Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Apr 1.
Published in final edited form as: Res Nurs Health. 2014 Jan 9;37(2):90–97. doi: 10.1002/nur.21582

Night and Day in the VA: Associations between Night Shift Staffing, Nurse Workforce Characteristics, and Length of Stay

Pamela B de Cordova 1, Ciaran S Phibbs 2, Susan Schmitt 3, Patricia W Stone 4
PMCID: PMC3959218  NIHMSID: NIHMS552094  PMID: 24403000

Abstract

In hospitals, nurses provide patient care around the clock, but the impact of night staff characteristics on patient outcomes is not well understood. The aim of this study was to examine the association between night nurse staffing and workforce characteristics and the length of stay (LOS) in 138 Veterans Affairs (VA) hospitals using panel data from 2002 through 2006. Staffing in hours per patient day was higher during the day than at night. The day nurse workforce had more educational preparation than the night workforce. Nurses’ years of experience at the unit, facility, and VA level were greater at night. In multivariate analyses controlling for confounding variables, higher night staffing and a higher skill mix were associated with reduced LOS.

Keywords: night shift, human capital, nursing workforce, length of stay, staffing

Although the intensity of nursing work may differ at certain hours during the day, nurses are essential in ensuring patient safety at all hours. Patient outcomes are worse when critical events occur at times other than on weekdays from 9 a.m. to 5 p.m. (Barnett, Kaboli, Sirio, & Rosenthal, 2002; Becker, 2007; Bhonagiri, Pilcher, & Bailey, 2011; Peberdy et al., 2008). In a comprehensive systematic review of patient and employee outcomes on off-shifts (i.e., nights and weekends), we also found that both patient and employee outcomes were worse on off-shifts than on more regular hours (de Cordova, Phibbs, Bartel, & Stone, 2012). Although evidence suggests that patients admitted to hospitals on off-shifts have worse outcomes than those admitted during the day, there is a paucity of evidence on the staffing and workforce characteristics on off-shifts that may explain these worse outcomes.

Better nurse staffing is known to be associated with better patient outcomes in hospital settings (Aiken, Clarke Sloane, Sochalski, & Silber, 2002; Glance et al., 2012; Kane, Shamliyan, Mueller, Duval, & Wilt, 2007; Liang, Tsay, & Chen, 2012; Sidani, Covell, & Antonakos, 2011; Stone, Pogorzelska, Kunches, & Hirschhorn, 2008; Wiltse Nicely, Sloane, & Aiken, 2012) and in skilled nursing and long-term care facilities (Horn, 2008; Kim, Harrington, & Greene, 2009; Konetzka, Stearns, & Park, 2008). Adequacy of nurse staffing is a key variable associated with patient safety (Clarke, 2007). National nursing organizations recognize the importance of nurse staffing to patient safety and have implemented recommendations for nurses and administrators (Weston, Brewer, & Peterson, 2012).

Characteristics of the nursing workforce, including nurses’ training, education, and experience, also affect patient safety and outcomes. Over the past 10 years, evidence has grown that a better-educated workforce is associated with better patient outcomes (Aiken, Clarke, Cheung, Sloane, & Silber, 2003; Blegen, Goode, Park, Vaughn, & Spetz, 2013). For example, researchers recently found that hospitals whose registered nurse (RN) workforces had higher educational levels had lower occurrences of pressure ulcers, heart failure mortality, post-operative deep vein thrombosis, and shorter length of stay (LOS) (Blegen et al., 2013). In better nurse work environments with improved staffing and a more educated workforce, patients have almost a 10% decreased odds of mortality and death from complications (Aiken et al., 2011).

Both nurse staffing numbers and workforce characteristics may be associated with the outcome of LOS. LOS is an indicator of both adverse patient outcomes and health care efficiency. LOS has been endorsed by the National Quality Forum as a quality measure, is an important fiscal measure for hospitals, and is negatively associated with nurse staffing (Kane et al., 2007; NQF, 2008). A range of adverse patient outcomes (e.g., falls, potentially preventable infections, medication errors) are associated with longer LOS (Zhan & Miller, 2003). Quality of nursing care is related to LOS because high-quality care can prevent complications, including infections, pressure ulcers, and falls, all of which increase LOS. Arguably, a more educated and experienced workforce would be more skilled in preventing complications and decreasing LOS.

Patients experience worse outcomes at night than during the day (Caughey et al., 2008; Laupland, Shahpori, Kirkpatrick, & Stelfox, 2008; Peberdy et al., 2008). However, no evidence was found on whether these outcomes are related to nurse staffing or workforce characteristics on night shifts. Previous research has been limited to cross-sectional analyses, due to lack of availability and/or cost of longitudinal data.

This study is the first to use longitudinal data to examine the impact on LOS of both nurse staffing and nurse workforce characteristics. The purpose of this study was to examine the association between night nursing inputs (i.e., staffing and workforce characteristics) and LOS. The study was a sub-analysis of data compiled for a larger econometric analysis of the association between the skill level and the stability of the nursing workforce on nursing-sensitive patient outcomes. The research question was: What is the relationship between LOS and RN staffing levels, skill mix, and experience on the night shift? The hypotheses that were tested were: 1) RN staffing at night is negatively related to LOS; 2) RN skill mix at night is negatively related to LOS; and 3) RN experience at night is negatively related to LOS.

Theoretical Framework

The study was based on the theory of human capital, in which training and competencies improve productivity in one’s profession (Buhr, 2010). As initially proposed by economists, human capital can be general or specific (Becker, 1964). General human capital is often acquired through formal education and years of experience on the job. A hallmark of general human capital is that it increases productivity in all environments, and an individual can apply this expertise in different work settings (e.g., hospitals). Specific human capital is acquired from years of experience in a particular setting (e.g., hospital or unit). This type of human capital is often considered on-the-job training and improves productivity only in the environment in which the individual works.

Given the Institute of Medicine’s recommendation (2010) to increase the educational preparation of the nursing workforce, it is important to understand the association between nurses’ general human capital, such as education, and quality of care. Nurses’ specific human capital, in the form of experience in a particular nursing care setting, also may be related to quality outcomes. Newer graduate nurses often possess less specific human capital, both within the facility and on the unit. Contract nursing personnel may also have less specific human capital because they are unfamiliar with the hospital or the unit to which they are assigned. Therefore, levels of human capital in a nursing workforce, in addition to staffing, may be associated with patient outcomes.

In this study, general human capital was conceptualized as education preparation and prior nurse experience. The operational definition of educational preparation was the percentage of RN hours worked by nurses who had either associate or baccalaureate degrees. Prior nurse experience was operationalized as the number of years of nursing experience prior to joining the VA. Specific human capital was conceptualized as nurses’ experience on the unit, within the facility, and within the VA. The operational definitions for these variables were the number of years the RN worked on the unit, within the specific facility, and within the VA system.

Methods

Design and Data Source

A longitudinal (panel) dataset was developed from 138 VA acute care hospitals for the period October 2002 through 2006. Two university institutional review boards and a local VA Research & Development Committee approved the study. Patient and nurse data were drawn from the hospitals’ administrative and electronic data sources and compiled into a single dataset by the research team. For this study, only medical, medical/surgical, surgical, step-down, and telemetry units were included in the analyses. Psychiatric units, ambulatory services, and intensive care units were excluded.

For each unit, monthly measures of each variable were created. Based on the timing of the VA pay rate’s shift differential, day hours were defined as beginning at 7 a.m. and ending at 6 p.m., and night hours were 6:01 p.m. through 6:59 a.m. The final sample of unit-monthly observations (N = 8,243) represented 185 medical, medical/surgical, surgical, step-down, and telemetry units.

Variables

Length of stay (LOS)

The outcome variable was defined as the average LOS for patients admitted to the unit in each month. When calculating the average unit LOS, each patient was assigned to where (unit) and when (month) the patient was admitted to the hospital. If a patient’s LOS spanned more than one month, it was attributed to the month when the patient was admitted. The LOS was a minimum of one day, and if the LOS was longer than one year, the patient was considered an outlier and removed from the data. If a patient was transferred from acute care to a sub-acute care unit, the sub-acute days were not included in the total LOS. However, acute care LOS did include intensive care days, if a patient was transferred from a medical/surgical unit to an intensive care unit.

Staffing variables

The data for monthly staffing levels were obtained from the VA budget system, which tracks nursing personnel hours appropriated to each nursing unit by type of personnel. Productive work hours excluded vacation and sick time but included education time (e.g., when a nurse attended an in-service program). Worked hours also included instances when nurses worked on units other than their primary units (i.e., floating). An adjustment for time spent floating was made based on an average percentage of work effort allocated to other units.

Staffing variables were calculated separately for each type of personnel, which included RNs, Licensed Practical Nurses (LPNs), and Unlicensed Assistive Personnel (UAPs). For each personnel type, an average hour per patient day (HPPD) variable was created for each month divided by patient bed days.

The VA accounting data only tracked the nursing hours by month and did not distinguish between shifts. To distinguish between different shifts, the VA payroll data was used to calculate the proportion of each type of staff that was paid a shift differential. These data were then used to create staffing variables for day and night hours.

Human capital variables

VA payroll data were also used to create human capital variables for the educational preparation of nurses and years of RN experience. Education level was operationalized separately for day and night and defined as the percentage of total RN hours worked by nurses with baccalaureate degrees and the percentage of total RN hours worked by nurses with associate degrees.

The experience variables were broken down to include prior RN experience as well as unit, facility, and VA system tenure. Prior experience was defined as the number of years of experience prior to joining the VA. Unit tenure was tracked from the unit each nurse worked on during each pay period and defined as the mean number of years nurses worked on the unit. The payroll data were unavailable prior to 1995; therefore, facility tenure, which did not depend on the older pay period data, was used as a proxy for unit tenure for those nurses who were working on the unit in the first month of 1995. Because most nurses who changed units switched during the first year of employment, using the proxy for unit tenure was thought to be acceptable. Finally, all of the education and experience variables were weighted by the actual hours worked by each of these nurses during the month.

Covariates

Three groups of covariates were included in the regression models: patient, unit, and nurse characteristics. Patient characteristics included age, diagnosis-related groups (DRG) weights to adjust for patient characteristics (mean DRG for the unit in the month and four dummy variables where the DRG weights were categorized into quantiles), and the Elixhauser index of co-morbidity to control for differences in patient conditions (mean average index for patients on the unit in the month). The unit characteristic that was controlled for was the mean number of admissions to unit in the month. Nurse characteristics included the mean percentage of RN hours provided by contract nurses on the unit in the month. Contract nurses were considered a covariate because these nurses may possess high general human capital (overall years of experience) but less specific human capital due to lack of experience on the unit, within the facility, or within the VA system. All of the covariates were averaged at the unit level for the month of the observation.

Model Specifications

To break the strong correlation between nurse staffing and the human capital variables on the night and day shifts, the night shift variables were modeled as the difference between the night and day variables. This difference (differential) was calculated by using the monthly variables for nights minus the daytime average for the year. Use of the annual average for day shift prevented a change in a month’s day variable from affecting that month’s night differential, so the variation in the night variables would be driven by changes in the night staffing and not by changes in the day staffing. Therefore, the model’s specifications include differences between night and day shift, while controlling for the day variables.

Analytic Methods

All analyses were conducted at the unit level. Descriptive statistics were summarized for day and night shifts. T-tests were computed to identify differences between day and night for each of the variables. The outcome variable LOS was tested against the normal distribution assumption; LOS was log-transformed and is reflected as such in the tables presented here.

A primary model was developed to test the three hypotheses simultaneously. Multiple variations of the model were created to determine robustness of the results. Models also were examined with and without educational preparation. When education variables were tested, models only included the associate degree RNs because there was a strong correlation (r=0.6) between the percentage of hours worked by baccalaureate-prepared and associate-degree RNs. Time dummies (47 dummies for 48 months) were included in the models to account for seasonality and the overall time trend. Robust standard errors were used to account for clustering of units within hospitals.

Fixed effects analysis was used to control for unobservable variables (e.g., work environment) correlated with the outcome. The fixed effects analysis allowed for examination of variation within units over time while controlling for observed and unobserved time-invariant attributes of a unit (e.g., nursing leadership) that had not changed over the four years. Fixed effects models allow for the control of unobserved factors by removing the mean effects of each variable for each unit. The estimates are based on the within-unit variance and do not allow any inferences to be made about the overall level of each variable. For example, fixed effects models did not provide an estimate of the effect of higher average staffing levels. Rather, they measured the marginal effect of a unit having higher staffing, relative to its mean staffing level.

In addition to the multivariable models, correlational matrices of staffing and human capital variables were examined. Some correlations between predictor variables were moderate, but regression diagnostics were used to determine that most collinearity was below commonly accepted standards where the variance inflation factors (VIFs) were <10. The VIFs ranged from 1.07 for HPPD to 2.77 for VA tenure. The correlation coefficients were modest between the education and experience variables (r=0.5). The highest correlation was among experience variables, specifically between VA and facility tenure (r=0.7) and unit and facility tenure (r=0.7). Therefore, facility tenure was removed from the models. All analyses were conducted at the unit level, and all data were analyzed using STATA 11.1 (Stata Corporation, College Station, Texas).

Results

Staffing and Workforce Characteristics

In four years of data, there were 8,243 unit-monthly observations of 185 medical-surgical units in 138 VA hospitals across the US. Overall, there were significant differences between day and night hours for staffing, education, and experience of the nursing workforce (Table 1). RN staffing was greater during the day than at night (4.3 HPPD vs. 3.4). However, the percentage of care hours provided by RNs at night was higher than the percentage of hours provided by RNs during the day (64.6% vs. 60.2%). Similarly, the percentage of HPPD provided by UAPs was higher during the day than at night (15.5% vs. 11.0%).

Table 1. Differences in Day and Night Staffing and Human Capital in VA Hospitals (N = 8,243 unit-monthly observations).

Day Night p value
Mean SD Mean SD
Staffing HPPD 4.3 2.0 3.4 1.8 <0.01
% HPPD provided by RNs 60.2 0.2 64.6 0.2 <0.01
% HPPD provided by LPNs 24.2 0.2 24.3 0.2 <0.01
% HPPD provided by UAPs 15.5 0.2 11.0 0.1 <0.01
Education % hours provided by Associate degree RNs 43.2 0.2 45.3 0.2 <0.01
% hours provided by Baccalaureate degree RNs 37.0 0.2 35.4 0.2 <0.01
RN Experience* Prior experience 3.5 1.9 3.4 2.2 <0.01
Unit tenure 4.1 2.1 5.0 2.9 <0.01
Facility tenure 6.5 3.2 7.2 3.8 <0.01
VA tenure 7.5 3.8 9.0 4.6 <0.01
Open in a new tab

Notes. VA = Veterans Affairs, HPPD = Hours Per Patient Day, RN = Registered Nurse, LPN = Licensed Practical Nurse, UAP = Unlicensed Assistive Personnel.

*

Experience variables were weighted by the actual hours worked by each of these RNs during the month.

Human capital levels differed during the day and at night. Specifically, the educational preparation of nurses was higher during the day. During the day, 37.0% of the hours worked were provided by baccalaureate-degree nurses as compared to 35.4% of the hours at night (p<0.01). Conversely, the percentage of night hours worked by associate-degree nurses was slightly higher than during the day (45.3% vs. 43.2%, p<0.01). Prior experience was the only experience variable that was slightly higher during the day (3.5 years vs. 3.4 years, p <0.01). The remaining experience variables were higher at night than during the day, with the greatest difference found in VA tenure. For example, at night, the units had a mean of 9.0 years of VA experience, compared to 7.5 years of experience during the day (p <0.01).

Relationship to Length of Stay

In the multivariable fixed effect models, day staffing (p<.01) was negatively associated and the difference between night and day staffing (p=0.05) was marginally negatively associated with LOS (Table 2). The differential effect indicated that when the difference between night and day staffing became smaller (shift staffing became more similar), the LOS decreased. Each hour that staffing was lower on nights relative to day shift was associated with an increase in LOS of 1.5%.

Table 2. Multivariate Analysis of Relationship of Nurse Staffing and Human Capital (Education and Experience) to LOS in VA Hospitals (N = 8,243 unit-monthly observations).

β Coefficient 95% CI p value
Staffing Day shift HPPD −0.03 y[−0.04, −0.02] <0.01
Difference in HPPD* −0.02 [−0.03, −0.01] 0.05
% HPPD provided by LPNs 0.05 [−0.11, 0.01] 0.13
Difference in % HPPD provided by LPNs* 0.02 [−0.08, 0.12] 0.42
Day shift % HPPD provided by UAPs 0.20 [0.05, 0.36] 0.01
Difference in % HPPD provided by UAPs* 0.18 [0.03, 0.33] 0.02
Education Day shift % hours provided by Associate degree RNs −0.01 [−0.06, 0.43] 0.77
Difference in % hours provided by Associate degree RNs* 0.05 [0.01, 0.10] 0.03
Experience Day shift prior experience −0.01 [−0.01, 0.01] 0.13
Difference in prior experience* −0.01 [−0.01, 0.01] 0.98
Day shift unit tenure −0.01 [−0.02, −0.01] 0.10
Difference in unit tenure* −0.01 [−0.02, −0.01] 0.02
Day VA tenure 0.01 [−0.01, 0.02] 0.69
Difference in VA tenure* −0.01 [−0.01, 0.01] 0.29
Open in a new tab

Notes. LOS = Length of Stay, VA = Veterans Affairs, HPPD = Hours Per Patient Day, LPN = Licensed Practical Nurse, UAP = Unlicensed Assistive Personnel. Model controlled for % contract nurses, admissions, Elixhauser, DRGs (weighted in quantiles), age, and time dummies.

*

Difference is defined as the average of night variable for the month minus the average of day variable for the year.

The percentage of HPPD provided by UAPs during the day, that is, presence of a larger proportion of UAPs in relation to RNs, was associated with longer LOS. The differential effect for UAPs at night was also significant; as the percentage of hours worked by UAPs at night increased in comparison to day shift (and the percentage of hours worked by licensed nurses went down), the LOS increased. The percentage of nursing hours provided by LPNs was not significantly related to LOS.

Differences in education of the night and day workforce had an independent effect on LOS. A greater percentage of hours worked by associate-degree nurses at night in comparison to during the day was associated with an increase in LOS. Longer unit tenure among night shift staff than on the day shift was negatively associated with LOS, although this effect was relatively small.

When the educational variables were removed from the analysis (Table 3), day staffing was still negatively associated (p <0.01) and the differential between day and night staffing was still marginally negatively associated (p =0.05) with LOS. Again, the findings suggested that, for each hour that staffing was lower on nights relative to day shift, the LOS may increase by 1.5%. Additionally, the percentage of care hours provided by day shift UAPs and the differential between day and night shift UAP staffing were positively associated with LOS. The differential effect of unit tenure also was significantly related to LOS, as in the first model.

Table 3. Multivariate Analysis of Relationship of Nurse Staffing and Human Capital (Experience Only) to LOS in VA Hospitals (N = 8,243 unit-monthly observations).

β Coefficient 95% CI p value
Staffing Day shift HPPD −0.03 [ −0.04,−0.02] <0.01
Difference in HPPD* −0.02 [ −0.03,−0.01] 0.05
Day shift % HPPD provided by LPNs −0.01 [ −0.06, 0.04] 0.62
Difference in % HPPD provided by LPNs* −0.01 [ −0.10, 0.10] 0.92
Day shift % HPPD provided by UAPs 0.19 [ 0.04, 0.36] 0.02
Difference in % HPPD provided by UAPs* 0.18 [ 0.02, 0.33] 0.03
Experience Day shift prior experience −0.01 [ −0.01, 0.01] 0.14
Difference in prior experience* −0.01 [ −0.01, 0.01] 0.77
Day shift unit tenure −0.03 [−0.02 −0.01] 0.07
Difference in unit tenure* −0.01 [ −0.02, −0.01] 0.03
Day VA tenure 0.01 [ −0.01, 0.01] 0.67
Difference in VA tenure* −0.01 [ −0.01, 0.01] 0.32
Open in a new tab

Notes. LOS = Length of Stay, VA = Veterans Affairs, HPPD = Hours Per Patient Day, LPN = Licensed Practical Nurse, UAP = Unlicensed Assistive Personnel. Model controlled for % contract nurses, admissions, Elixhauser, DRGs (weighted in quantiles), age, and time dummies.

*

Difference is defined as the average of night variable for the month minus the average of day variable for the year.

Discussion

On these VA acute care units at night, there were fewer nurses and a less-well-educated workforce, but experience on the unit, in the facility, and in the VA was greater at night, suggesting that a core group of experienced nurses may choose to work at night. Differences between nurse staffing at night and during the day were associated with unit LOS.

Although much more evidence supports an association between poor staffing in general and poor patient outcomes in (Kane et al., 2007), the findings from this study are consistent with the limited research available on night shift staffing. For example, in a cohort study of 569 adult ICU patients, patients with fewer nurses at night had an increased risk of re-intubation (relative risk 5.7; 95% CI 2.4-13.7; p <0.01) and pulmonary failure (relative risk 3.6; 95% CI 1.3-10.1; p < 0.01) (Dimick, Swoboda, Pronovost, & Lipsett, 2001). The same research team found a 39% increase in hospital LOS when ICU nurses cared for three or more patients at night (Amaravadi, Dimick, Pronovost, & Lipsett, 2000)

Night staffing has been found to be less than optimal in non-VA settings. For example, Needleman et al. (2011) found over a four-year period (2003-2006) in one large non-VA institution that 29.7% of 11,650 night shifts, compared to 13.7% of 11,663 day shifts, were below a target level of RN staffing, and patient mortality increased when RN staffing fell below the target level (Hazard ratio 1.02; 95% CI 1.01-1.03; p<0.01) (Needleman et al., 2011).

Skill mix was associated with LOS in this study, which is congruent with what other researchers have found. In a cross-sectional study of 2,545 nurses, when there were higher proportions of RNs in a unit’s skill mix, there was more teamwork and less reported missed care (β = 0.436, p<0.01) (Kalisch & Lee, 2011). Nurse admininstrators who increase the percentage of nursing care provided by UAPs and decrease RN skill mix as a cost-saving mechanism may ultimately increase LOS.

There are some differences between the nursing workforce in hospitals in the VA system as compared to the private sector. For example, VA nurses are less likely to resign, which is attributed to the VA’s extensive retirement benefits (VA, 2013). VA hospitals have standardized technology across institutions, so nurses who work within the VA system can transfer between hospitals and still maintain specific human capital within the VA. However, the private and federal systems are alike in other ways. Although VA hospitals are governed by federal law rather than state laws for nurse staffing, because VA hospitals aim to attract and retain nurses in a given state, they are often pressured to provide staffing levels mandated in state regulations. Similarly, VA RN salaries are competitive because they are set by law to equal the average of the RN salaries for hospital-based RNs in the local market.

The findings from this study can inform nurse managers and administrators of VA hospitals about the impact of decreasing nurse staffing at night. Given an apparently lower nursing workload at night (e.g., no elective surgery, fewer patient admissions, etc.), nurse managers may be inclined to reduce staffing levels at night in order to reduce cost. However, this decision can lengthen patient stay and ultimately increase hospital expenditures by increasing the risk for infection and other complications. Patient acuity, level of care, and volume can change at any time, and nurse staffing models should be flexible and facilitate shift-to-shift decisions in response to patients’ needs and census (Needleman et al., 2011). Nurse managers are also encouraged to consider the education level and experience of the VA workforce in off-shift planning. For example, increasing the years of experience of nurses who work on the unit at night may facilitate shorter LOS. Differences in workforce characteristics at night as compared to during the day may be associated with other patient outcomes.

Limitations

The data analyzed for this report now ranges from seven to 11 years old. However, the VA nursing workforce has remained relatively stable over the past ten years. Recent recommendations to increase educational preparation of nurses and recent staffing laws would not necessarily pertain to VA federal hospitals.

These analyses were conducted at the unit level and not the patient level because of the nature of the data available at the time. As a result, the outcome variable, LOS, and the staffing variable, HPPD, are both operationalized with days as the denominator, increasing the risk of endogeniety bias. This risk would be lessened if analyses could be conducted with patient-level data. For the model specifications, using fixed effects was appropriate and decreased the risk of omitted variable bias. However, inclusion of other variables may have strengthened the analysis. For example, nurses are not the only health care providers that affect quality of care. Data on physicians and other ancillary personnel should be included; unfortunately, these were not available at the time of the analyses. Less supervision or higher levels of fatigue at night were not measured and also may be related to poor patient outcomes. Finally, these results are generalizable only to VA hospitals at this time.

Conclusion

Relative to the day shift, better nurse staffing and better skill mix at night was associated with shorter LOS. Nurses working at night differed in education and experience from those working during the day. Night nurses’ educational preparation and years of experience on the unit was negatively associated with LOS; however, the effects were small. Improving staffing at night in VA acute care hospitals may improve LOS.

Acknowledgements

This project was supported by grant number 053420 from the Robert Wood Johnson Foundation, R36HS018216 from the Agency for Healthcare Research and Quality, IIR 09-362 from the Veterans’ Affairs, and by the Sigma Theta Tau Alpha Zeta Chapter. The corresponding author acknowledges postdoctoral fellowship support from the National Institute of Nursing Research training grant, Advanced Training in Nursing Outcomes Research (T32-NR-007104, Linda Aiken, PI).

Contributor Information

Pamela B. de Cordova, College of Nursing, Rutgers, The State University of New Jersey, 110 Paterson Street, New Brunswick, NJ 08901, Phone: 848-932-0424, Fax: 732-932-7745, pam.decordova@rutgers.edu.

Ciaran S. Phibbs, VA Palo Alto Healthcare System, Menlo Park, CA; Department of Pediatrics, Stanford University School of Medicine, Stanford, CA.

Susan Schmitt, VA Palo Alto Healthcare System, Menlo Park, CA.

Patricia W. Stone, Columbia University School of Nursing, New York, NY.

References

  1. Aiken LH, Cimiotti JP, Sloane DM, Smith HL, Flynn L, Neff DF. Effects of nurse staffing and nurse education on patient deaths in hospitals with different nurse work environments. Medical Care. 2011;49:1047–1053. doi: 10.1097/MLR.0b013e3182330b6e. doi:10.1097/MLR.0b013e3182330b6e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aiken LH, Clarke SP, Cheung RB, Sloane DM, Silber JH. Educational levels of hospital nurses and surgical patient mortality. JAMA. 2003;290:1617–1623. doi: 10.1001/jama.290.12.1617. doi: 10.1001/jama.290.12.1617 [doi]290/12/1617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002;288:1987–1993. doi: 10.1001/jama.288.16.1987. doi: joc20547. [DOI] [PubMed] [Google Scholar]
  4. Amaravadi RK, Dimick JB, Pronovost PJ, Lipsett PA. ICU nurse-to-patient ratio is associated with complications and resource use after esophagectomy. Intensive Care Medicine. 2000;26:1857–1862. doi: 10.1007/s001340000720. [DOI] [PubMed] [Google Scholar]
  5. Barnett MJ, Kaboli PJ, Sirio CA, Rosenthal GE. Day of the week of intensive care admission and patient outcomes: A multisite regional evaluation. Medical Care. 2002;40:530–539. doi: 10.1097/00005650-200206000-00010. [DOI] [PubMed] [Google Scholar]
  6. Becker DJ. Do hospitals provide lower quality care on weekends? Health Services Research. 2007;42:1589–1612. doi: 10.1111/j.1475-6773.2006.00663.x. doi: 10.1111/j.1475-6773.2006.00663.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Becker G. Human capital: A theoretical and empirical analysis. Columbia University Press; New York: 1964. NBER General Series No.8. National Bureau of Economic Research. [Google Scholar]
  8. Blegen MA, Goode CJ, Park SH, Vaughn T, Spetz J. Baccalaureate education in nursing and patient outcomes. Journal of Nursing Administration. 2013;43:89–94. doi: 10.1097/NNA.0b013e31827f2028. doi: 10.1097/NNA.0b013e31827f2028. [DOI] [PubMed] [Google Scholar]
  9. Bhonagiri D, Pilcher DV, Bailey MJ. Increased mortality associated with after-hours and weekend admission to the intensive care unit: a retrospective analysis. Medical Journal of Australia. 2011;194:287–292. doi: 10.5694/j.1326-5377.2011.tb02976.x. [DOI] [PubMed] [Google Scholar]
  10. Buhr KJ. Does increasing education increase the probability of promotion? The case of registered nurses in Canada. Journal of Continuing Education in Nursing. 2010;41:123–132. doi: 10.3928/00220124-20100224-05. doi: 10.3928/00220124-20100224-05. [DOI] [PubMed] [Google Scholar]
  11. Caughey AB, Urato AC, Lee KA, Thiet MP, Washington AE, Laros RK., Jr Time of delivery and neonatal morbidity and mortality. American Journal of Obstetrics and Gynecology. 2008;199:491–495. doi: 10.1016/j.ajog.2008.03.046. [DOI] [PubMed] [Google Scholar]
  12. Clarke SP. Registered nurse staffing and patient outcomes in acute care: looking back, pushing forward. Medical Care. 2007;45:1126–1128. doi: 10.1097/MLR.0b013e31815ccaaf. doi: 10.1097/MLR.0b013e31815ccaaf. [DOI] [PubMed] [Google Scholar]
  13. de Cordova PB, Phibbs CS, Bartel AP, Stone PW. Twenty-four/seven: A mixed-method systematic review of the off-shift literature. Journal of Advanced Nursing. 2012;68:1454–1468. doi: 10.1111/j.1365-2648.2012.05976.x. doi: 10.1111/j.1365-2648.2012.05976.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dimick JB, Swoboda SM, Pronovost PJ, Lipsett PA. Effect of nurse-to-patient ratio in the intensive care unit on pulmonary complications and resource use after hepatectomy. American Journal of Critical Care. 2001;10:376–382. [PubMed] [Google Scholar]
  15. Glance LG, Dick AW, Osler TM, Mukamel DB, Li Y, Stone PW. The association between nurse staffing and hospital outcomes in injured patients. BMC Health Services Research. 2012;12:247–254. doi: 10.1186/1472-6963-12-247. doi: 10.1186/1472-6963-12-247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Horn SD. The business case for nursing in long-term care. Policy, Politics, and Nursing Practice. 2008;9:88–93. doi: 10.1177/1527154408320420. doi: 10.1177/1527154408320420. [DOI] [PubMed] [Google Scholar]
  17. Institute of Medicine . The future of nursing: Leading change, advancing health. The National Academies Press; Washington, DC: 2010. [PubMed] [Google Scholar]
  18. Kalisch BJ, Lee KH. The impact of teamwork on missed nursing care. Nursing Outlook. 2010;58:233–241. doi: 10.1016/j.outlook.2010.06.004. doi: S0029-6554(10)00266-6 [pii]10.1016/j.outlook.2010.06.004. [DOI] [PubMed] [Google Scholar]
  19. Kalisch BJ, Lee KH. Nurse staffing levels and teamwork: a cross-sectional study of patient care units in acute care hospitals. Journal of Nursing Scholarship. 2011;43:82–88. doi: 10.1111/j.1547-5069.2010.01375.x. doi: 10.1111/j.1547-5069.2010.01375.x [doi] [DOI] [PubMed] [Google Scholar]
  20. Kane RL, Shamliyan TA, Mueller C, Duval S, Wilt TJ. The association of registered nurse staffing levels and patient outcomes: Systematic review and meta-analysis. Medical Care. 2007;45:1195–1204. doi: 10.1097/MLR.0b013e3181468ca3. doi: 10.1097/MLR.0b013e3181468ca3 [doi]00005650-200712000-00011 [pii] [DOI] [PubMed] [Google Scholar]
  21. Kim H, Harrington C, Greene WH. Registered nurse staffing mix and quality of care in nursing homes: A longitudinal analysis. Gerontologist. 2009;49:81–90. doi: 10.1093/geront/gnp014. doi: 10.1093/geront/gnp014. [DOI] [PubMed] [Google Scholar]
  22. Konetzka RT, Stearns SC, Park J. The staffing-outcomes relationship in nursing homes. Health Services Research. 2008;43:1025–1042. doi: 10.1111/j.1475-6773.2007.00803.x. doi: 10.1111/j.1475-6773.2007.00803.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Laupland KB, Shahpori R, Kirkpatrick AW, Stelfox HT. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. Journal of Critical Care. 2008;23:317–324. doi: 10.1016/j.jcrc.2007.09.001. [DOI] [PubMed] [Google Scholar]
  24. Liang YW, Tsay SF, Chen WY. Effects of nurse staffing ratios on patient mortality in Taiwan acute care hospitals: A longitudinal study. Journal of Nursing Research. 2012;20:1–7. doi: 10.1097/JNR.0b013e3182466ddc. doi: 10.1097/JNR.0b013e3182466ddc [doi] 00134372-201203000-00003 [pii] [DOI] [PubMed] [Google Scholar]
  25. Manojlovich M, Sidani S, Covell CL, Antonakos CL. Nurse dose: linking staffing variables to adverse patient outcomes. Nursing Research. 2011;60:214–220. doi: 10.1097/NNR.0b013e31822228dc. doi: 10.1097/NNR.0b013e31822228dc [doi] [DOI] [PubMed] [Google Scholar]
  26. Needleman J, Buerhaus P, Pankratz VS, Leibson CL, Stevens SR, Harris M. Nurse staffing and inpatient hospital mortality. New England Journal of Medicine. 2011;364:1037–1045. doi: 10.1056/NEJMsa1001025. doi: 10.1056/NEJMsa1001025 [doi] [DOI] [PubMed] [Google Scholar]
  27. NQF National Quality Forum Endorses Consensus Standards for Quality of Hospital Care. 2008 Retrieved from http://www.qualityforum.org/projects/hospital_outcomes-and-efficiency_I.aspx.aspx.
  28. Peberdy MA, Ornato JP, Larkin GL, Braithwaite RS, Kashner TM, Carey SM, Berg RA. Survival from in-hospital cardiac arrest during nights and weekends. JAMA. 2008;299:785–792. doi: 10.1001/jama.299.7.785. doi: 10.1001/jama.299.7.785. [DOI] [PubMed] [Google Scholar]
  29. Stone PW, Pogorzelska M, Kunches L, Hirschhorn LR. Hospital staffing and health care-associated infections: a systematic review of the literature. Clinical Infectious Disease. 2008;47:937–944. doi: 10.1086/591696. doi: 10.1086/591696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. VA Careers Why VA? Become a VA nurse. 2013 Retrieved from http://www.vacareers.va.gov/assets/common/print/Nursing_Brochure.pdf.
  31. Weston MJ, Brewer KC, Peterson CA. ANA principles: the framework for nurse staffing to positively impact outcomes. Nursing Economic$ 2012;30:247–252. [PubMed] [Google Scholar]
  32. Wiltse Nicely KL, Sloane DM, Aiken LH. Lower mortality for abdominal aortic aneurysm repair in high-volume hospitals is contingent upon nurse staffing. Health Services Research. 2013;48:972–991. doi: 10.1111/1475-6773.12004. doi: 10.1111/1475-6773.12004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Zhan C, Miller MR. Excess length of stay, charges, and mortality attirbutable to medical injuries during hospitalization. JAMA. 2003;290:1868–1874. doi: 10.1001/jama.290.14.1868. doi:10.1001/jama.290.14.1868. [DOI] [PubMed] [Google Scholar]

RESOURCES