Health care costs in the United States consume a significant proportion of gross domestic product (GDP) (17 to 18 percent, or nearly $3 trillion, annually1) and are increasing at a rate 1.1 percent above the rate of growth of GDP.2 Issues of health care quality, access, safety, and equity remain pressing concerns as well. Further, compared with health care delivery in other developed economies, the U.S. health care delivery system has been found to be less effective, costing more and returning less.3
As the purchaser of health care for a third of all Americans, CMS currently accounts for a large proportion of that cost. In addition, as its covered population grows, and as its mission expands, CMS has a significant role in helping to control the cost and improve the quality of health care. Fundamental to modern businesses are the information and the information technology (IT) they rely on to operate effectively. Thus it is essential
2 Sean P. Keehan, Andrea M. Sisko, Christopher J. Truffer, John A. Poisal, Gigi A. Cuckler, Andrew J. Madison, Joseph M. Lizonitz, and Sheila D. Smith, 2010, “National Health Spending Projections Through 2020: Economic Recovery and Reform Drive Faster Spending Growth,” Health Affairs 30(8):1-12.
3 David A. Squires, 2011, The U.S. Health System in Perspective: A Comparison of Twelve Industrialized Nations, New York, N.Y.: The Commonwealth Fund, available at http://www.commonwealthfund.org/Content/Publications/Issue-Briefs/2011/Jul/US-Health-Systemin-Perspective.aspx, last accessed August 1, 2011.
that, at a minimum, CMS have access to information that could help it to work with the rest of the health care industry to meet those objectives.
Key policy perspectives such as moving toward what the Institute of Medicine calls a learning health care system4 have to be addressed by CMS, since federal programs are of such scale that CMS’s efforts will have significant impact on moving the nation as a whole forward. Indeed, evidence exists that improved quality performance can mitigate costs while also improving outcomes.5 Although fundamental public policy issues with respect to health care reform, safety and quality enhancement, cost management, and the appropriate federal role in these matters are being debated (issues that are beyond the scope of this report), it is inarguable that CMS must not only ensure the appropriateness of its direct outlays but also support, both proactively and reactively, efforts to reduce costs in the health care system as a whole. No matter what payment models it operates under now or in the future, CMS will have to be able to both fulfill its fiduciary responsibility and ensure that eligible people are receiving appropriate care.
What planning must CMS embrace to fulfill these responsibilities? Central to the effort is that CMS develop a vision of its role in modern health care and strategies for realizing it—an overall strategy for CMS as a whole, along with a strategic technology plan for comprehensive, incremental development of effective information systems, and strategies for quickly addressing near-term issues.
For the purposes of this report the terms “modernization” and “transformation” refer to two ends of a spectrum of possible transitions for components and subcomponents of an information system. Modernization refers to modest or evolutionary transitions; transformation refers to significant or revolutionary transitions. Use of the term “modernize” or
4 A system that is “designed to generate and apply the best evidence for the collaborative healthcare choices of each patient and provider; to drive the process of discovery as a natural outgrowth of patient care; and to ensure innovation, quality, safety, and value in health care.” See Institute of Medicine, 2011, Digital Infrastructure for the Learning Health System, Washington, D.C.: The National Academies Press, p. 2.
5 See, for instance, B.L. Hall et al., 2009, “Does Surgical Quality Improve in the American College of Surgeons National Surgical Quality Improvement Program?,” Annals of Surgery 250:363-376; J.B. Dimick et al., 2004, “Hospital Costs Associated with Surgical Complications: A Report from the Private-Sector National Surgical Quality Improvement Program,” Journal of the American College of Surgeons 199(4):531-537.
“transform” presumes that the transition being discussed is at one end of the scale or the other.
CMS has done an admirable job in meeting its growing responsibilities over time and has continued to improve its information systems. Since the introduction of a comparatively simple Medicare program in 1965, CMS (as well as its predecessor organizations) has effectively managed to meet its expanding responsibilities and has surmounted significant technology challenges in order to do so. CMS built and manages an extremely complex distributed transaction-processing system for the fee-for-service payment systems for Medicare Parts A and B. CMS also manages a large variety of core payment, monitoring, and reporting systems. Moreover, CMS has delivered robust technology systems to meet the requirements of new programs. It has consistently done so under severe time and budget constraints, a notable example being the delivery of Medicare Part D in 2006.6
In spite of these historical successes, however, the challenges facing CMS today are daunting. In order to continue to provide payments according to current requirements, to meet growing requirements for health-care-related data and analysis, and to support new payment models, CMS will have to modernize or transform at least some of its information systems.
Complexity and Dynamism of the Health Care Enterprise
Originally designed to perform straightforward Medicare eligibility determination and to manage and disburse health claim payments to hospitals and medical professionals,7 CMS’s systems are now expected
6 Considered a successful implementation, the Medicare Part D program entailed complicated interconnections between several federal agencies and deployment of complex websites for users, and was implemented with all statutorily mandated functionality within the statutory deadlines.
7 To provide some sense of scope and scale, the committee notes that CMS handles 240,000+ new Medicare beneficiaries monthly and must process approximately 200,000 notices of deaths of beneficiaries monthly. There are nearly 3 million eligibility inquiries daily. More than 14 million monthly transactions from Medicare Advantage and Part D plans and more than 1.2 billion Medicare fee-for-service (FFS) claims are processed annually. CMS receives, processes, and stores more than 5.3 million Part D prescription drug events daily. In terms of dollar outlay, CMS makes nearly $1 billion in FFS payments daily and must calculate and pay more than $13.1 billion monthly for Medicare Advantage and Part D (Tony Trenkle, 2010, “CMS Systems,” presentation, TechAmerica Federal Committee Meeting, Baltimore, Md., April 20). By Medicare statute, much of the day-to-day administration of the program is delegated to private contractors. Functions such as paying providers (processing reimbursement claims), enrolling providers and suppliers in the Medicare program, educating providers about Medicare billing requirements, and processing appeals are performed by Medicare administrative contractors (MACs), fiscal intermediaries (FIs), and carriers. Generally, MACs perform these functions for Parts A and B providers, FIs for Part A providers, and carriers for Part B providers. In 2003, the Medicare Modernization Act required the secretary to implement FFS contracting reform and to replace FIs and carriers with MACs by 2011. CMS has contracted with 15 A/B MACs (each responsible for processing the claims from several states) to process Part A and B claims and with 4 durable medical equipment (DME) MACs to process DME supplier claims (CRS, 2010, “Medicare Primer,” R40425, available at http://aging.senate.gov/crs/medicare1.pdf, last accessed September 12, 2011). In terms of upcoming procurements, CMS expects to consolidate the present 15 MAC jurisdictions into 10 in a phased process over several years (per information available at https://www.cms.gov/MedicareContractingReform/04_VisionofFutureFeeforServiceMedicareEnvironment.asp, last accessed October 15, 2011). CMS works with nearly 5,000 hospitals (CMS, 2011, “Hospital General Information,” available at http://data.medicare.gov/dataset/Hospital-General-Information/v287-28n3, last accessed September 12, 2011), more than 11,000 home health agencies (CMS, 2011, “Home Health Care Facilities,” data set available at http://data.medicare.gov/dataset/Home-Health-Care-Facilities/6jpm-sxkc, last accessed September 12, 2011), more than 65,000 DME providers (CMS, “DME Supplier,” data set available at http://data.medicare.gov/dataset/DME-Supplier/p7kk-c8cp, last accessed September 12, 2011), approximately 16,000 skilled nursing providers (CMS, 2011, “SNF Provider ID Information,” data set available at http://www.cms.gov/CostReports/Downloads/SNFProviderID06302011.zip, last accessed September 12, 2011), and more than 5,000 dialysis facilities (CMS, 2011, “Dialysis Facility Compare—Listing by Facility,” data set available at http://data.medicare.gov/dataset/Dialysis-Facility-Compare-Listing-by-Facility/23ew-n7w9, last accessed September 12, 2011).
to handle a much more complex and demanding set of activities, including special payments for certain procedures, prescription drug payments, aggressive discovery and recovery of erroneous payments, reporting of Medicare data showing comparisons of medical quality across hospitals and medical practices, differential payments to hospitals that train doctors, and medical provider payments adjusted for actual inclusion of good medical practices as defined by independent medical-professional entities. The 2009 Health Information Technology for Economic and Clinical Health (HITECH) Act and the 2010 Patient Protection and Affordable Care Act (PPACA) also impose new requirements, as described in Chapter 1.
CMS’s programs are thus entering a very dynamic phase, with many changes expected or anticipated, including:
• A shift toward pay for performance or value and away from fee for service;
• The need to manage and report on multiple “experiments” in reforming payment for delivery for health care, and potentially to scale them up more broadly depending on the results of the innovations;
• The incorporation and use of some clinical data, and even population health data, obtained from a variety of sources such as electronic medical devices and electronic health record data whose collection is facilitated through HITECH;
ally, MACs perform these functions for Parts A and B providers, FIs for Part A providers, and carriers for Part B providers. In 2003, the Medicare Modernization Act required the secretary to implement FFS contracting reform and to replace FIs and carriers with MACs by 2011. CMS has contracted with 15 A/B MACs (each responsible for processing the claims from several states) to process Part A and B claims and with 4 durable medical equipment (DME) MACs to process DME supplier claims (CRS, 2010, “Medicare Primer,” R40425, available at http://aging.senate.gov/crs/medicare1.pdf, last accessed September 12, 2011). In terms of upcoming procurements, CMS expects to consolidate the present 15 MAC jurisdictions into 10 in a phased process over several years (per information available at https://www.cms.gov/MedicareContractingReform/04_VisionofFutureFeeforServic eMedicareEnvironment.asp, last accessed October 15, 2011). CMS works with nearly 5,000 hospitals (CMS, 2011, “Hospital General Information,” available at http://data.medicare.gov/dataset/Hospital-General-Information/v287-28n3, last accessed September 12, 2011), more than 11,000 home health agencies (CMS, 2011, “Home Health Care Facilities,” data set available at http://data.medicare.gov/dataset/Home-Health-Care-Facilities/6jpm-sxkc, last accessed September 12, 2011), more than 65,000 DME providers (CMS, “DME Supplier,” data set available at http://data.medicare.gov/dataset/DME-Supplier/p7kk-c8cp, last accessed September 12, 2011), approximately 16,000 skilled nursing providers (CMS, 2011, “SNF Provider ID Information,” data set available at http://www.cms.gov/CostReports/Downloads/SNFProviderID06302011.zip, last accessed September 12, 2011), and more than 5,000 dialysis facilities (CMS, 2011, “Dialysis Facility Compare—Listing by Facility,” data set available at http://data.medicare.gov/dataset/Dialysis-Facility-Compare-Listing-by-Facility/23ew-n7w9, last accessed September 12, 2011).
• New legislative mandates and CMS responsibilities; and
• Changes in public policy expectations, especially increasing expectations regarding transparency, fraud resistance, timely assessments, delivery of care of improved quality and equity, and greater involvement of key stakeholders.
Controlling the growth of health care costs is essential. So-called coordination of care is central to all meaningful health care delivery models that are attempting to address cost issues, including patient-centered medical homes,8 accountable care organizations,9 and global payments.10 Coordination of care is key whether health care delivery models are sponsored by employers, private payers, public payers, or provider organizations. Any practical and scalable implementation of these models requires the exchange of electronic medical data. Moreover, the core infrastructure is already being built, admittedly in fits and starts, for example in the form of provider electronic health record and electronic medical
8 An approach to providing comprehensive primary care for children, youth, and adults, the patient-centered medical home (PCMH) is a health care setting that facilitates partnerships between individual patients and their personal physicians and, when appropriate, the patient’s family. Key principles in this approach to care are personal physician relationships; physician directed medical practice teams with collective responsibility for ongoing, wholeperson-oriented coordinated patient care; improved quality and safety; and enhanced access to care that features increased availability, communication, open scheduling, and pay-forquality approaches. Many private insurers are experimenting with payment schemes to motivate the adoption of the PCMH model, and to test the validity of the hypothesis that the model reduces costs and improves health. (See American Academy of Family Physicians [AAFP], American Academy of Pediatrics [AAP], American College of Physicians [ACP], and American Osteopathic Association [AOA], 2007, “Joint Principles of the Patient-Centered Medical Home,” concept paper, available at http://www.acponline.org/advocacy/where_we_stand/medical_home/approve_jp.pdf, last accessed August 1, 2011.)
9 An accountable care organization (ACO) is a network of doctors and hospitals that shares responsibility for providing care to patients. The intent is to use ACOs to “make providers jointly accountable for the health of their patients, giving them strong incentives to cooperate and save money by avoiding unnecessary tests and procedures.” (See Jenny Gold, 2011, “Accountable Care Organizations, Explained,” National Public Radio online, January 18, available at http://www.npr.org/2011/04/01/132937232/accountable-care-organizationsexplained, last accessed August 1, 2011.)
10 Global payments are an alternative to the fee-for-service model traditionally used for Medicare payments. They would give hospital and care provider groups lump sums, intended as incentives to increase efficiency and quality of care and to stop patients from returning to the hospital for preventable conditions. The premise is that the fee-for-service payment model does not encourage systematic improvements in health care quality and efficiency, but instead acts as an incentive for care providers to allow patients to remain sick so that they require additional medical services and treatments. (See Robert Steinbrook, 2009, “The End of Fee-for-Service Medicine? Proposals for Payment Reform in Massachusetts,” New England Journal of Medicine; Health Policy and Reform, available at http://healthpolicyandreform.nejm.org/?p=1247, last accessed August 1, 2011.)
record systems, electronic prescriptions, and state-level health information exchanges, and in the networking of intelligent medical monitoring devices. It seems clear that there is no turning back from the digitization of health care.11
In terms of meeting legislative requirements, CMS has begun to handle new types of clinical data12 and is working with states to set up insurance exchanges with eligibility determination, subsidy determination, and interfaces with other insurance programs.13 States are also involved in actually administering and disbursing meaningful-use incentive payments. These efforts require new systems. To accomplish the necessary coordination requires working with systems not only across CMS, but also across federal agencies, states, local governments, and profit/nonprofit entities. Such coordination would be a complex challenge even if all participants had unlimited resources and were working cooperatively with each other.
Irrespective of any future legislative changes that might once again modify the role of CMS, if CMS is to continue as a purchaser of health care in any form, it will almost certainly have to go beyond paying for “encounters” to paying for “episodes” and outcomes, integrated across financial and clinical dimensions. Further complicating matters, major changes are occurring in how health care data are collected and used to measure the effectiveness of specific clinical options. Consistent collection of rich data at the point of care can and will provide near-real-time sources of information that will form the basis of a learning health care system. Although CMS’s role and specific responsibilities with respect to such data are in flux, the agency clearly will have to be able to cope with changes in the broader health care system of which it is a part.
In addition to anticipated changes to CMS’s own programs are the significant changes that the broader health care, practice, and public policy environments are undergoing, including:
• Ongoing evolution and increasing sophistication of technology and a constantly changing set of technical options;
• An increasing need to manage the delivery of care based on medical advances empowered by genomics and proteomics;
• An infrastructure involving widespread use of electronic health records and regional data repositories;
11 Institute of Medicine, 2011, Digital Infrastructure for the Learning Health System: The Foundation for Continuous Improvement in Health and Health Care: Workshop Series Summary, Washington, D.C.: The National Academies Press, available at http://www.nap.edu/catalog.php?record_id=12912, last accessed August 1, 2011.
12 42 CFR 495.8, Demonstration of meaningful-use criteria.
13 42 CFR 425.5, Accountable Care Organizations.
• Uncertainty and some ambivalence regarding the essential value of access to health data for populations and individuals as a good competing with privacy concerns;
• Growing awareness of the diverse needs of racial/ethnic and other populations experiencing disparities in health care;
• Increasing needs to integrate important non-clinical health-related data with clinically generated medical data to assess and improve health; and
• Consumer access to online medical information and social networks.
Centrality of Data to Modernization and Transformation Efforts
CMS’s systems today are meeting current requirements but are challenged to meet emerging demands. Considering just the data layer, CMS information systems consist of silos of data stores for each major CMS program—Medicare Part A (hospital), Part B (outpatient), Part C (managed care), and Part D (prescription drugs)—in addition to data stores for quality clinical data, and other special data sets. Over the years, software has been added to meet new requirements for data breakouts and data interchange between the stores corresponding to Parts A, B, C, D, and other data stores. Some integrated databases have been developed to meet the needs for specific analyses, but these integrated data stores are not designed for (nor readily capable of) the interoperability and flexibility required for emerging (but not yet completely defined) new mission requirements.
There are software code bases specific to the various program silos, with some containing legacy code dating back 40 years. Such software has been continuously modified, resulting in transactionally robust but difficult to maintain sets of independent systems. Today it is time-consuming and costly for CMS to effect even annual or quarterly routine changes that legislative mandates require.14 And new mandates require new applications (or another layer of code) that then introduce additional complexity and new opportunities for breakdowns and errors.
It was possible, albeit challenging, to manage these systems when the add-ons and new requirements demanded, in essence, more detailed data breakouts within the individual systems (corresponding to programs). Increased efforts to reduce payment errors and fraud, among other needs,
14 Examples include updates to the various fee schedules used to determine provider payments such as for ambulance services; clinical laboratory testing; durable medical equipment, prosthetics, orthotics, and supplies; pharmaceutical reimbursements; and physicians fees schedule. (See CMS, 2011, “Fee Schedule—General Information,” website, available at https://www.cms.gov/feeschedulegeninfo/, last accessed August 1, 2011.)
have meant that new, larger, and more time-sensitive data sets have to be matched with data within these systems and across data sets corresponding to Parts A, B, C, D, and other silos (storing data associated with individuals and providers). To begin to address these sorts of challenges, CMS has developed some data warehouses, such as the chronic condition data warehouse.15 But increasingly, added demands will continue to change how CMS accesses and stores information and are likely to require ever greater flexibility in how CMS analyzes data.
If the programs CMS administers were to remain fundamentally static and the only issue were the growth in claims volume driven by an aging population, it is possible that CMS’s current systems could be evolved and adapted to satisfy the anticipated growth in transaction volume. However, these systems, built for different purposes and in a different era of technical sophistication, appear to lack the flexibility that might enable their expeditious evolution to address emerging challenges. The status quo is, therefore, not a realistic option.
In summary, not only is CMS faced with a near-term growing and complex workload, but its role is also central to addressing rapidly growing health care costs and the U.S. federal budget deficit. The programmatic requirements that CMS must meet now and in the rest of this decade require underlying changes in its systems that will be extremely challenging to meet through the structure of its systems today, and these changes cannot be delayed given the agency’s legislative mandates.
Although the daunting and substantial challenges confronting CMS might suggest a need for creating, all at once, a large all-encompassing system through one huge development effort, such “big bang” approaches to systems modernization almost always fail.16 Even when the end-state
15 As noted on its website, http://www.ccwdata.org/, “The CMS Chronic Condition Data Warehouse (CCW) provides researchers with Medicare and Medicaid beneficiary, claims, and assessment data linked by beneficiary across the continuum of care. In the past, researchers analyzing data files were required to perform extensive analysis related to beneficiary matching, deduplication, and merging of the files in preparation for their study analysis. With the CCW data, this preliminary linkage work is already accomplished and delivered as part of the data files sent to researchers.”
16 See, for example, Edward Cone, 2002, “The Ugly History of Tool Development at the FAA,” website “Baseline,” available at http://www.baselinemag.com/c/a/Projects-Processes/The-Ugly-History-of-Tool-Development-at-the-FAA/, last accessed August 1, 2011; NRC, 2004, A Review of the FBI’s Trilogy Information Technology Modernization Program, Washington, D.C.: The National Academies Press, available at http://www.nap.edu/catalog.php?record_id=10991, last accessed August 1, 2011; Vivek Kundra, 2010, 25 Point Implementation Plan to Reform Federal Information Technology Management,
seems clear, the record of success of such approaches is poor. Often they are driven by a technology agenda, thereby missing the principal business objectives, or “requirements creep” results when large-scale approaches attempt to solve too many problems at once. “Big bang” projects are, by their nature, so extensive and complex that the sheer technical task is simply too large for all but the most technically sophisticated organizations.
It is also important to note that the requirements that would be fulfilled by very ambitious large systems cannot be expected to remain fixed over time. Large systems take many years to develop, during which time the underlying initial needs often change. To succeed, such systems must have the ability to hit a moving target. Further complicating this situation is the observed fact that successful systems are themselves agents of change. Especially in cases where a new system brings unprecedented capability to stakeholder groups (especially new ones), the advent of the system and its capabilities often change the perceptions, expectations, and desires of these communities. Thus, the arrival of a new system can wash back upon the overall system in the form of needed change that is often not cosmetic, but instead can be radical.
In recognition of the drawbacks and risks of the big bang approach, the committee urges that CMS pursue an incremental approach to the development of the information systems that will be needed. But the nature of this incremental approach requires careful consideration and planning. A reactive, year-by-year and program-by-program approach to upgrading CMS systems, for example, is unlikely to succeed in meeting the new and emerging demands on CMS; nor is separate incremental consideration of new requirements, new communities, and new programs. By contrast, a comprehensive approach implemented incrementally can help to head off the issues that arise when increments are considered in isolation.
Successful information systems and data repositories must change and evolve constantly, in size, in reach, and in the stakeholder constituencies that interact with them. The reason for this ongoing evolution is the organic nature of the relationships between information services and human stakeholders. It has been widely observed that new kinds of information and new systems for making information available entice new users, and stimulate new uses by previous users.
The challenges that CMS faces cannot be met simply by upgrading
December 9, Washington, D.C.: White House, Office of the Chief Information Officer, available at http://www.cio.gov/documents/25-Point-Implementation-Plan-to-Reform-Federal%20IT.pdf, last accessed August 1, 2011; NRC, 2010, Critical Code: Software Producibility for Defense, Washington, D.C.: The National Academies Press, available at http://www.nap.edu/catalog.php?record_id=12979, last accessed August 1, 2011.
information technology. In some sense, all of IT resolves eventually to the processing of data by computers, databases, operating systems, and networks. Some of the most critical and difficult challenges are those arising from understanding and addressing stakeholder needs and CMS business considerations. These considerations are the ones that must be articulated in order to create the precise requirements that are to be met by information technologies. A strategy that emphasizes pursuing a comprehensive incremental approach to understanding and ultimately meeting the needs and requirements of CMS stakeholder communities and CMS internal business units will, in the committee’s view, serve CMS and the country well.
To be most effective, modernization and transformation efforts need to be focused on specific and concrete business issues and objectives, not on information technology per se. However, business decisions should be made with a clear understanding of the capabilities of advanced IT solutions.
CMS’s mission and organization are very complex—of the scope and scale of the largest private enterprises and other large federal agencies. As a result only evolutionary, iterative approaches are likely to succeed. The tension between meeting routine demands for processing and claims payment while simultaneously developing the flexibility and capacity to move toward newer capabilities will have to be managed carefully.
Although this report’s focus is on information technology systems, IT systems and the organizations that support them do not exist independently of the other parts of an enterprise. Developing a coherent and effective vision for IT at CMS is dependent on establishing a vision for CMS as a whole. IT should be seen throughout the agency, by Congress, and by stakeholders as a means for supporting the effective performance of CMS’s activities, businesses, and programs. In order to establish a compelling vision for IT at CMS, the agency itself must have a clearly expressed view of how it intends to (or believes it will) function in the future. Thus, CMS will need to develop a strategic plan that is broadly accepted; its strategy will be recorded in an evolving document that will require periodic updating as mandates are refined, technology progresses, stakeholder communities are engaged increasingly effectively, experience is amassed, and the health care delivery system as a whole changes.
As CMS develops a long-range vision of its role in the health care system, that vision will have to be refined iteratively over time, but core elements of the vision are almost certain to persist, forming the backbone of
new generations of IT systems. New delivery models, widespread use of electronic health records, and new kinds of registries tracking population health data and performance information mean that CMS not only will have to continue to focus on its traditional fee-for-service business model but also will have to establish an increasing focus on supporting a fee-forperformance model. This new emphasis will likely entail developing new capabilities, such as mechanisms for devising and assessing approaches to defining and collecting the measurements that will be needed to support effective comparisons of cost and quality.
More evidence-based medicine within a large-scale learning health care system17 may mean that CMS will have to be prepared to handle growing amounts of clinical information (Chapter 5 addresses this issue in more detail). CMS will increasingly find itself in the information business rather than being primarily a transaction processor. More coordination of care will mean that CMS must be prepared to look beyond medical care in the narrow sense and engage productively with data regarding broader social determinants of health—the economic and social conditions that affect people’s lives and health—in order to make valid assessments of quality and outcomes. What is needed, rather than a “medicalization” of the social determinants, is a shift from a medical model to a more comprehensive health model. A clear understanding of how this shift will or should evolve does not yet exist. However, such a transition is integral to any U.S. effort to move toward a sustainable and balanced approach to providing a sensible value-and science-driven array of health care services.18
The foundation for this vision already exists at CMS, as articulated in the “Triple Aim”:19
Improving the U.S. health care system requires simultaneous pursuit of three aims: improving the experience of care, improving the health of populations, and reducing per capita costs of health care. (p. 759)
17 See the Institute of Medicine’s Learning Health System series of workshop reports, published by the National Academies Press, Washington, D.C.
18 A number of recent reports have cited this need to consider additional socioeconomic factors in improving health care quality. See, for example, NRC, 2009, Race, Ethnicity, and Language Data: Standardization for Health Care Quality Improvement, Washington, D.C.: The National Academies Press, available at http://www.nap.edu/catalog.php?record_id=12696, last accessed August 1, 2011. See also NRC, 2011, Improving Access to Oral Health Care for Vulnerable and Underserved Populations, Washington, D.C.: The National Academies Press, available at http://www.nap.edu/catalog.php?record_id=13116, last accessed August 1, 2011.
19 Donald M. Berwick, Thomas W. Nolan, and John Whittington, 2008, “The Triple Aim: Care, Health, and Cost,” Health Affairs 27(3):759-769.
The perspectives of stakeholders are vital. Whether they be internal to CMS, other government entities, beneficiaries, researchers, or the public at large, the ultimate users of systems must fundamentally guide what CMS IT provides and how it delivers it. The future needs of the stakeholders are uncertain, even to them, and indeed, some important stakeholder communities may not currently realize their future importance as stakeholders. Their perspectives, nevertheless, must be weighed heavily as they become known. This can be done most effectively through an incremental approach in which stakeholder communities are identified and engaged increasingly effectively over time.
Information technology, medical technology, the shape and form of the U.S. health care system, and CMS’s role and mandate will always be evolving and will thus, to some extent, remain a moving picture. Within this evolving context, a strategic technology plan for CMS is needed to guide planning, acting, and making ongoing refinements based on accumulating experience. CMS’s strategic technology plan should be grounded not only in admittedly difficult projections about the future but also in a clear assessment of current conditions. Among the current technical issues at CMS that have become evident, for example, has been the historical tendency for little sharing of resources among CMS systems families, along with resulting redundancy that not only increases run-rate costs but also significantly complicates any new development—issues that CMS is aware of and has taken steps to begin to address. This type of systematic analysis should be continued and accelerated to identify not only the potential “shared services” but also the direct and indirect cost implications and finally to serve as the baseline for the modernization and transformation of systems.
Especially important are the connections between a CMS strategic technology plan and the agency’s overall strategic plan: the technology plan must explicitly tie to and support an overall CMS strategic plan and vision. In addition, IT strategic planning requires engagement and ownership at the highest levels of the organization and cannot be effectively driven solely by IT organizations within CMS. According to information the committee was able to gather, CMS recognizes the importance of this engagement. However, IT’s role cannot simply be to take orders. Although IT historically has been viewed, both in industry and in government, as a tactical resource, experience and the literature have shown that it is not possible to make strategic decisions without considering their impact on IT and the impact of IT. In this case, IT refers both to IT systems within
the organization and, in a broader sense, to technological evolution and developments in the external environment.
An effective strategic technology plan also requires an understanding of who all the stakeholders are and what their needs are, a clear statement of business objectives, and a documented and coherent view of the current state of IT across the enterprise. The plan should help maintain a focus on reduced overlap in functionality among systems, the leveraging of existing technologies, a standards-based orientation, and an agencywide understanding that the plan is dynamic and iterative.
The strategic technology plan should be based on a clear and wellarticulated understanding of the current and anticipated funding flows and structures. How are resources within CMS allocated for IT sustainment, modernization, and transformation? What proportions are allocated to each? What is the expected magnitude of the modernization and transformation workload compared to the size of the IT workforce?
Because funding for enterprise IT efforts typically is most readily available when it is tied to specific programmatic or other objectives in legislation, CMS’s strategic technology plan should be sufficiently broad and flexible to encompass opportunities for funding that might arise out of specific program mandates, but also detailed enough to serve as a roadmap for long-term modernization and/or transformation of CMS’s systems. The plan, which will of necessity be iterated as CMS’s mission and the broader environment within which it operates continue to change, should be developed incrementally, guided in part by the impacts, results, and acceptance within the relevant stakeholder communities of initial efforts and systems.
In the committee’s view, development of a CMS strategic technology plan is critical. The rationale for the development of such a plan is multifaceted— such a plan would contribute to realizing the following benefits:
• Rationalize the process of making the difficult, and necessarily long-term, decisions about systems replacement, evolution, and transformation. Without an overarching plan, effort and resources are often deployed suboptimally or even incorrectly.
• Provide context and background for the funding requests for IT, and provide a mission-driven rationale for and prioritization of individual initiatives and funding requests.
• Foster coordination of efforts to gain increased efficiencies—reducing solely program-based systems development.
• Ensure that the entire complex CMS organization and its stakeholders understand the overall direction and intent of IT use at CMS. Widespread understanding of and support for key objectives are critical to achieving the coherence in complex systems that can allow for flexibility
within and between units as they adapt to meet changing objectives over time.
• Identify long-term requirements for resources and align those resources effectively. As discussed further in Chapter 4, CMS needs to have enough in-house technology expertise to make sound technology decisions on its own without having to rely on contractors. Although it must inevitably depend on contractors for advice and up-to-the-minute information, CMS should reserve for itself final decisions that are based on sufficiently deep in-house understanding of underlying information and computer science and technology.
• Mitigate the risks in execution of IT systems modernization and transformation by recognizing current and potential future interdepen-dencies among systems.
• Facilitate the alignment of the core and contracted parties.
In addition to incorporating the principles outlined above, a strategic technology plan should also include four key components: an enterprise architecture framework, explicit priorities and a roadmap, human capital requirements, and periodic planned review and iteration of the plan itself—each discussed in more detail below.
Enterprise Architecture Framework
The environment in which CMS operates is large, complex, and challenging. CMS has multiple functions and roles, each of which is supported by one or more information ecosystems—the IT required to build, develop, operate, assess, and evolve one or perhaps multiple business functions, including the people who design, build, maintain, and operate the systems—what CMS refers to as a “family of systems.” Information ecosystems20 are complex; no single individual can understand one in totality. Different families of systems are managed by different groups; in the case of CMS, some of these groups are internal to the organization, managed by CMS employees in whole or in part, and some are external to CMS. Just as with some business functions, information ecosystems may also be externally created and maintained. The CMS “families” of related
20 Inherent in the nature of an information ecosystem is the fact that it is very broad and diverse and that it is constantly growing, encompassing and interconnecting ever more diverse entities. An information ecosystem includes various information repositories and the information-processing capabilities that build and use them. But it also encompasses entire stakeholder groups that are the clients and contributors that provide both the rationale for the information repositories and the sources of the raw data and aggregated information contained in the repositories.
and interdependent technologies are examples of information ecosystems, and they include both internal and external components.
Although one might assume that each business function maps directly to a single information ecosystem, this need not be the case. For example, within CMS, the systems that support Medicare Parts A and B fee for service include processes that rely on systems for determining eligibility, and they provide data that may be shared with other functions. A “global information ecosystem” is a notion that encompasses the entire set of information ecosystems in an enterprise.
To properly document, plan, and execute any modernization or transformation of a complex global ecosystem, it is necessary to take a structured approach. The discipline of enterprise architecture (EA) provides one such approach. A common EA framework is that of Zachman21 (the originator of the term)—and the committee’s recommended approach reflects some of the foundational concepts articulated by Zachman, namely: begin with an overall CMS strategic plan, translate its priorities to a set of future target/re-engineered business processes, and ensure that the business processes drive the information ecosystems’ modernization or transformation.
It is the complexity of ensuring that different aspects of an enterprise maintain an overall integrated approach that drives the need for a disciplined and coherent approach to IT modernization or transformation. Without such an approach, short-term or ad hoc IT programs will lead to unforeseen difficulties, will drive up lifetime costs, and will not serve the business needs of the enterprise.
The purpose of the EA discipline is to have a documented target at the ecosystem level and a target for each of the component systems of the ecosystem, all of which serve to provide a description of a coherent whole. Having a target, even while recognizing that it will move, permits components of a large and complex organization to advance in the same direction. At a more tactical level, an EA establishes target standards and rules of the road for individual component types, for instance, databases, servers, languages, and libraries.
An EA framework, once created, is used in planning and in driving prioritization. EA is also used in execution of a plan as a method for understanding interdependencies and of course, incorporating standards. A risk is that the EA function can become a bureaucratic roadblock without sufficient compensating benefit.
21 John Zachman, 1987, “A Framework for Information Systems Architecture,” IBM Systems Journal 26(3).
Explicit Priorities and a Roadmap
Prioritization is critical to the effectiveness and success of a strategic technology plan. Not every need is equal, not every item is an imperative, nor is every scenario equally likely. In addition to articulating future capabilities, a strategic technology plan must also identify current platforms that are having increasing difficulty in meeting requirements— so-called burning platforms—and must assess risks and define near-term mitigation strategies. A roadmap that articulates how an organization is expected to move from its current state to its anticipated target state is critical. For example, given that CMS already has a robust transaction processing system, enhancements to that system should be considered carefully and be made according to a well-architected, phased delivery and implementation plan. An effective roadmap provides clear direction but includes the capacity for course corrections, since flexibility is key to accommodating the kinds of uncertainties inherent in the CMS environment.
CMS should seek to leverage modernized and/or transformed systems across programs in order to increase efficiencies and reduce redundancies. Although CMS is unique in some ways—with the result that there will not always be off-the-shelf solutions to its unique issues—CMS should develop its own solutions only when other alternatives have been seriously evaluated and rejected.
At the same time, adapting to changes in technology will be an ongoing challenge. Emerging technologies should be evaluated carefully with respect to known requirements and adopted opportunistically; unproven trends should be avoided. Achieving this capability will require expansion and strengthening of in-house technological skill sets. An emphasis on adhering to open, or at least published, standards will maximize longterm benefits. A strategic technology plan can help to reinforce this goal, forestalling deployment of ad hoc or heavily proprietary solutions in the heat of demanding requirements, or in the expectation of short shelf life.
A CMS strategic technology plan will serve as a roadmap for future IT-related efforts only if everyone in the agency is, at the very least, aware of its existence and importance. It is crucial that a well-crafted strategy be recognized as a necessary tool and not viewed as an expensive doorstop.
Human Capital Requirements
Understanding human capital resources and requirements is a critical component of any strategic plan. A CMS strategic technology plan should outline what types of personnel will be needed for the future (for example, clinical informaticians, data architects, and so on) and articulate a strategy for obtaining those skills. The plan should identify which
technical skills need to be CMS core competencies, which the agency should then move rapidly to put in place. Clearly articulating a strategy and plan regarding in-house staffing versus contracting is also important. Contracting for near-term work early on to make quick progress will likely be required in parallel with building up internal capacity to deal with longer-term issues.
To most effectively meet its mandates and make progress in developing systems to support its mission, CMS should carefully consider and clearly articulate what its core technical competencies need to be. This strategizing effort should include deciding what IT services CMS must provide for itself and what can be handled by other parties. For IT products and services that might be left to others to provide, CMS should determine how best to help foster a vibrant marketplace for those products and services in the health care sector. CMS’s strategic technology plan should include explicit rationales for such decisions based on value and performance and should incorporate ongoing review to determine if current approaches are working. Chapter 4 elaborates on human capital needs in more detail.
Periodic Planned Review
Multiyear plans are inevitably subject to modification based on changing priorities and a changing landscape of technology, policy, and on-the-ground exigencies. A strategic technology plan must account for uncertainties and evolution in a rapidly evolving health care delivery and payment system as well as changes to and uncertainties about CMS’s role. Planning at CMS, as in any government agency, will also need to account for the uncertainties associated with constant changes in its mandates and expectations as expressed in legislation and rule making. Moreover stakeholder groups will change over time—in who they are, what they want, and how their importance should be weighted. There are also inevitable changes in and uncertainty about technology.
Periodic review is essential to ensuring an effective strategic technology plan. It will be important to institutionalize the process of continually reexamining and updating priorities and to implement a governance structure that can ensure resolution of conflicts and clarity of leadership to achieve sufficient momentum for action. Potentially even more challenging is coping with evolving, competing, and conflicting long-term visions and aspirations for the agency itself. Competing visions of the health care system, of CMS’s role, and of technical evolution need to be acknowledged and reflected in evolving plans.
The strategic technology plan should articulate and continually update the goals for the target ecosystems that will result from all mod-
ernization and transformation efforts. Iterative development of a comprehensive view will be ongoing and continuous: CMS should expect that the need to change and evolve its IT systems will be continuous and perpetual. There should be no expectation on the part of CMS or its stakeholders, including Congress, of an “ultimate” or finished CMS IT system. Long-range strategic planning should not be based on the expectation of such a system. CMS should instead plan for continuous iteration, and should be guided by the continuous pursuit of an increasingly broad and well-articulated global view—expressed as its strategic technology plan—of both its business and the systems that support the satisfaction of its business requirements. Chapter 3 addresses operationalizing this approach in more detail.
A forward-looking strategic technology plan is essential to coping with ongoing changes in mandates and requirements both now and in the future. However, requirements such as several provisions of HITECH and the PPACA that affect CMS in the near term, as described in Chapter 1, will have to be addressed now in parallel with the development of a strategic technology plan. Although it may be ideal to wait for completion of a strategic plan, fast-track programs developed in parallel are often required to deal with the realities of stakeholder demands. It is important, though, not to completely decouple meeting immediate needs from maintaining a long-term perspective. Sometimes it is necessary, and even wise, to take an expeditious approach in the short term, even at the potential cost of longer-term difficulties. For instance, implementing a relatively “quick and dirty” version of a new function might be reasonable in the short term to, say, resolve uncertainty about the value of and demand for that function. Such a step, however, should be taken only with full cognizance of the technical debt incurred: that is, if, for example, the function does turn out to be valued, it may be necessary to make follow-on investments to integrate the functionality into the larger strategic vision and into business and information ecosystems.
Among the possible benefits of linking long-term planning to fasttrack programs are the imposition of some pragmatism in longer-term planning efforts, which might otherwise devolve into purely theoretical exercises, and the forcing of hard decisions that can help to clarify longer-term planning, keeping it from dragging on indefinitely. Shorterterm efforts might have to be thought of as prototypes of a sort, aimed at gathering the more precise knowledge and understanding that can serve as the basis for more solidly conceived and implemented permanent solutions to be implemented in subsequent iterations.
To ensure success in addressing both tactical and strategic challenges, it is essential that fast-track efforts be focused on demands that are both well defined and reasonably constrained. In the context of a steadily developing comprehensive strategic plan, approaches such as the following might keep near-term projects on track and contribute to their effectiveness:
• Identify some services or capabilities that will be needed in a future global ecosystem and for which there is a very clear understanding of what to do, and then begin to create and instantiate these services or capabilities. Despite some risk that what appears to be clear will eventually prove to be more complicated than originally thought, a careful choice of services or capabilities is likely to forestall any dramatically wrong outcomes, and any subsequent adjustment (if necessary) would be modest.
• Take an element of the larger strategic framework and ask what pieces of it can be accomplished in less than N months. Then do those pieces, even if their place in the overall sequence is less than ideal.
• Take a plan for addressing a specific business function and ask which pieces really require an understanding of the ecosystem by focusing in a disciplined way on defining those components while at the same time moving ahead on those pieces that are truly unique to the specific system.
• Recognize that “quick and dirty” implementation of some functions can provide great business value. When explicitly acknowledged as eventually “throw away,” such efforts do not diminish the importance of disciplined, longer-term planning.
The following list of examples of fast-track efforts that might be undertaken soon is neither prioritized nor meant to be exhaustive:
• Make CMS claims data available to support assement of experimental programs internally (for example, ACO models, innovation demonstrations, and so on). The timescale for evaluation of programs, delivery models, and so on should not be constrained by access to data.
• Provide CMS claims data quickly and efficiently to qualified external organizations such as the state-level All Payer Claims Databases (APCDs).
• Provide faster access to CMS claims data for pre-pay fraud detection, which may involve better integration of Medicare and Medicaid data.
• Develop and implement process for CMS to collect needed data from clinical registries, electronic health records where they exist, and patient surveys.
The enormous challenge of moving from a siloed set of information ecosystems to information ecosystems that can support the needs of new applications and new requirements cannot be underestimated. Chapter 3 provides guidance in the form of a meta-methodology for planning a comprehensive incremental modernization and transformation of CMS systems.