physician and care manager 7. Frequency of communication between care manager and patient
Pts 5 5
30 10 15
5
30
100
Figure 3. Summary of Physician Office Link measures.
groupings. Subsequent focus groups used a “pencil and paper” exercise during which the groups of measures and associated labels were tested. The result is a prototype shown in Figure 4 that needs further validation and testing www.turner-white.com
to determine its effectiveness in helping consumers select physicians and hospitals. In addition to receiving information on physicians, patients in the Bridges to Excellence program are engaged to Vol. 10, No. 8 August 2003 JCOM 443
BRIDGES TO EXCELLENCE Doctor Information Dr. Robert Smith FAMILY PRACTICE ID NO: 00046688833 03 My philosophy of care 518.472.4584 518.472.4620 fax
[email protected]
Address & Hours
Staffing
Credentials
Hospital Affiliation
997 Glen Cove Avenue Glen Head, NY 11545
• 2 Nurses • 3 Technicians • 1 On-call doctor
NY Medical College, M.D., 1969 St. Lukes–Roosevelt, 1992 AM Board of Internal Medicine, 1994
Mt. Sinai Medical Center Westchester Medical Center Columbia Presbyterian Medical Center
Monday–Thursday 10–5 Friday, Saturday 11–4
Performance Report: Effectiveness of Care Overall Doctor: Average score:
Diabetes Care
Cardiac Care
5 6
Overall Doctor: Average score:
Clinical Information Systems & Evidence-Based Medicine Basic registries and follow-up
Patient Experience of Care
✔ 100%
Patient Education and Support Educational resources
✔ 70%
4 5
Care Management Care of chronic conditions
Electronic registries, prescription and test ordering
Referrals for risk factors and chronic conditions
Preventable admissions
Electronic medical records
Quality measurement and improvement
Care of high-risk medical conditions
Key ✔ Provider has fulfilled the requirements for the measure
Doctor/Patient Interactions
Access and Office Systems
Communication
Organizational access
Interpersonal treatments
Visit-based continuity
Knowledge of patient
Clinical team
Health promotions Integration Key Patient trust
Your provider Average provider
Relationship duration
Figure 4. Physician report card prototype.
better understand their condition and encouraged to improve or stabilize it. Work by Wagner et al [12] has demonstrated that it is not possible to get the full yield from a chronic care management model without robust patient in444 JCOM August 2003 Vol. 10, No. 8
volvement, a view that was strongly echoed by physicians involved in developing Bridges to Excellence. They strongly felt that if a physician performance measure was based on patient outcomes, patients also should have an incentive to www.turner-white.com
REPORTS FROM THE FIELD improve their outcomes. The result is Diabetes Care Rewards, a novel program to encourage employees and family members to take a more active role in managing their condition. The program includes tools, support, and information for diabetic patients to monitor their self-care activities; patients receive points for lowering their HbA1c level and following care guidelines. Patients can accumulate points to qualify for rewards offered by participating employerspurchasers, such as vouchers for lower copayments on doctor visits or prescriptions or coupons for products not routinely covered by health benefits (eg, sugar-free candy). In focus groups, consumers indicated that a monetary or quasimonetary reward was very important and would keep them focused on achieving better outcomes. Designing the Program Implementation—The “How” There are 3 CTQs that drove the majority of the operational design for Bridges to Excellence: (1) make the rewards as meaningful as possible by consolidating the bonuses in a single payment, (2) make the program administratively simple for purchasers, plans, and providers, and (3) don’t cause the plans to open up their provider contracts or do anything that would disrupt current network arrangements. These CTQs forced us to eliminate many options (eg, having each plan administer and pay the bonuses) that would have been easy for purchasers to implement but would have been counter to what the customers and other stakeholders wanted. One of the core principles in designing a new program using the Six Sigma methodology is not to retrofit a solution into an existing infrastructure, because the existing infrastructure may not meet the needs of your customers. As a result, the operational framework chosen by Bridges to Excellence was to hire an independent third party (Medstat, Ann Arbor, MI) as the overall program manager. Medstat’s role is to aggregate data files from plans, creating a master patient/physician/purchaser grid that defines the number of patients per physicians for whom a bonus could be paid, and enables the participating purchaser to quickly gauge what their maximum exposure would be if all physicians met the performance measures. In addition, Medstat invoices each purchaser on a quarterly basis reflecting all the bonuses that have to be paid to physicians that meet the performance measures, and then pays the physician a lump sum bonus, across all participating purchasers. This structure ends up being administratively simple because it is not dependent on a specific health plan or network arrangement and does not require a plan to modify its existing contractual arrangements with network physicians. In fact, the health plan’s role is limited to sending the data file to Medstat (although they can also become involved in helping physicians in their network meet
www.turner-white.com
VALUE-BASED PURCHASING
the performance measures, and in engaging patients with diabetes to enroll in the self-care activities). For employers, it does require signing a few agreements so that the data can flow between their plan and Medstat, and they are bound by the terms of the program to pay the bonuses and engage their employees in better self-care. Conclusion Designing a new product or service in a fragmented health care system with stakeholders who can have highly divergent needs was far from easy. However, the framework provided by the Six Sigma methodology enabled stakeholders to make trade-offs and build consensus, ensuring that the design would, overall, have wide appeal among purchasers, providers, plans, and patients. In fact, Bridges to Excellence has already succeeded in many of its objectives. The program is adaptable: in creating an organizational structure that is not dependent on health plans, plan designs, or contractual relationships between plans and providers, many of the barriers to soliciting the participation of multiple purchasers have been avoided. The measures are objective and directly actionable by providers. An actuarial business case for providers and purchasers has been created, which aligns incentives around higher quality care. Whether the incentives are sufficient to encourage physicians to have their performance assessed will be a real test of success, and is dependent in large part on the number and size of participating purchasers. Early results from the first intervention markets in Louisville, KY, Cincinnati, OH, and Boston, MA, indicate that the incentives are sufficient. Another key test will be to validate that better performance leads to measurable defect reductions and savings for purchasers. These two key tests are closely related, as future participation by public purchasers, in particular, Medicare, will be necessary to provide the volume needed to create a robust business case for providers, and budget neutrality will require that savings offset the increased payments. Irrespective of the success of this particular project, a new chapter in U.S. health care has begun: purchasers, firm in the belief that improved quality does indeed cost less, have stated that they are willing to pay more for better performance provided that the performance is transparent to the consumer. In turn, leading providers and provider organizations appear willing to publicly release measures of their performance in the belief that rewarding quality will enable them to fund the infrastructure changes that are necessary to further improve processes and quality.
Corresponding author: Francois de Brantes, Program Leader, Health Care Initiatives, 3135 Easton Turnpike, W2A, Fairfield, CT 06828.
Vol. 10, No. 8 August 2003 JCOM 445
BRIDGES TO EXCELLENCE References 1. 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. 2. Blendon RJ, DesRoches CM, Brodie M, et al. Views of practicing physicians and the public on medical errors. N Engl J Med 2002;347:1933–40. 3. Birkmeyer JD, Birkmeyer CM, Wennberg DE, Young M. Leapfrog patient safety standards: the potential benefits of universal adoption. Available at www.leapfroggroup.org/ safety1.htm. Accessed 28 Jul 2003. 4. Galvin RS. The business case for quality. Health Aff (Millwood) 2001;20:57–8. 5. de Brantes F, Galvin RS. Creating, connecting and supporting active consumers. Int J Med Marketing 2001;2:73–80. 6. Bailit M, Dyer MB. Provider incentive models for improving quality of care. Available at www.nhcpi.net/monographs.cfm. Accessed 28 Jul 2003. 7. Bodenheimer T, Wagner EH, Grumbach K. Improving primary care for patients with chronic illness. JAMA 2002;288: 1775–9, 1909–14. 8. Centers For Medicare and Medicaid Services. Physician group practice demonstration. Available at www.cms.hhs.gov/ healthplans/research/PgpDemo.asp. Accessed 28 Jul 2003. 9. Integrated Healthcare Association. Pay for performance. Available at www.iha.org/Ihaproj.htm. Accessed 28 Jul 2003. 10. National Committee for Quality Assurance. The diabetes physician recognition program. Available at www.ncqa.org/ dprp/dprpmain.htm. Accessed 28 Jul 2003. 11. National Committee for Quality Assurance. The state of health care quality, 2002. Washington (DC): The Committee; 2002. 12. Wagner E, Sandhu N, Newton K, McCulloch D, et al. Effect of improved glycemic control on health care costs and utilization. JAMA 2001;285:182–9. 13. The CDC Diabetes Cost-effectiveness Group. Cost-effectiveness of intensive glycemic control, intensified hypertension control, and serum cholesterol level reduction for type 2 diabetes. JAMA 2002;287:2542–51.
14. Vijan S, Hofer TP, Hayward RA. Cost-utility analysis of screening intervals for diabetic retinopathy in patients with type 2 diabetes mellitus. JAMA 2000;283:889–96. 15. Bodenheimer T, Lorig K, Holman H, Grumbach K. Patient self-management of chronic disease in primary care. JAMA 2002 ;288:2469–75. 16. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computerbased clinical decision support systems on physician performance and patient outcomes: a systematic review. JAMA 1998; 280:1339–46. 17. Elson RB, Connelly DP. Computerized patient records in primary care. Their role in mediating guideline-driven physician behavior change. Arch Fam Med 1995;4:698–705. 18. Hammond KW, Prather RJ, Date VV, King CA. A providerinteractive medical record system can favorably influence costs and quality of medical care. Comput Biol Med 1990;20: 267–79. 19. Tierney WM, Miller ME, McDonald CJ. The effect on test ordering of informing physicians of the charges for outpatient diagnostic tests. N Engl J Med 1990;322:1499–504. 20. Bates DW, Kuperman GJ, Rittenberg E, et al. A randomized trial of a computer-based intervention to reduce utilization of redundant laboratory tests. Am J Med 1999;106: 144–50. 21. Jones DL, Kroenke K, Landry FJ, et al. Cost savings using a stepped-care prescribing protocol for nonsteroidal antiinflammatory drugs. JAMA 1996;275:926–30. 22. Research by Cap Gemini Ernst & Young US LLC: e-prescribing in a multi-center group. Available at www.allscripts.com/ ahcs/news_2.asp?S=2017&ID=2. Accessed 28 Jul 2003. 23. TouchScript medication management system: financial impact analysis on pharmacy risk pools. Available at www. allscripts.com/ahcs/epres/epres.pdf. Accessed 28 Jul 2003. 24. Galvin R, Milstein A. Large employers’ new strategies in health care. N Eng J Med 2002;347:939–42. 25. Marshall MN, Shekelle PG, Leatherman S, Brook RH. The public release of performance data: what do we expect to gain? A review of the evidence. JAMA 2000 ;283:1866–74. 26. Hibbard JH, Jewett JJ. Will quality report cards help consumers? Health Aff (Millwood) 1997;16:218–28.
Copyright 2003 by Turner White Communications Inc., Wayne, PA. All rights reserved.
446 JCOM August 2003 Vol. 10, No. 8
www.turner-white.com
