Towards a New Era in Diabetes Quality
Brian Leas, MS, MA, Kathryn M. Kash, PhD, Bettina Berman, BS, RN, Albert Crawford, PhD, Richard Toner, MS
Author Affiliations: At the time this project was conducted and the manuscript drafted, all authors were affiliated with the Thomas Jefferson University School of Population Health, Philadelphia, PA
This research was funded by the National Changing Diabetes Program© of Novo Nordisk, Inc.
Disclosure: The authors have no conflicts of interest to disclose.
The manuscript that follows was prepared in September 2009, but never submitted for publication. However, the findings and themes may be informative, especially in the context of renewed efforts to change the landscape of quality measurement in healthcare. Questions or comments are welcome: .
Abstract
Diabetes care has been the focus of a tremendous volume of performance assessment and quality improvement efforts in recent years, but many observers are dissatisfied with the extent of the resulting impact. Through a series of key informant interviews with health policy leaders across the U.S., we identified five themes which may need to be addressed to achieve more significant long-term improvement in the quality of care delivered to patients with diabetes. These elements include: 1) strong national leadership that has credibility and can build momentum for change; 2) consideration and potential adoption of new frameworks for care, such as the patient-centered medical home and the population health paradigm; 3) development of a coherent measurement strategy that can reduce the burden of measurement while also addressing key gaps in current performance assessment initiatives; 4) establishment of national data sources and health information technology embedded with tools to support quality measurement and improvement; and 5) reforming healthcare financing to help add value to the most significant aspects of diabetes care.
Introduction
Diabetes imposes a severe clinical and economic burden on the United States and around the world. Nearly 24 million Americans currently have diabetes,1 a number expected to double in the next two decades.2 Fifty-seven million Americans have significant risk factors for diabetes1, which in turn increases their risk for heart disease. Costs associated with diabetes are enormous, with recent studies estimating the total economic burden at $218 billion annually in direct and indirect costs in the United States.3-5
The impact of diabetes can often be controlled through management of risk factors and behaviors, and monitored by tracking indicators of disease severity. For patients with diabetes, clinical guidelines are well established and application of them can result in decreased morbidity and mortality.6,7 For those at risk for diabetes,appropriate primary prevention strategies can avoid or delay onset.8 But despite all these tools, patient outcomes continue to disappoint. The acknowledged need for improvement hasled in part to a multitude of efforts designed to routinely assess the performance of healthcare providers. Quality metrics have been developed for dozens of conditions, but because of the impact of diabetes on individual patients, health care delivery systems, and national economies, it is among the most frequently measured.
Developing measures is a crucial first step towards improved patient outcomes, but many factors affect how measures are put into practice and whether they achieve their intended results. In a recent study,we examined the current state of diabetes quality measurement, focusing on the breadth of measures and how they are gathered and utilized.9While conducting the research we interviewed a broad spectrum of leaders in diabetes care, performance assessment, and quality improvement, who identified significant themes that point to major challengesand opportunitiesfor moving forward diabetes quality measurement and improvement. These themes are presented in detail below, and possible directions for change are addressed.
Research Design and Methods
As part of our study of diabetes quality measurement weperformed an environmental scan to identify organizations that develop, promote, or use diabetes quality measures, including providers, payers, purchasers, patients, and government agencies. Representativestate-level and regional organizations that demonstrated significant innovation addressing diabetes quality were also included. Following the environmental scan, we conducted interviews of leaders within these organizations that are familiar with quality measures and improvement strategies. Between July and October, 2008, we interviewed 21individuals representing 17 different organizations. A full list of interviewees’ organizations is included in Appendix 1. Leaders within other organizations provided input during and after the interview process, but were not formally interviewed; these include the American Diabetes Association, American Association of Clinical Endocrinologists, and Agency for Healthcare Research and Quality
Interviews were semi-structured and lasted between 30 minutes and one hour. Most interviews were conducted by a two-person team and on the telephone, and a few were completed face-to-face or by a solo interviewer. Interviewees were promised individual confidentiality and were not compensated for their time. Two interview protocols were used. A broad version designed for institutions that develop measures addressed seven topics: 1) organizational motivation for assessing diabetes quality; 2) experience developing and using measures; 3) efforts to implement measurement or promote utilization; 4) perception of the strengths and weaknesses of specific measures; 5) thoughts about the impact of quality measures; 6) effects of incentives or disincentives associated with measurement; and 7) expectations and recommendations for the future use of diabetes quality measurement. A briefer version intended for stakeholders that do not create measures consisted of the latter four subject areas. The protocols are summarized in Appendix 2.
Results
All of the interviewees identified diabetes as a condition ideally suited to quality measurement and improvement, often citing its prevalence, morbidity and cost as reasons for its prominence in quality initiatives. Proper diabetes management, which includes monitoring a wide variety of clinical indicators, also translates readily to process and outcome measurement. Many interviewees considered diabetes quality improvement efforts to be among the great successes of the quality movement because they have been associated with increased tracking of key clinical processes, such as measuring hemoglobin A1c and performing foot exams, and improvements in some clinical outcomes, such as reductions in the rate of lower extremity amputations and better control of blood glucose levels.10-12
Despite these positive developments, they also noted that diabetes prevalence and costs continue to rise, while disease severity and complications have remained largely unchanged. Interviewees described an environment that is overwhelmed with measurement and quality improvement efforts. For example, physicians and other stakeholders reported that it is increasingly difficult to chart a clear path for developing effective and accountable care processes. All of the respondents felt that current efforts might yield incremental improvement in some processes and proximate measures, but will not result in long term improvements in patient outcomes. Several strategies for addressing these challenges were suggested and are presented below.
Five noteworthy and related themes emerged from the interviews and point towards possible directions for change. Many of the interviewees described the need for 1) unified leadership; 2) a guiding vision; 3)a coherent measurement strategy; 4)valid data at the patient level and on a national scale; and 5) resource allocation aimed at promoting better practice. Each theme is explored in detail below. Many of these themes have been addressed in the recent literature.13,14 Improving diabetes outcomes will likely require a distillation and integration of the following current streams of thought on quality of care.
Leadership
Many stakeholders have a strong interestin diabetes quality, and this diversity was represented by the interviewees, whoincluded: specialist and primary care physicians, other caregivers and related health professionals, health systems, insurers, employers, government agencies, quality measurement and improvement institutions, and patient advocates. Each of these groups has a unique set of constituents, operating principles, goals, strategic priorities, and funding sources. Many have a long history of varying interests and objectives that are not necessarily aligned to support quality improvement. Interviewees believed that these differences havehinderedthe coordination of strategies and dampenedcollegial interaction. “Stakeholders bring their personal perspectives to the table”, one person reported, noting that resolving these differences is a major challenge. Another person felt that “multi-stakeholder efforts are necessary, but current initiatives have been focused on the politics of a small community” of actors.
Reflecting these distinctions, leadership at the national level was reported to be fragmented. The National Quality Forum (NQF), charged by a Presidential Commission to integrate health care quality initiatives, has sought to serve as the arbiter of quality metrics. The Ambulatory Care Quality Alliance (AQA) aims to play a central role in guiding the practical use of quality measures.15 The National Committee for Quality Assurance (NCQA) spearheaded measure development at the health plan level, and has expanded its efforts considerably to include assessment of individual physicians and other entities. The leadership of these and other institutions is notable and many stakeholders look to them for guidance. Nevertheless, perspectives on these organizations varied considerably in the interviews. Some respondents, for example, expressed frustration that NQF measures are not universally accepted as the norm (“NQF should be the key,” said one person), while others voiced concern that NQF is “getting too close to measure development” rather than endorsement. Similarly diverse views were expressed about many of the organizations whose roles we examined.
Nevertheless, working relationships between many of these stakeholders have in factresulted in genuine collaboration around important initiatives in recent years. For example, several high profile measurement and improvement organizations, such as the NCQA16 and NQF,17 intentionally create advisory boards populated with multiple interests to stimulate collaboration and generate broadly supported measurement activity. The recent introduction by the Centers for Medicare and Medicaid Services (CMS) of its Physician Quality Reporting Initiative (PQRI) features prominent roles for the American Medical Association, NCQA, NQF and others in development and testing of its measures.18Although it is too early to determine how PQRI is impacting care, some interviewees have seen stronger collaboration result. “PCPI [AMA’s Physician Consortium for Performance Improvement] is a good multi-stakeholder process”, one person reported, and “we’re getting along better” were typical comments, thoughwe were also told “we still have a ways to go.”
Against this complex backdrop of competing organizational priorities and growing support for collaboration, interviewees discussed the need for stronger national leadership in the quality arena. Many sectors felt that such leadership should come from within their own ranks, while others asserted that responsibility lies elsewhere. Balancing these attitudes remains a challenge for NQF and other groups that seek to lead. Many respondents hoped to see a new force for diabetes leadership in the near future. They suggested that this might be served by a broad-based group in the model of the former National Diabetes Quality Improvement Alliance (a multi-stakeholder initiative that helped yield consensus on many of the current quality measures), or it could be guided by government, with an individual or office charged with such a task. While defining the exact nature or parameters of leadership was not a purpose of these interviews, several people suggested that such an individual or entity might best be positioned externally to the organizations currently in the quality domain, and thus outside their history and politics.
Vision
One of the major priorities for new leadership would be to develop a long term vision for the future of diabetes care. For many years disease management concepts and the chronic care model generated a range of strategies and programs for addressing diabetes and other chronic illnesses.19,20 These approaches place significant value on patient self-management, and redesigning care systems to incorporate evidence based medicine and support clinician decision making. Many respondents, though, questioned whether this remains the best overall approach, as prevalence, morbidity and costs associated with diabetes continue to rise. “We treat patients and populations, not diseases,” one person said. “Patients are more complex than one disease,” said another. Most interviewees suggested that it is time for health care providers to consider and embrace new care concepts in order to generate greater progress and respond to changes in the broader health care environment. Two such models which were cited repeatedly by our respondents were the patient-centered medical home, and thepopulation health paradigm.
The patient-centered medical home aims to restructure care delivery and provide a full array of primary care services to patients by focusing on the role of a primary care provider, and strengthening linkages between the patient, primary care physician, other physicians, and the patient’s family.21 Built on the recognition that primary care is crucial to positive health outcomes,22 the medical home model seeks to coordinate care across multiple providers and settings, while improving the safety and quality of services. Patients with diabetes receive optimal care when they have a broad team of providers, including endocrinologists, ophthalmologists or optometrists, diabetes educators, and others, but the key role of the primary care physician was frequently cited by the interviewees. “Coordination of care provided by a primary care doctor is the key” to improvement, said one respondent, and this sentiment was echoed by others. This approach relies in part on information technology and patient registries, and also reflects many of the core tenets of the chronic illness model. The medical home has gained enormous support from leading organizations nationally, including NQF, NCQA, and physician professional societies.
The population health model has simultaneously begun to gain ground as an established framework for addressing clinical care, guiding performance assessment, and setting health care goals.23-25The crucial aspect of the model as it relates to diabetes quality is its linkage of primary health determinants with long-term health outcomes across a given population. These two factors – determinants and outcomes – are at the center of the vision our respondents articulated for diabetes care.
Understanding the determinants of health for diabetes is critically important to its management. Determinants include a range of factors such as those used to develop clinical guidelines and can include things like health behaviors, access to care, and socioeconomic considerations. Our evaluation of diabetes quality measures identified these areas as noticeably lacking in useful metrics, and the interviews confirmed this assessment. Health behaviors may be particularly important because they are directly linked with risk factors for the onset and progression of diabetes. Patients’ behavior, along with their access to care and socioeconomic status, can also impact the availability and success of primary and secondary prevention strategies. Factors which increase the risk for diabetes also have broader relevance as key determinants of the risk for heart disease. Diabetes is itself a risk factor for heart disease, and several respondents emphasized that a population health model incorporated in a medical home enables physicians to devise treatment plans that fully appreciate the interrelationship between all of these components.
Health outcomes are also a critical component of the population health model of care, and a focus on outcomes was specified by many of our interviewees as a significant gap in current quality measurement. “Measures that matter reflect the big picture” of patient care, one explained. Proximal outcomes of treatment, such as hemoglobin A1c values, are often the main focus of performance assessment. Rates of extreme long-term outcomes such as blindness, lower extremity amputation, or mortality are tracked with much less frequency. Interviewees felt that these outcomes, which are incredibly important to patients, could be tracked more consistently and incorporated into measurement and improvement activities. Taken together, the patient-centered medical home and a population health model are two approaches that hold promise for redesigning care for patients with diabetes and other chronic illnesses.
Coherent Measurement Strategy
Our research highlighted the enormous breadth and volume of current diabetes measurement. In conducting a review of diabetes metrics developed by leading organizations, we found nearly 150 different measures. Many of these were variations on key clinical indicators such as hemoglobin A1c, lipids, and blood pressure, although more than twenty different aspects of diabetes care measurement were identified. Often these measures differ from each other in their specifications, such as the selection of a target goal, or criteria for including patients in a data set, or setting a timeframe for measurement. For example, we identified 21 measures of hemoglobin A1c, and 22 ways of tracking lipid management.