in the Quality Chasm report. The most important gaps identified by the committee are measures of efficiency, equity, and patient-centeredness.

Efficiency

Substantial work is under way on the development of measures of efficiency that can represent the value of medical care. Prior Institute of Medicine (IOM) studies have endorsed the basic concept of avoiding waste: a more efficient care process or delivery system will produce an equal or better outcome at lower cost. The key is to be able to measure both quality and resource use for well-defined episodes of care. The following principles guided the committee’s thinking in this area: (1) measures of efficiency should be based on episodes of care of adequate duration so that the quality of care and/or outcomes of treatment can be reliably determined; (2) the scope of services and time window of observation should be broad and long enough to ensure that providers being evaluated cannot improve their apparent efficiency simply by shifting costs to other providers or to periods outside the window of observation; (3) multiple measures of efficiency (i.e., of costs and quality) for a given provider are preferable because performance may vary across the types of service provided (e.g., care for diabetes versus congestive heart failure); and (4) when possible, reliance should be placed on measures that have been reported in the peer-reviewed literature to enhance both affordability and validity.

Two broad types of efficiency measures warrant consideration: longitudinal and episodic. The committee recommends an aggressive research agenda to develop and pilot test efficiency measures of both types.

Longitudinal efficiency An example of measures of longitudinal efficiency for defined populations over relatively prolonged periods is 1-year mortality and resource use for acute conditions. The feasibility of collecting these data has been demonstrated for different types of care delivery systems (Tarlov et al., 1989; Ware et al., 1996), hospitals (Fisher et al., 2004; Guadagnoli et al., 1995), and for regional care systems within the United States (Fisher et al., 2003a,b). Such data have also been used to monitor the impact of the introduction of a prospective payment system on hospitalized patients (Kahn et al., 1990). In addition to the measure of longitudinal efficiency recommended by the committee for the starter set of measures detailed in Chapter 4—1-year mortality, resource use, and functional status after acute myocardial infarction—attention should be paid to collecting long-term follow-up data on additional conditions for which longitudinal outcomes and costs can be reliably assessed. Candidates include hip fracture and colorectal cancer, given their relative frequency, the high rates of hospitalization associated with these conditions (allowing population-based comparisons of outcomes at the com-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement