cates that general medical–surgical nurses would likely contribute much of the data on hospital-wide nurse staffing; together, however, ICU, operating room (OR), and labor and delivery nurses could also reasonably be expected to exert significant influence on reported aggregate, hospital-wide nurse staffing levels.

This is an important point because the only source of staffing data on all types of inpatient hospital units (i.e., the California Office of Statewide Health Planning and Development [OSHPD])5 shows that ICU, labor and delivery (apart from other obstetrics), and step-down/transition units have considerably higher average nurse staffing levels than medical–surgical and other hospital nursing units (Spetz et al., 2000). These data are presented later in this chapter.

When staffing levels are based on hppd estimates6 from staffing studies that combine nurses in direct patient care positions with those in administrative or other non–direct care positions and are aggregated across multiple hospital units, nurse staffing levels such as those in Table 5-3 are produced.

The high nurse staffing levels suggested by these estimates are in contrast to the unit-specific data and direct patient care nurse-specific data produced by the Donaldson, Cavouras, and Aiken studies cited above. The higher nurse staffing levels in Table 5-3 also reflect the limitations of the available data sources on nurse staffing. The American Hospital Association (AHA) data used in several of the studies included in Table 5-3 aggregate all nursing staff (direct-care nurses and nurses in administrative positions) across all inpatient and outpatient care units, thereby producing higher levels of nurse staffing. State data sets often can distinguish nursing staff by cost center (and thereby by nursing unit), but may suffer from incomplete data. For example, in the Lichtig et al. (1999) study cited above, seven California hospitals did not submit cost reports, 26 submitted reports but did not include data on nursing hours, and 8 reported unrealistic nursing hours. Better understanding of actual nurse staffing levels is provided by studies that have examined staffing levels within specific types of patient care units.


OSHPD’s survey of hospitals is considered to be the most comprehensive in the United States and is held up as a model for other states. In spite of some limitations, it captures data from nearly every hospital in the state, and data are provided by cost center, allowing examination of care delivered by distinct care units as opposed to all hospital patient care units in the aggregate (Spetz et al., 2000).


Staffing estimates are sometimes calculated by dividing 24 hours by the number of hppd for a facility; e.g., 24 hours divided by 6.2 hppd = 3.9 patients per nurse. Other studies estimate hospital-wide nurse staffing levels using other measures, including the ratio of FTE hospital RNs to total hospital-adjusted days (Kovner and Gergen, 1998) and RN staffing per 100 occupied beds (Bond et al., 1999).

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