dental care, skin condition, and medications (Morris et al., 1990). Because many persons tend to stay in nursing facilities for considerable lengths of time, often for months or years, process measures tend to assume greater importance than they do in hospitals, where the average length of stay is 7 days (Kane, 1988; Kane and Kane, 1988).
A number of studies of nursing home quality have examined process measures (Zimmer, 1983, 1989; Zimmer et al., 1986). Some of these measures describe how personal services to residents are provided. These measures include help with activities of daily living (ADL) and provision of special services. At the same time, in high-quality institutions, staff avoid overuse of psychotropic medications (chemical restraints) and physical restraints. Critical to provision of high-quality care is a patient-specific care plan. Finally, residents have basic rights that society accords to other individuals. Thus, these rights also constitute elements of quality captured by the process measures.
The outcomes of nursing home care include changes in health status and conditions attributable to the care provided or not provided. Outcomes of long-term care are "most fairly expressed in terms of the relationship between expected and actual outcomes." For some nursing home residents, realistic expectations for the outcomes of care may be maintained levels of health or slower-than-expected rates of decline, rather than improved health (R.L. Kane, 1995, p. 1379). The currently used measures of outcome include global measures such as mortality rates and rehospitalization rates (Lewis et al., 1985; GAO, 1988a,b; Spector and Takada, 1991); summary measures of functional status; and specific indicators such as incidence of facility-acquired pressure sores and urinary incontinence (Nyman, 1989b). Satisfaction of both residents and their families are also quality indicators because nursing home care and professional performance encompass more than the provision of technical services (Hay, 1977). Ultimately, determining the expected and actual outcomes of care for nursing home residents will require sophisticated and increased attention to assessment of individuals' initial health status, quality of life, sociodemographic characteristics, and the nature of treatment provided (e.g., palliative or curative), with the goal of determining the outcomes attributable to treatment after controlling for other variables (R.L. Kane, 1995).
Nursing home residents and the primary missions of nursing homes vary, as well as the way in which variations affect how specific quality-of-care measures should be interpreted. At the risk of oversimplification, there are three types of residents: (1) those who use the facilities for recovery and rehabilitation following an acute hospital stay; (2) the terminally ill; and (3) persons with multiple chronic conditions and cognitive and functional impairments who are expected to stay in nursing facilities for the rest of their lives. The second and third types of