PART 2: Measuring quality of performance: What’s in store for 2008 and beyond

Critical gaps and weaknesses in the quality enterprise are being addressed in a coordinated effort to set priorities and goals

President and CEO, National Quality Forum, Washington, DC

Senior Vice President for Performance Measures, National Quality Forum, Washington, DC

Part 1 of this article described the history of a national effort to establish a quality measurement framework and the many players who have contributed to the process. Good progress has been made, but important work lies ahead. In Part 2, we detail next goals and the steps we must take to achieve them.

  How national priorities and goals for quality improvement would help

Providers cannot measure and improve in all areas at once. A coordinated national priority and goal-setting process is needed to identify a limited number of high-leverage areas having the greatest potential to enhance quality and slow the rate of growth in health care expenditures. The absence of national priorities and goals for quality improvement is a basic impediment to the development of a coherent national system of performance measurement and improvement.

Correcting imbalances in measure development. In the absence of a national process, each measure developer sets its own agenda for development. Most measure developers continue to focus on a specific clinical area such as diabetes care, or on a provider setting such as the hospital.

Most measures tend to focus on disease-specific process measures (eg, annual foot exam for diabetes), as opposed to intermediate outcome measures (eg, level of glycosylated hemoglobin for diabetes control). Critical areas for which no individual provider is solely accountable tend to be overlooked. Examples are care coordination, longitudinal care of chronic illness, and end-of-life care.

Too little attention is paid to measures of patient engagement in decision making, and patient knowledge and preparedness to implement their treatment plans. Inadequate attention is also paid to high-volume, high cost procedures, and to measures of “overuse” as well as “underuse.”

Reconciling the needs of different stakeholders. A number of important stakeholder groups have legitimate vested interests in setting their own priorities. The AQA and HQA are important decision-making bodies that prioritize and recommend measures to CMS and other purchasers for use in public reporting.

Professional societies and organizations involved in accreditation and certification also have measurement agendas that fit their own evaluation objectives. And consumers and private-sector purchasers, and those organizations that represent these constituencies (e.g., the Consumer/Purchaser Disclosure Group, the AARP, the Leapfrog Group, the National Business Coalition on Health), each have efforts underway aimed at setting national performance improvement agendas.

Setting realistic expectations. A coordinated national priority and goal-setting process is needed to a) focus provider and practitioner attention on a core set of high-leverage areas having the greatest potential to enhance quality and slow the rate of growth in health care expenditures; b) systematically raise the bar of performance expectations over time; and c) assure the efficient and effective deployment of scarce resources for measure development and quality improvement. The national priorities and goals are meant to serve as a core set of key areas for quality improvement; they do not preclude the identification of additional focus areas by particular specialty groups, communities, or other stakeholders.

Modest steps are now being taken to establish an initial set of national priorities and goals. In 2008, NQF will convene for the first time a national advisory panel, National Priorities Partners, to establish a “National Priorities Agenda.” Co-chaired by Margaret O’Kane of NCQA and by Don Berwick of the Institute for Healthcare Improvement, the National Priorities Partners includes representatives of national organizations (found at www.jfponline.com). Because of the roles the partners play in fashioning payment incentives, public reporting, and quality improvement and professional development programs, this panel should have the leverage needed to encourage the health system to focus on achieving set priorities.

  Coordinated measure development

Due to the sizable number of organizations and resources involved in measure development, coordination is critical. NQF has been working collaboratively with measure developers to create guidelines and common conventions for measure development and testing, and much of this work is now reflected in NQF’s measure submission requirements.¹

Through its endorsement process, NQF is now requiring a greater degree of measure harmonization. For example, whenever possible, measures roll up (eg, post-surgical infection rates should be calculated the same way for surgeons and hospitals); measures are setting-neutral (eg, measures of pain management should be the same for nursing homes and home health patients); and measures related to specific groups apply common conventions (eg, paired process and outcome measures for patients with depression should employ the same denominator population for each measure).

Numerous methodologic issues in the performance measurement arena also require continued attention and action. For example, performance measurement at the physician level frequently encounters the following concerns:

• Sample size: individual physicians may not see sufficient numbers of patients with a particular condition to provide an adequate sample size.

• Risk adjustment: current methods of risk adjustment may not provide fair comparisons among physicians.

• Exclusions: the absence of a standardized approach to excluding patients from eligibility for specific quality measures could lead to “gaming” on the part of providers.

• Attribution: the inability to attribute specific care to individual physicians or group practices limits the effectiveness of performance measurement.

• Measurement mode: we must better understand the effect of measurement using administrative data, clinical data, or data derived from electronic health record (EHR) systems. The lack of clinical detail in most administrative data may make it difficult to allow comparisons with measures based on medical record review. For example, a higher level of clinical detail is often needed to identify patients who should be excluded from a measure.

• Improvability: measures should move us beyond proving basic competency to achieving a higher level of performance with improvability. Quality measures frequently assess documentation of important aspects of care (eg, were discharge instructions given?), rather than the actual care delivered (eg, did the patient understand the discharge instructions?). We also need a system that retire certain quality measures that have reached a performance ceiling.

• Unintended consequences: physicians might practice “cherry picking,” to avoid the sickest patients and thereby appear to achieve better performance.

Moving toward composite measures. As the number of measures has increased, so too has interest in the development of composite measures or summary metrics. Much work is ongoing to determine how best to construct composite measures that will be most meaningful to potential users.

Expansion of electronic health records. Lastly, over the coming 5–7 years, many if not most health care providers will switch from paper to electronic health records (EHRs). EHRs pose enormous opportunities and challenges. Ultimately interoperable EHRs and personal health records (PHRs) should ease the burden of data collection, support real-time measurement and performance feedback/decision-making, and open up opportunities to measure aspects of performance for which we currently lack documentation (eg, handoffs, patient compliance).

While many experts expect EHRs to serve as the performance measurement backbone of the future, much work is needed to precisely specify the high-quality data elements needed for performance measurement. This is a high priority of America’s health information community, and various collaborative efforts between the performance measurement community and the EHR community are underway to develop EHRs that support performance measurement and improvement. These efforts include public-private partnerships such as the Quality Workgroup of the American Health Information Community.²

A future issue of CCP will carry more in-depth coverage on using EHRs, registries, merged data sources, etc, to collect information on quality measures.

References

1.  National Quality Forum, “Measure Submission Guide,” www.qualityforum.org/pdf/Measure%20Submission%20Form-Version1.1.doc, accessed September 19, 2007
2.  Health Information Technology, Quality Workgroup, www.hhs.gov/healthit/ahic/quality, accessed September 19, 2007) and various efforts by the AMA, NCQA, and NQF.