necessary care. Once necessary care is provided, one would expect to see no additional improvement in health outcomes from greater numbers of staff. This point is supported by the above-referenced CMS study of nursing home staffing, which identified a threshold level of nurse staffing above which no further improvements in patient outcomes were detected (CMS, 2001).
An HCO’s staffing level is traditionally considered a structural measure of quality that can affect the processes and outcomes of care (Donabedian, 1980; IOM, 1996). In nursing homes, the processes of care include a range of nursing activities, such as assistance with ADLs and monitoring of health status; therapeutic services, such as dressing changes and administration of medications; and other nursing activities, such as the management of incontinence. The outcomes of care can be measured as weight loss, pressure ulcers, incontinence, or other markers of physical decline (Zimmerman et al., l995).
In long-term care, higher staff levels and lower RN turnover have been shown to be related to better care processes, such as lower urinary catheter use, better skin care, and better resident participation rates (Spector and Takada, 1991). Inadequate nurse staffing is correlated with inadequate feeding assistance and poor oral health (Kayser-Jones, 1996, 1997; Kayser-Jones and Schell, 1997; Kayser-Jones et al., 1999). NAs with inadequate time to provide care have been documented to cut corners in order to manage their workloads (Bowers and Becker, 1992).
Schnelle et al. (2002) conducted a blinded study to determine whether there were differences in the quality of care processes among 34 randomly selected California long-term care facilities with different staffing levels. Three groups of homes were identified in the sample. Group 1 (nine homes at the 0 to 25th percentile of staffing levels) reported 2.7 mean total (RNs, LVNs, and NAs) direct-care hours per resident/day (hprd). Group 2 (six homes in the 75th to 90th percentile) reported 3.4 hprd; and Group 3 (six homes in the 91st to 100th percentile) reported 4.9 hprd. During a 3-day on-site visit, research staff used standardized protocols for direct observation, resident assessment, resident interview, and medical record review to assess 16 care processes delivered by NAs and 11 care processes delivered by licensed nurses. NAs in Group 3 homes reported significantly lower resident care loads across the day and evening shifts in 2001–2002 (7.6 residents per NA) compared with NAs in all of the remaining homes. Group 3 homes also performed significantly better on 12 of 16 care processes implemented by NAs compared with all other remaining homes combined. Residents in the Group 3, or highest-staffed, homes were significantly more likely to be out of bed and engaged in activities during the day and to receive more feeding assistance and incontinence care. The researchers concluded that there is a relationship between nursing home reports of total staffing, NA reports of resident care load, and the quality of implementa-