Understanding the factors that directly influence hospital staffing will help finance managers make the tough decisions about resource allocation that are needed in today's highly competitive value-based market.


The fact that labor costs constitute more than half of all hospital operating expenses creates a significant opportunity for hospitals to use labor cost management to preserve operating margins under current fee-for-service payment. Based on this premise, we performed an analysis that delved into data published by the Centers for Medicare & Medicaid Services (CMS) to identify financial and operational factors that influence staffing levels (measured by FTE/adjusted occupied bed) in hospitals, allowing for ownership characteristics and case mix. Using common statistical tools and techniques such as linear regression, our analysis sought to quantify across a large number of organizations the strength of the association between staffing levels and factors thought to influence staffing. We believed that important insight and guidance for hospital management could be gleaned from measuring the statistical extent to which common factors in hospitals may influence staffing levels.

Findings

As noted in the sidebar discussion of the study methodology, the analysis assessed the significance of four factors on FTE/AOB staffing:

  • Ownership and control
  • Operational characteristics
  • Financial measures/ratios
  • Clinical measures

Sidebar: Hospital Staffing Study: Methodology

The study found that, in terms of ownership and control, investor-owned hospitals tended to staff at an average of 0.348 FTEs per adjusted occupied bed (FTE/AOB) lower than the national average. This finding makes sense in view of the efficiencies obtained by facilities in investor-owned chains, where administrative overhead functions are highly consolidated and staffing control programs are well established. Similarly, not-for-profit hospitals with a multihospital affiliation were found to have a 0.341 FTE/AOB lower staff level than facilities without such affiliations (either investor-owned or not-for-profit)-again, likely due to multihospital systems' administrative economies of scale. It is important to note that investor-owned ownership and multihospital affiliation were two separate variables in the analysis to account for stand-alone investor-owned hospitals or multihospital system affiliations among not-for-profit facilities.

Interestingly, specialty hospitals tended to operate at a much higher level-i.e., 1.298 FTE/AOB. Considering the operational norms in those facilities, where physician owners/partners may call for higher levels of nurse staffing, that observation seems reasonable. The notion of economies of scale appears further supported in this analysis, as a 1 percent increase in hospital occupancy percentage appears to drive a 0.019 FTE/AOB decrease below the national average. Conversely, patient severity is associated with an increase in FTE/AOB staffing levels of about 0.228 FTE/AOB for every 1.0 increase in a hospital's case mix index.

The financial measures selected for the study from prior financial ratio analyses (i.e., cash flow margin, days in receivables, average age of plant, contractual adjustment percentage, cash flow to total debt, current ratio, and days cash on hand) did not appear to have a significant influence on staffing levels, with the exceptions of cash flow margin, days in receivables, and contractual adjustment percentage. For every 1 percent increase in cash flow margin in the analysis, staffing levels dropped by 0.106 FTE/AOB, which could indicate that lower staffing levels increased margins by reducing labor expense. The revenue cycle may have a negligible effect on staffing, as the findings indicate that an increase of one day in days in receivables translates to a 0.002 FTE/AOB increase.

This result seems counterintuitive; a hospital with higher receivables might be expected to staff at lower levels in patient accounting functions. This finding may warrant further examination in a future analysis. Also, the use of any outsourcing of revenue cycle functions was not measurable in terms of impact on accounts receivable or in the FTEs assigned to this function.

Finally, a 1 percent increase in contractual allowance (expressed as a percentage of gross revenue) appears to precipitate a small 0.02 FTE/AOB reduction from the average. These observations appear reasonable; an increase in receivables may be countered by adding resources to improve collection, while increased deductions could suggest a decline in net revenues available to pay labor costs.

Some interesting observations were noted with respect to nurse staffing. First, it appears that higher proportions of nursing FTEs (registered nurses [RNs], licensed vocational nurse [LVNs], and nursing aides) translate into some overall labor efficiencies, as FTE/AOB decreases by 0.026 FTE below the national average for each 1 percent increase in the percentage of hospital FTEs in the nursing disciplines. Conversely, within the nursing division, the higher the proportion of nursing FTEs who are RNs, the higher the overall FTE/AOB staffing level appears to be, as a 1 percent increase in the proportion of nurse staffing provided by RNs was found to correlate with a 0.06 FTE/AOB increase in staffing levels. This finding appears reasonable as lower-level nursing disciplines could cross-cover other patient care functions and yield some economies of scale, while higher proportions of RN staffing could be viewed as inefficient from a productivity standpoint. This finding also appears consistent with the findings in the nursing management literature, but suggests that hospitals with a higher proportion of RN staffing do not reap efficiencies from the higher clinical skill levels of their staffing complement. Overall, this finding may make a defensible case for a lower level of nursing skill mix, although the impact on quality deserves further study in the context of the current value-based purchasing environment.

The statistically significant variables noted in our analysis are summarized in the exhibit below. Any variable that was not significant in the analysis (based on the strength of association with FTE/AOB staffing) is not included.

Hospital Staffing Study: Significant Regression Results
Hospital Staffing Study: Significant Regression Results

The significance of the California nurse staffing mandate on overall FTE levels in hospitals appeared substantial, with the mean FTE/AOB level for hospitals in California observed at 5.97 versus 5.00 in other states-i.e., almost one FTE higher in California. To further examine this issue, the analysis was run with the same variables for hospitals outside of California. The proportion of total hospital hours in nursing was quite similar (37.9 percent in California versus 37.4 percent in other states). Further, the proportions of RN and LVN hours were also quite similar-77.3 percent in California versus 74.7 percent elsewhere. Thus, although a hospital in California may have slightly higher overall staffing levels, the difference does not appear to be a result of the staffing mandate unique to this state.

Operation as a specialty hospital, multi-hospital affiliation, occupancy percentage, and Medicare case mix index all showed the same direction and amount of influence for California facilities as these variables did in other states.

Implications

Our national analysis of 1,721 U.S. hospitals produced compelling evidence about the four explanatory factors associated with staffing levels-i.e., ownership type, operational characteristics, financial measures, and nursing measures. In terms of ownership type, the regression results indicate that investor-owned hospitals have significantly lower staffing levels than do not-for-profit hospitals,whereas government-owned hospitals and not-for-profit hospitals do not differ significantly in staffing levels.

Hospital occupancy rates are inversely associated with staffing levels, which might correlate with staffing cost management difficulties in lower-occupancy hospitals and some degree of economies of scale with higher-occupancy levels. This possibility, coupled with the study's findings on nursing labor, underscores the importance of staffing levels and quality of care as the country moves forward with payment reform and value-based purchasing.

However, these findings were not analyzed in conjunction with clinical outcomes and should be considered with some caution. At least one study reported in 2002 that lower staffing levels are associated with a higher frequency of adverse patient outcomes, such as urinary tract infection, upper gastrointestinal bleeding, and hospital-acquired pneumonia. Moreover, a 2011 study of 197,961 admissions found increased mortality when staffing levels were below target levels. A 2014 study, involving 55,159 older adults in 303 hospitals in four states, found no significant association between ICU staffing and the probability of death, due to a lack of staffing variation among the hospitals studied, but did find that a higher proportion of bachelor's-prepared critical care nurses was associated with a lower probability of death.

A study published in 2012 examined the association between RN staffing and patient safety in Florida hospitals between 1996 and 2004. The researchers used two measures of RN staffing- RN FTEs and RNs per adjusted patient day-and analyzed their association with four Agency for Healthcare Research and Quality (AHRQ) patient safety indicators (PSIs). They found that low staffing levels were associated with a higher number of all PSIs. Failure to rescue had the most significant relationship to low RN staffing. These findings from recent literature provide compelling evidence of the association between RN staffing and patient outcomes.

In general, it appears that financial constraints do not have a material influence on staffing as measured by FTE/AOB. Instead, clinical services, patient acuity, and economies of scale from higher occupancy rates appear more influential. In particular, a higher proportion of nurse staffing appears to contribute to greater overall efficiency expressed as lower FTE/AOB, while controlling for other relevant factors. Considering value-based purchasing metrics associated with higher nurse staffing levels, this can be a useful finding-increased efficiency and improved quality performance are associated with higher nurse staffing levels. From a managerial perspective, this conclusion may be troubling, as hospital payments are increasingly constrained both by competition and by payment reductions. However, health policy makers should take note of the empirical evidence that shows that clinical need-not financial concerns-drives hospital staffing decisions.

As payment reform takes hold, senior finance and operations executives will need to manage a mix of financial, operational, and clinical metrics. Finance leaders no longer can focus solely on the financial aspects of the business; nor can operations leaders focus solely on operations. Instead, a transformational manager is needed to command multidisciplinary skill sets, and the finance executive must be better informed about how staffing decisions can improve efficiency and drive value.

The findings of our analysis provide mixed and limited support for the premise that financial health is the only factor associated with lower acute care staffing levels. In today's environment, the disconnect between financial health and staffing levels will challenge hospitals where operational characteristics support higher staffing levels. Hospitals in weaker financial condition may even find that operational needs call for potentially unaffordable staffing levels. In any event, it is important for healthcare stakeholders to understand that operational and clinical conditions-especially patient acuity, occupancy levels, and specialty operation-exert more influence on hospital staffing than financial factors do.

Sidebar: Using OLS Regression to Complete Healthcare Operational Analyses


Jeffrey R. Helton, PhD, FHFMA, is assistant professor of healthcare management, Metropolitan State University of Denver, and a member of HFMA's Colorado Chapter.

Joseph S. Coyne, DrPH, MPH, is professor, Department of Health Policy and Administration, and director, Center for International Health Services Research & Policy, Washington State University, Spokane, Washington, and a member of HFMA's Washington-Alaska Chapter.

Footnotes

a. Kazahaya, G., “Harnessing Technology to Redesign Labor Cost Management Reports,” hfm, April 2005.

b. Furukawa, M., Raghu, T., and Shao, B. “Electronic Medical Records, Nurse Staffing,

and Nurse-Sensitive Patient Outcomes: Evidence from California Hospitals, 1998–2007,” Health Services Research, August 2010.

c. Coyne, J., and Helton, J., How Prepared Are U.S. Hospitals for the Affordable Care Act A Financial Condition Analysis of US Hospitals in 2011. Journal of Health Care Finance, October/November 2014; Coyne, J.S., Richards, M., Short, R., et al., “Hospital Cost and Efficiency: Does Hospital Size Really Matter?” Journal of Healthcare Management, May/June 2009; Coyne, J.S., and Singh, S.G., “The Early Indicators of Financial Failure: A Study of Bankrupt and Solvent Health Systems,” Journal of Healthcare Management, September 2008.

d. Staggs, V., and He, J., “Recent Trends in Hospital Nurse Staffing in the United States,” The Journal of Nursing Administration, July/August 2013.

e. Needleman, J., Buerhaus, P., Matike, S., et al., “Nurse-Staffing Levels and the Quality of Care in Hospitals,” The New England Journal of Medicine, May 30, 2002.

f. Needleman, J., Buerhaus, P., Pankratz, V.S., et al., “Nurse Staffing and Inpatient Hospital Mortality,” The New England Journal of Medicine, March 17, 2011.

g. Kelly, D., Kutney-Lee, A., McHugh, M., et al., “Impact of Critical Care Nursing on 30-Day Mortality of Mechanically Ventilated Older Adults,” Critical Care Medicine, May 2014.

h. Unruh, L., and Zhang, N., “Nurse Staffing and Patient Safety in Hospitals,” Nursing Research, January-February 2012.

Publication Date: Saturday, August 01, 2015

YOU MIGHT ALSO BE INTERESTED IN: