Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120
Presentation
The association between costs and nurses' perception of practice environment in acute inpatient nursing
Background/Objective: Several studies have assessed the association between nurses' perception of their practice environment in inpatient settings and patient outcomes. To date, no studies have assessed the association between nurses' perception of their practice environments and costs of care. It is reasonable to hypothesize that practice environments that are perceived as more favorable have higher costs than less favorable environments. Our objective was to assess the association between nurses' perception of their practice environments and costs of care.
Methods: We used data from two previous studies, one in which we administered a mailed survey to all nursing personnel in 124 VA medical centers to assess nurses' perceptions of practice environment, among other elements of nursing care, the other in which we obtained cost data from VA administrative databases to assess costs in acute medical-surgical non-intensive care inpatient nursing units in the same hospitals. We assessed practice environment using the five scales in the Practice Environment Scale (PES) endorsed by the National Quality Forum. We merged these files using unit-level identifiers. We have completed bivariate analyses at the unit level and are currently completing patient-level multivariate regressions using hierarchical modeling to control for clustering by nursing unit and hospital.
Results: The survey had a response rate of approximately 30% among registered nurses (RNs). We selected nursing units with over 5 RN respondents, for a total of 99 acute, non-intensive inpatient care nursing units. Mean total cost per day was $1821 (std. dev. 614), while mean nursing cost per day was $563 (s.d. 134). Mean RN hours per patient day were 4.8 (1.6), and mean RN skill mix was 64% (12). Mean values of the five sub-scales of the PES ranged from 2.0 to 2.8 on a 0 to 4 scale. Of the five sub-scales, only one, the satisfaction with staffing sub-scale, was significantly correlated with either cost variable. The correlation was 0.32 (p = 0.0013), suggesting that as satisfaction with staffing increases, nursing costs per day also increase. All subscales were significantly and positively correlated with RN hours per patient day, despite the lack of correlation with nursing costs. There were no significant correlations with RN skill mix.
Conclusions: It is noteworthy that there were relatively few significant bivariate correlations at the unit level between RN perceptions of unit practice environment as measured by the PES, and costs of care. Assuming these null findings persist as we complete the patient-level multivariate hierarchical analyses, this suggests that it should be possible to improve RN perceptions of unit practice environment without significantly increasing costs. However, the significant positive correlations with RN hours per patient day suggest that the null findings with respect to cost may be related to complex interactions between costs of care and nurse staffing.