L. Allen Dobson, Jr, MD; Marcia Guida James, MS, MBA; Margaret E. O’Kane, MPH; Peter Salgo, MD; and Jed Weissberg, MD
Continuous improvement is a core tenet of the quality enterprise, and measurement is necessary to gauge improvement. Therefore, there is a growing emphasis on measurement throughout the healthcare spectrum.
To meet the demands of the quality enterprise, the measurement field must continue to expand and evolve. Healthcare reform legislation will accelerate and shape the trajectory of that evolution. Along the way, significant barriers to proper measurement must be addressed.Measuring Quality: Challenges and Limitations
In the reform era, measuring and gauging quality are essential but problematic. The major barriers are discussed below.Inherent system complexity equals variations, lack of consensus, and lack of coordination
The US healthcare matrix is vast and extremely complex. Complex problems defy simple solutions. There are few national consensus measures. Existing quality-related definitions, approaches, mandates, and measures vary widely between stakeholders and from place to place, and often overlap and conflict. At present, the federal government (eg, CMS, Agency for Healthcare Research and Quality [AHRQ]), the Joint Commission, the NCQA, professional associations and societies (eg, AMA’s Physician Consortium for Performance Improvement), private corporations, and individual researchers have all developed measures used by health plans, hospitals, physician practices, and long-term care providers to compare performance.
Although some stakeholders are capable of measuring their own performance, they measure only against themselves, and information is seldom shared. Moreover, because quality standards are still evolving, many hospitals and health plans distrust the available data and have not committed to the systemwide measurement effort, although the PPACA will require sharing and measurement. To some extent, mistrust and reluctance to share might be understandable.
Quality Measurement: 10 Key Areas of Evolution
Patient-level outcomes—better health
• Linking health risks and outcomes
• Avoiding complications (eg, hospital-acquired infections, medical errors, and medication errors)
• Patient-reported functional status, health-related quality of life, and experience of care
Processes of care—better care
• Technical effectiveness
• Coordination of care and transitions to care settings
• Alignment with patient preferences: shared decision making
Cost and resource use (overuse, misuse, waste)
• Total cost of care across episodes of care
• Indirect cost: employee absence or reduced productivity
Developing ways to aggregate complex information for measuring continuumwide performance (eg, composite metrics, efficiency data, deltas, dashboards)
Measuring disparities throughout the continuum (eg, stratification, clinical and socioeconomic status risk profiles)
Harmonizing measures across sites and providers
Shared accountability across patient-focused episodes of care
Meaningful use of HIT and clinical decision support
Measures for patients with multiple chronic conditions
Measuring and reporting performance systemwide
Under a VBP construct, reporting data and quality results can impact provider compensation.Moving targets
In the dynamic and evolving reform environment, “best practices,” “quality,” and “quality improvement” are moving targets. Definitive quality solutions are therefore elusive.Measuring at the population level
Patient-centered care is a cornerstone of reform and VBP. “After all,” said Ms James, “patients are on the receiving end of the care we deliver.” For overall reform to succeed, measurement must extend beyond individual patients to the patient population, the community, and beyond, to wider geographical regions and the entire nation.Unaligned legacy system
The measures currently used in the quality enterprise were developed for other purposes, and are not readily adaptable for publicly reporting and rewarding quality performance. Although the legacy system does report some performance measures, those measures are not necessarily aligned with identified national priorities. The reverse is also true: some national priorities have been identified without corresponding performance measures.Lack of/incomplete information and metrics; limits of HIT
Randomized controlled trials are the gold standard of evidence- based medicine; their results inform best practices for attaining optimum outcomes. However, the evidence base is lacking, incomplete, or mixed for many conditions, and especially for patients with multiple comorbid conditions. In those instances, discrete metrics are either unavailable or not yet feasible. For example, most trials exclude the frail elderly with multiple comorbidities, who often react to treatment differently from healthy young people or people with a single condition.
Even when the evidence base is sufficient, developing meaningful quality metrics is time-consuming, expensive, and difficult. Multiple, disparate data points must be aggregated and integrated. EHRs and other HIT should help resolve the problem. “The evolution of quality management pivots on automated records,” said Mr White. “That’s where many outcome measures can be found—not just outcomes for individual patients, but for an entire practice or population.”
The quality enterprise poses nuanced performance questions. To answer them, HIT must be configured to identify gaps in care, report outcomes in the same format from one record to the next, and help improve practice- or populationwide patient management. However, full HIT capability is not yet widely available or easily linked to other essential data, and current EMRs fall short. In 2000, RAND Health published quality measures specifically for frail elderly patients that took their multiple health issues into account. “A great measure set,” said Dr Weissberg. However, those measures are not readily extractable from claims data or EMRs. Dr Dobson commented, “I need—but current EMRs don’t provide— all the information I need to manage all of the patients I care for, whether or not they’re physically sitting in my office.”
PDF is available on the last page.