Minding the Gaps

The desire to improve care by holding clinicians accountable for the quality and costs of care delivered is a theme running through recent health policy proposals and changes. Providers report on a flurry of quality measures that tend to be based on discrete patient-provider encounters. But these snapshots measure fragments of care that cannot easily be pieced together to fully assess a patient's health care experience.

In a recent RAND study, researchers posit that it is time to focus health care quality measurement efforts on rendering a more complete picture of care. By measuring on the basis of an episode of care — management of a condition through diagnosis, treatment, and subsequent follow-up — data on patients' care and experiences could be more comprehensive.

RAND researchers worked with medical professionals to investigate how to construct episode-based care measurement. They started by mapping episodes of care for six conditions and identifying where quality metrics would be applied. Significant gaps existed; there were no quality measures for large portions of the care episodes. Most current metrics focus on specific events or items, such as recovery from a surgical procedure or blood work values, instead of on broader concepts, such as clinical improvement or degree of control over one's disease progression.

With the gaps identified, the researchers constructed five categories of quality measures:

• what happened (outcome measures)
• how the patient's health changed (delta measures)
• the amount of time for which a patient's health was less than optimal (integral measures)
• care for which quality depended on previous events (contingent measures)
• groups of measures taken throughout the episode (composite measures).

Some of these categories, such as outcome, already have measures available, but others do not. Ideally, those who develop and test quality measures would use this analysis to create measures to fill the gaps so that care can be evaluated over time and across settings in a comprehensive manner.

The researchers acknowledge that, in some cases, the data may not be available yet to feed into each type of measure. But they argue that using this framework in the context of the many changes in health care could help guide development of such efforts as electronic health records and modeling of new payment and care delivery methods. Implementation of these quality measures would require substantial investments in data capture, analytics, and process design, as well as consensus on the episode-based framework. When it comes to these measures, though, incremental progress may be the best way forward.