Staggs, Vincent S. PhD; He, Jianghua PhD
Nursing care quality has received a great deal of attention from researchers, patient safety organizations, and policy makers over the past 2 decades, with efforts to study and improve nurse staffing reaching a crescendo in the 1990s. The Institute of Medicine Committee on the Adequacy of Nurse Staffing in Hospitals and Nursing Homes published its report in 1996,1 the American Nurses Association launched the National Database of Nursing Quality Indicators® (NDNQI®) in 1998,2 and the California Assembly passed legislation in 1999 leading to mandated minimum nurse-to-patient ratios.3 In the new millennium there has been growing public interest in the quality and availability of healthcare, and efforts to study and improve nurse staffing have continued.
Staffing levels have been linked to patient outcomes in numerous observational studies.4-6 Although study designs that allow for causal inference have been lacking,7 and results in some cases have been mixed8 or counterintuitive,9,10 a substantial body of evidence suggests that higher nurse staffing levels are generally associated with lower rates of adverse outcomes for patients.11
There is also evidence that nursing staff composition affects patient outcomes.12 Staff composition variables linked to better patient outcomes include proportion of bachelor of science in nursing degree (BSN) nurses,4 proportion of BSN nurses with specialty certification,13 and proportion of nursing care provided by RNs.8,14,15 Use of temporary staff has been associated with worse patient outcomes in a few studies,16-18 and some have expressed concern about its effect on quality.19 However, other findings suggest null or even beneficial effects of temporary staff use,20-22 and Aiken and colleagues20(p341) have argued that “the assumption that the use of temporarily assigned registered nurses (RNs) has an adverse impact on quality of care…may be more of a myth than reality.”
There are also associations between nurse staffing and variables related to nurse retention. Researchers have linked nurse staffing levels to RN job satisfaction and burnout23,24 and have linked the proportion of nursing hours provided by RNs to rates of RN turnover.25 Others have observed a positive association between proportion of temporary nurses (as reported by nurses) and permanent nurses’ reported intent to leave.20 These associations are important because keeping nurses in the nursing profession is important. By one 2009 estimate, US demand for RNs will exceed supply by roughly 260000 full-time equivalents (FTEs) by 2025.26 This does not take into account the effects of the 2010 Affordable Care Act, which is expected to increase the number of Americans covered by health insurance by 32 million.27
National trends in nurse staffing reflect changes in the quantity and quality of nursing care available to patients, indicate whether the value placed on nursing by hospitals is changing, and tell us where nursing resources are being allocated. However, trends in nurse staffing, like trends in other indicators of nursing care quality, have not been well studied. There is evidence that rates of inpatient falls28 and hospital-acquired pressure ulcers29 have been decreasing, but a full description of how the nation’s quality of nursing care is changing requires an analysis of trends in nurse staffing levels and nursing staff composition.
About the Study
The primary purpose of this study was to describe recent trends in total nurse staffing levels, nursing skill mix (the proportions of nursing care provided by RNs, licensed practical/vocational nurses [LPNs], and assistive personnel), and use of temporary RNs in US hospitals. Although several studies of nurse staffing in California have been published since mandated nurse/patient ratios took effect in 2004,3,30,31 recent nationwide trends in staffing have not been examined in detail.
A secondary purpose was to examine the effects on nurse staffing of 2 potentially influential events: the recession that began at the end of 2007 and the 2008 Centers for Medicare & Medicaid Services (CMS) rule change ending payment for certain reasonably preventable hospital-acquired complicating conditions,32 a number of which are recognized as being sensitive to nursing care.33 The number of RN FTEs in the workforce increased dramatically in 2007 to 2008 because of the worsening overall employment picture,26 but it is not known what effects the recession or the CMS rule change had on nurse staffing levels, nursing skill mix, or use of temporary nurses in hospitals.
We focused our analyses on data from the NDNQI,2 which collects unit-level data on nursing-related measures from more than 1900 acute care hospitals. Research related to the NDNQI is carried out with the approval of the University of Kansas Human Subjects Committee. For comparison, we also analyzed hospital-level data from the American Hospital Association (AHA) annual survey database.34 Both databases are national in scope, but the AHA annual survey database includes data from more than 80% of US hospitals, whereas about 30% of US hospitals participate in the NDNQI. Based on a comparison of 2011 NDNQI data with fiscal year (FY) 2010 AHA data, the NDNQI attracts a disproportionately high number of private nonprofit hospitals, hospitals in metropolitan areas, hospitals with 100 or more staffed beds, and Magnet®-designated hospitals.
We extracted monthly NDNQI staffing data for January 2004 through December 2011 for all adult general care units (ie, medical, surgical, and medical-surgical units) and intensive care units (ICUs) in US general hospitals. Each unit’s cohort was defined as the earliest year it submitted staffing data to the NDNQI. Units in the 2011 cohort were excluded for having insufficient data to inform the trend analysis. The final sample comprised 2634 medical units, 1895 surgical units, 3561 medical-surgical units, and 2822 ICUs in 1499 hospitals. After aggregating monthly data to the quarter level, there were 183477 unit-quarters of data in the data set (16.8 quarters per unit, on average).
Participating units in NDNQI hospitals report the productive nursing care hours worked each month by unit-based nursing personnel devoting over half their shift to direct patient care responsibilities. The NDNQI sums the reported hours worked by RNs, LPNs, and assistive personnel on each unit to compute the month’s total nursing care hours. Units also report the monthly hours worked by direct care RNs who are not hospital employees but are temporarily assigned to the unit, as well as the monthly number of patient days.
We computed staffing measures separately for general care units and ICUs. We aggregated monthly data to the quarter level by summing patient days and nursing hours across months for each unit, excluding quarters for which a unit had fewer than 2 months of staffing data. For descriptive purposes, we collapsed across units, computing total nursing hours per patient day (TNHPPD) for each quarter by summing total nursing care hours across units and dividing by the sum of patient days across units. We similarly computed quarterly RN hours per patient day (RNHPPD).
To describe trends in nursing skill mix, we computed the quarterly proportion of total nursing hours worked by RNs, by LPNs, and by assistive personnel by summing the appropriate nursing care hours across units for the quarter and dividing by the sum of total nursing care hours. The quarterly proportion of RN hours provided by temporary nurses was computed in a similar manner. We also computed the proportion of hospitals using LPNs (on general care units or ICUs) in each quarter.
We examined trends in TNHPPD and RNHPPD, nursing skill mix, percentage of RN hours provided by temporary nurses, and proportion of hospitals with LPN hours by plotting each variable’s average value across time. For all measures except proportion of hospitals using LPNs, we analyzed data for general care units and ICUs separately. We collapsed across the 7 cohorts after verifying that their broad trends on each measure were generally consistent.
After visually examining these time trends, we tested whether the linear trends were statistically significant. Quarterly TNHPPD and RNHPPD were computed for each unit by dividing the sum of its nursing care hours by the sum of its patient days for the quarter, and the quarterly proportion of hours provided by RNs was computed by dividing RN hours by total nursing hours. We fit linear mixed models using the MIXED procedure in SAS 9.2 to model TNHPPD, RNHPPD, and proportion of hours provided by RNs as functions of the study quarter (numbered from 1 to 32). Given the hierarchical, clustered nature of the data (multiple measures across time for each unit), we included a random unit intercept in each model to account for the nonindependence of each unit’s repeated measures. To model the time trend in the proportion of hospitals using LPNs, we used SAS’ GLIMMIX procedure to fit a logistic mixed model that included a random hospital intercept.
We extracted data for general medical and surgical hospitals from the AHA annual survey database for FY2006-2010. Children’s hospitals, acute long-term care facilities, and other specialty hospitals were excluded. Hospitals respond to the annual survey based on a completed 12-month period. We limited the sample to years in which a hospital was open for at least 360 days, had at least 365 inpatient days, and reported the number of full- and part-time RNs employed at the end of the reporting period. There were 4683 hospitals in the final data set (4.2 years of data per hospital, on average).
We estimated the number of full-time RNs, LPNs, and assistive personnel employed for each hospital-year in the sample using AHA’s method of counting part-time personnel as half-time. We measured staffing in each survey year using total nursing FTEs per 1000 patient days and RN FTEs per 1000 patient days, computed by summing the FTEs across hospitals for the year, dividing by the sum of hospitals’ inpatient days for the year, and multiplying by 1000. We also computed the annual proportions of total nursing FTEs accounted for by RNs, LPNs, and assistive personnel.
By raising RN staffing levels, NDNQI hospitals have been steadily raising total nurse staffing levels on general care units and ICUs (Figure 1). From 2004 to 2011, TNHPPD and RNHPPD on general care units increased by 0.9 (11.5%) and 1.1 (22.9%), respectively. Changes were less pronounced on ICUs, where staffing levels were much higher to begin with; TNHPPD and RNHPPD increased by 0.5 (3.0%) and 0.7 (4.8%), respectively. All 4 trends were statistically significant (P values < .001).
We observed a statistically significant upward trend in the proportion of hours provided by RNs on general care units (P value < .001). Whereas this proportion increased by 9.7% during 2004-2011 for general care units (Figure 2), it only increased from 88.9% to 89.5% for ICUs. We did not test this trend for significance. Meanwhile, the proportion of nursing care hours provided by LPNs has been decreasing, dropping by more than half on both general care units and ICUs (Figure 2). We found a significant downward trend across time in the proportion of hospitals using LPNs to provide nursing care (P value < .001), the percentage falling from 85.6% in 2004 to 67.6% in 2011.
Use of temporary RNs dropped precipitously from mid-2008 through mid-2009 on both general care units and ICUs (Figure 3). The annual share of RN hours provided by temporary nurses on general care units fell from a high of 3.8% in 2004 to a low of 1.1% in 2010. On ICUs, temporary nurses accounted for 4.1% of RN hours in 2004, 2006, and 2007, but only 1.6% in 2010. We did not test these trends for statistical significance because of their marked nonlinearity.
Staffing trends in the AHA sample were similar. During FY2006-2010, total and RN nursing FTEs per 1000 patient days increased by 1.1 (13.6%) and 1.1 (18.6%), respectively (Figure 4). The share of total nursing FTEs accounted for by RNs went from 72.8% to 75.4% (a 3.6% increase), and the share of LPN FTEs declined from 7.4% to 5.4% (a 27% decrease) (Figure 5).
This is the 1st published study examining nationwide trends in hospital nurse using hospital- and unit-level data from 2 national databases. Hospitals have increased total nurse staffing levels by raising RN staffing levels, use of LPNs is in decline, and RNs are accounting for a steadily increasing share of nursing care hours and nursing FTEs. Use of temporary RNs in NDNQI hospitals dropped sharply beginning in 2008 and has remained low.
Nurse Staffing Levels
The increase in staffing levels we observed on general care units was caused in part by a 5.7% decline in patient days across the years of the study. However, even as patient volume decreased, average hours of total and RN nursing care increased by 5.0% and 15.2%, respectively. On ICUs, average patient days increased by 4.5% during the study, but average total and RN hours increased even more (8.0% and 8.8%, respectively).
Despite recession-related economic difficulties,35 hospitals have been increasing RN staffing levels. We think that the attention given to nurse staffing over the past 2 decades has led to a greater appreciation of its importance among hospital administrators and the wider healthcare community. Another potential explanation is that hospitals increased staffing levels because of the 2008 CMS rule change in the hope of preventing costly never events. However, there is no evidence from this study to support this hypothesis; the increase in nurse staffing levels has been slow and steady. Perhaps, recession-related economic constraints prevented hospitals from making more than incremental changes to nurse staffing levels in response to the rule change.
Use of Temporary Staff and LPNs
Economic factors were likely responsible for the rapid decline in RN hours provided by temporary staff from mid-2008 through mid-2009 (Figure 3). Temporary RNs can be expensive,20 and hospitals apparently took advantage of the surge in RNs looking for work due to the recession26 by replacing temporary staff with permanent staff as a cost-saving measure. It could be argued that the CMS rule change was partly responsible for this dramatic change, but given the limited, inconclusive evidence that use of temporary staff has adverse effects on patient outcomes20 and the lack of evidence in this study that the CMS change had any effect on nurse staffing levels, this explanation is not supported. The decline in the use of LPNs observed in this study appears to be part of a trend toward phasing LPNs out of acute care.
Limitations and Conclusion
Generalizability is not a major concern for the AHA sample, which included 4683 hospitals (81.4%) of the 5754 hospitals eligible for registration with AHA. The NDNQI sample was smaller, and the set of NDNQI hospitals is not a random or proportional sample, so caution is in order here. The trends we observed in the NDNQI sample may differ somewhat from trends in non-NDNQI hospitals, but we think our findings are indicative of changes taking place on a broad scale, as suggested by the consistency between the findings for the NDNQI and AHA samples.
The differences between the 2 samples and the staffing measures used for each make their similar trends even more compelling. Both samples were limited to general hospitals. However, whereas the NDNQI sample was limited to adult ICUs and general care units in acute care facilities, the AHA data reflect hospital-wide nurse staffing. Moreover, the NDNQI measures were based on productive nursing hours actually worked by direct care personnel, not nursing FTEs estimated from counts of nursing personnel on the hospital payroll.
We think these trends represent good news for patients. There is some evidence that the trend toward higher RN staffing levels has been accompanied by declining rates of inpatient falls28 and hospital-acquired pressure ulcers,29 but more rigorous analysis is needed to determine if these co-occurring phenomena have a causal relation.
1. Wunderlich GS, Sloan FA, Davis CK; Institute of Medicine, eds. Nursing Staff in Hospitals and Nursing Homes. Washington, DC: National Academy Press; 1996.
2. Dunton N, Gajewski B, Taunton RL, Moore J. Nurse staffing and patient falls on acute care hospital units. Nurs Outlook. 2004; 52: 53–59.
3. Serratt T, Harrington C, Spetz J, et al. Staffing changes before and after mandated nurse-to-patient ratios in California’s hospitals. Policy Polit Nurs Pract. 2011; 12: 133–140.
4. Aiken LH, Cimiotti JP, Sloane DM, et al. Effects of nurse staffing and nurse education on patient deaths in hospitals with different nurse work environments. Med Care. 2011; 49: 1047–1053.
5. Blegen MA, Goode CJ, Spetz J, et al. Nurse staffing effects on patient outcomes: safety-net and non-safety-net hospitals. Med Care. 2011; 49: 406–414.
6. Needleman J, Buerhaus P, Pankratz VS, et al. Nurse staffing and inpatient hospital mortality. N Engl J Med. 2011; 364: 1037–1045.
7. Butler M, Collins R, Drennan J, et al. Hospital nurse staffing models and patient and staff-related outcomes. Cochrane Database Syst Rev. 2011; 7: 1–70.
8. Lake ET, Cheung RB. Are patient falls and pressure ulcers sensitive to nurse staffing? West J Nurs Res. 2006; 28: 654–677.
9. Mark BA, Harless DW. Nurse staffing and post-surgical complications using the present on admission indicator. Res Nurs Health. 2009; 33: 35–47.
10. Staggs VS. Nurse staffing, skill mix, and assaults on psychiatric units. Res Nurs Health. 2013; 36: 26–37.
11. Kane RL, Shamliyan T, Mueller C, et al. Nursing Staffing and Quality of Patient Care. Rockville, MD: Agency for Healthcare Research and Quality; 2007. AHRQ Publication No. 07-E005.
12. Lankshear AJ, Sheldon TA, Maynard A. Nurse staffing and healthcare outcomes: a systematic review of the international research evidence. Adv Nurs Sci 2005; 28: 163–174.
13. Kendall-Gallagher D, Aiken LH, Sloane DM, et al. Nurse specialty certification, inpatient mortality, and failure to rescue. J Nurs Scholarsh. 2011; 43: 188–194.
14. Needleman J, Buerhaus P, Mattke S, et al. Nurse-staffing levels and the quality of care in hospitals. N Engl J Med. 2002; 346: 1715–1722.
15. Staggs VS, Knight JE, Dunton N. Understanding unassisted falls: effects of nurse staffing level and nursing staff characteristics. J Nurs Care Qual. 2012; 27: 194–199.
16. Alonso-Echanove J, Edwards JR, Richards MJ, et al. Effect of nurse staffing and antimicrobial-impregnated central venous catheters on the risk for bloodstream infections in intensive care units. Infect Control Hosp Epidemiol 2003; 24: 916–925.
17. Phibbs CS, Bartel AP, Stone PW. A longitudinal study of the effects of nurse staffing levels, nursing characteristics, and job tenure on patient outcomes for VA inpatient care [abstract]. J Investig Med. 2009; 57: 116.
18. Robert J, Fridkin SM, Blumberg HM, et al. The influence of the composition of the nursing staff on primary bloodstream infection rates in a surgical intensive care unit. Infect Control Hosp Epidemiol. 2000; 21: 12–17.
19. May JH, Bazzoli GJ, Gerland AM. Hospitals’ responses to nurse staffing shortages. Health Aff. 2006; 25: 316–323.
20. Aiken LH, Clarke SP, Xue Y, et al. Supplemental nurse staffing in hospitals and quality of care. J Nurs Adm. 2007; 37: 335–342.
21. Newhouse RP, Johantgen M, Pronovost PJ, et al. Perioperative nurses and patient outcomes—mortality, complications, and length of stay. AORN J. 2005; 81: 508–528.
22. Xue Y, Aiken LH, Freund DA, Noyes K. Quality outcomes of hospital supplemental nurse staffing. J Nurs Adm. 2012; 42: 580–585.
23. Aiken LH, Clarke SP, Sloane DM, et al. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002; 288: 1987–1993.
24. Holden RJ, Scanlon MC, Patel NR, et al. A human factors framework and study of the effect of nursing workload on patient safety and employee quality of working life. BMJ Qual Saf. 2011; 20: 15–24.
25. Staggs VS, Dunton ND. Hospital and unit characteristics associated with nursing turnover include skill mix but not staffing level: an observational cross-sectional study. Int J Nurs Stud. 2012; 49: 1138–1145.
26. Buerhaus PI, Auerbach DV, Staiger DO. The recent surge in nurse employment: causes and implications. Health Aff. 2009; 28: w657–w668.
27. Institute of Medicine. The Future of Nursing: Leading Change, Advancing Health. Washington, DC: The National Academies Press; 2011.
28. He J, Dunton N, Staggs V. Unit-level time trends in inpatient fall rates of US hospitals. Med Care. 2012; 50: 801–807.
29. VanGilder C, Amlung S, Harrison P, Meyer S. Results of the 2008-2009 International Pressure Ulcer Prevalence™ Survey and a 3-year, acute care, unit-specific analysis. Ostomy Wound Manage. 2009; 55 (11): 39–45.
30. Aiken LH, Sloane DM, Cimiotti JP, et al. Implications of the California nurse staffing mandate for other states. Health Serv Res. 2010; 45: 904–921.
31. Donaldson N, Shapiro S. Impact of California mandated acute care hospital nurse staffing ratios: a literature synthesis. Policy Polit Nurs Pract. 2010; 11: 184–201.
33. National Quality Forum. National Voluntary Consensus Standards for Nursing-Sensitive Care: An Initial Performance Measure Set. Washington, DC: National Quality Forum; 2004.
34. American Hospital Association. AHA Annual Survey Database. Chicago, IL: Health Forum LLC; 2011.