A significant body of literature demonstrates that the amount of nurse staffing, as well as the mix, available to provide care to patients has an effect on patient outcomes. This relationship is present across multiple health care systems, and has been shown to correlate with a number of different medical outcomes. While these findings have impacted practice and policy, most studies of the relationship between nurse staffing and patient outcomes are still conducted at the level of the entire hospital, which averages patient outcomes across a wide variety of staffing levels. Detail at the unit level is essential to understanding the impact of nurse staffing variability, and to designing effective interventions. In one of the few studies assessing staffing at the unit and shift level over time, and using the difference between staffing targets and actual staffing over time, the gap was significantly associated with differences in patient mortality. Such a measure of nurse staffing allows organizations to set staffing target based on patient, unit, and organizational context, and design appropriate staffing interventions while still providing a comparable staffing metric. Health care systems struggle to define recommended or targeted levels of staffing. In response to national concerns about system practices that impact patient outcomes, the VA Office of Nursing Services (ONS) promulgated a Staffing Methodology (SM) directive in 2010, requiring the use of unit and facility level review of staffing and patient data to formulate custom recommendations for target staffing levels on each shift. The proposed study fills a gap in our understanding of how levels of nurse staffing affect patient outcomes. Variation in nurse staffing level persists at both the facility and unit level. It is not clear, however, how changes in staffing hours and staff mix at the unit level within facilities affect outcomes for patients. Critical questions about the type and quantity of nursing hours actually delivered to patients are not known, nor is the effect of variability in the gap between target and actual known. This project will use regression models to extend the current knowledge of how nurse staffing affects patient outcomes (mortality, failure to rescue, infections related to catheters, and risk-adjusted length of stay) with two major objectives. Aim One: Use the extensive VA data systems to conduct the most rigorous study to date on how nurse staffing levels and nurse affect patient outcomes. This will be the first large-scale examination of how shift- level staffing affects patient outcomes, AND the first large-scale study to simultaneously use unit-level data with controls for staffing levels, characteristics of the nurses, extensive patient risk-adjustment, and control for use of other health care staff, including physicians. We will examine the gap between targeted and actual nursing hours for each shift for the 66 VA facilities that use the AcuStaf nurse staffing program. We will also assess whether shift-to-shift variations in the characteristics of the nursing staff modify the impact of differences between targeted and actual hours. Aim Two: Carefully test the robustness of findings to alternative empirical specifications, levels of data aggregation, and inclusion of specific types of variables. Our goal is to significantly improve our understanding of how nurse staffing impacts the outcomes of inpatient care for Veterans. We will provide much more detailed data than is currently available on how nurse staffing affects patient outcomes. These results will extend beyond VA as they will represent the most definitive study of nurse staffing and patient outcomes to date, giving VA the best evidence available about how to make staffing decisions. These results will inform the VA Staffing Methodology to help insure that each shift is adequately staffed, with the expectation that appropriate nurse staffing will yield better patient outcomes. To facilitate dissemination, we will work with ONS to create presentations of the results, including clear presentations of how changes in nurse staffing (and the associated nursing costs) affect patient outcome.
NIH Reporter Project Information
None at this time.
TRL - Applied/Translational
Models of Care, Organizational Structure, Outcomes - System
None at this time.