In this chapter we will discuss measuring quality of care—general principles of measurement, measures for judging quality, and measures used for the local management and improvement of care—with a focus on measures of particular interest to Hospital Medicine physicians.

Measurement for Performance Assessment

The era of accountability has arrived in medicine. Consumers, purchasers, and providers of health care are increasingly interested in health care quality and safety. Health care value, in which value is a function of quality ÷ cost, is being closely scrutinized as the increasing cost of health care has outpaced inflation over several decades without a demonstrated concomitant increase in quality and safety.

Consumers want a high-value product. Selection pressures allow consumers of health care to choose between providers. Historically, the primary decision maker in assessing quality has been the individual patient, with no access to data and little understanding of how to assess quality of care. Reputation and word of mouth have been used to assess quality, and patients tend to follow the advice of their doctors when being referred for additional medical care. As health care costs increase, purchasers of health care on a larger scale look for more objective ways to assess quality. Research reflected in the Dartmouth Atlas of Health Care has demonstrated tremendous geographic variation in costs and outcomes within the United States. Paradoxically, higher expenditures on health care in the United States have not been associated with improved outcomes.

These factors provide the foundation for the push from large-scale purchasers of health care to demand that providers be able to objectively demonstrate the quality of care they provide. The Leapfrog Group, a consortium of health insurance purchasers, encourages hospitals to voluntarily report a variety of measures and makes the resulting quality ratings available to the public (http://www.leapfroggroup.org). The Centers for Medicare & Medicaid Services (CMS) and increasing numbers of commercial insurance companies provide incentives for hospitals and physicians to measure and report their health care quality. These purchasers apply selection pressures to improve quality by considering quality performance when choosing network providers or by paying differential rates for performance or participation in quality assessment. Beyond incentives, the mere fact of public reporting has influenced health care providers to turn attention and resources to improvement of quality and safety. Performance measures are markers of care used for public reporting or for incentive payment.

Organizations comparing performance commonly use one of two methods. The first is to compare one hospital’s performance against that of peer organizations. This is the method used in CMS’s public reporting and is frequently reported as a ranking among peers (often by percentile). This method allows for a direct comparison of quality between hospitals. Two cautionary notes are important. First, when there is a small range between organizations in performance, ranking may overstate differences in quality. Second, without performance targets the highest ranked hospitals have little incentive to improve, and overall performance by hospitals may be mediocre.

Another approach is to compare a provider’s performance with a benchmark; the provider either hits a target or does not. Targets are usually set arbitrarily (90% performance being a common target). When targets are set too high, care that is unnecessary may be provided to hit the targets. For example, consider the CMS Core Measure of percentage of patients with pneumonia receiving the first dose of antibiotics within six hours of arrival to the hospital. The level of scrutiny required to achieve 99% compliance with this measure may result in an unacceptable number of patients without pneumonia being treated with antibiotics in the effort not to miss a single patient with the diagnosis. In addition, this effort can divert attention away from other important aspects of clinical care.

Measurement for Local Quality Improvement

Internal audiences use measurement to improve care and to manage local performance. Pressures to improve quality may come from the external sources mentioned above or from internal sources such as risk assessments, clinical leaders, or the board. At the most basic level, constantly seeking to improve is at the core of our professional ethic in medicine. Comparison with prior performance, peers, or benchmarks can all be used for local improvement. Measurement within an organization may also serve to reduce variation. In general, reducing variation within a process leads to better outcomes, as described in Chapter 16.

Measurement forms the basis for scorecards, summaries of measures in multiple domains that are used by organizations to assess performance and to track progress. At the highest levels of the organization, a summary scorecard is commonly used to assess leadership performance toward achieving strategic goals. Components of the organization, such as sections of Hospital Medicine, can also be judged by such scorecards. See Table 15-1 for examples of candidate measures.

Without measurement, it is impossible to know if something is good or bad, the same or different. Individuals and organizations too often make decisions about health care without data. A patient who is about to undergo an elective hip replacement is likely to rely on suggestions from friends or a physician about which surgeon should do the procedure. Imagine instead if patients reviewed surgical outcomes and volumes to inform their decision. Large-scale consumers have started incorporating these types of objective quality measures into their decisions about health care purchasing and contracting. Instead of contracting with provider groups solely based on hospital affiliation and market share, they are considering things such as surgical volumes, readmission rates, complication rates, and patient satisfaction. While the science behind this measurement is still in its infancy, it is clearly moving to a more rational way of assessing care to make choices about which provider is the best for your needs.

It is also impossible to know if efforts to improve care have been successful without measuring the impact of a change. As discussed in Chapter 14, measurement is the answer to the question, “How do we know we are improving?” Too often we rush to “fix the problem” without a plan to measure the impact. It is only months or years later when the problem again surfaces that people go back and look to see why the solution was not as successful as hoped for. Without assessing an intervention, there is a real risk in wasting investment and efforts. There are some times when the luxury of measurement is not available or not needed because the results are so readily apparent. In organizations that make quality a high priority, these occasions should be rare. The default position should always be to measure.