Volume 47
&
Number 3
&
June 2015
Measuring Workload of Nurses on a Neurosurgical Care Unit John Nelson, Linda Valentino, Laura Iacono, Peter Ropollo, Natalia Cineas, Stephanie Stuart
ABSTRACT Aim of the Study: The aim of this study was to create a model of workload that could be used to manage workload and increase satisfaction of workload for nurses on a neuroscience care unit. Background: No study was found that delineated a model of workload that could be used to manage or improve satisfaction with workload for a neuroscience care unit at either the individual nurse or unit level. Methods: Staff, management, and a researcher collaboratively developed a model to examine workload on a neuroscience care unit. Forty-three independent variables of workload and the dependent variable of satisfaction with workload were studied over 28 days using stepwise regression. Stepwise regression is appropriate for model building. Criteria to enter any independent variable into a regression equation included correlating with the dependent variable of satisfaction with workload, validation of central tendency assumptions, and good data fit using residual diagnostics. Results: Independent variables of workload that explained the variance of satisfaction with workload included time (15.9%), undelegated work (4.0%), number of isolation patients (2.9%), individual employees (2.1%), number of patients (1.3%), and number of postoperative neurosurgical patients (1.1%). On the unit level, satisfaction with workload was predicted by time (42.5%) and the number of nurses on duty (7.7%). Conclusions: Satisfaction with workload as reported by staff nurses is predicted by both individual- and unit-level factors of workload. Staff input is crucial to the development of a model of workload on clinical specialty units like neuroscience care. Staff nurses identify key variables, otherwise overlooked, affecting workload and satisfaction and satisfaction with workload. Implications for Nursing Management: It is vital to develop unit-specific models of workload and consider both individual- and unit-level factors. Such models have potential for deeper research into both management and increasing satisfaction of workload at the level of clinical specialty/unit. Keywords: action research, nurse satisfaction, specified models, staff nurses, workload
Questions or comments about this article may be directed to John Nelson, PhD, at [email protected]. He is the President, Healthcare Environment, Inc., New Brighton, MN. Linda Valentino, MSN RN, is the Executive Vice President, Visiting Nurse Service of New York, New York, NY. Laura Iacono, MSN RN, is the Nurse Manager Neurosurgical Intensive Care Unit, Department of Nursing, North Shore University Hospital, Manhasset, NY. Peter Ropollo, BA DC BSN MPA, is a Nursing Administrator, New York Presbyterian Hospital, Columbia University Medical Center, New York, NY. Natalia Cineas, MS RN, is the Senior Director of Nursing, Mount Sinai St. Luke’s Hospital and Mount Sinai Roosevelt, New York, NY. Stephanie Stuart, BSN RN, is a Staff Nurse, Neurosurgery Care Unit, New York Presbyterian Hospital, Columbia Medical Center, New York, NY. No funding was provided for this study outside of investigators making their resources, including time, available to conduct this study. Study facility allowed nurses to complete their surveys over the 28-day study period. No conflict of interest was identified for authors of this study outside any recognition for conducting this study. No company supplied any additional resource to conduct this study. Copyright B 2015 American Association of Neuroscience Nurses DOI: 10.1097/JNN.0000000000000136
W
orkload for this study is defined as ‘‘the perceived amount of work in terms of pace and volume’’ (Spector, 1997, p. 358). Pace and volume of workload varies between professional roles (Jenaro, Flores, Orgaz, & Cruz, 2011) and from day to day (Mittman, Seung, Pisterzi, Isogoi, & Michaels, 2008) making measurement an amorphous process. Workload has been depicted using several models that configure nurses’ tasks, intensity of tasks, staffing/ resources, and external contributing factors (Morris, MacNeela, Scott, Treacy, & Hyde, 2007). The construct of workload is complex and context dependent, thus requiring careful consideration if workload is to be measured properly, delegated appropriately, and reimbursed correctly (Morris et al., 2007). The primary intent of this study was to show how workload was defined by frontline care providers and use the data to develop a model of workload specified within the context of a neurosurgical unit. Authors of this study used sociotechnical systems theory as a guide for staff, management, and researchers to develop a model of workload specified for the neurosurgical inpatient unit. It is hoped that creating models of workload specified at the unit/context
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E9
E10
Journal of Neuroscience Nursing
level will aid with both management and improvement of satisfaction of workload. Staff were the primary resource in identifying the pace and elements of work within the process of developing the model of workload for nurses on a neurosurgical care unit.
Background Several contributors to workload of nurses have been identified (Al-Kandari & Thomas, 2008; Henderson, Willis, Walter, & Toffoli, 2008; Kiekkas et al., 2007; Sawatzky-Dickson & Bodnaryk, 2009). Al-Kandari and Thomas (2008) found that several workload items had a statistically significant relationship with adverse patient events. Workload items included nurseYpatient load, number of unstable patients, occupancy rate of the unit/ward, and critical patient events (Al-Kandari & Thomas, 2008). Workload has been found to vary based on clinical specialty, as identified in a study in community health workload (Henderson et al., 2008). Henderson et al. found that community care workload included complexity of patient care, duration of the home visit, client support within the home, level of dependency, geographic location of the home, transfer to hospital for care if needed, case management availability, and number of activities carried out within home during each visit (Henderson et al., 2008). Workload in the intensive care unit (ICU) setting has been linked with acuity of patient, specifically because it relates to respiratory status of the patient as well as the diagnostics and treatments specific to the patient in the ICU (Kiekkas et al., 2007). Workload specific to the neonatal setting has been identified to include time for nursing care, documentation, communication, and observation of the patient (Sawatzky-Dickson & Bodnaryk, 2009). The most common elements of workload identified by nurses in Kuwait (n = 780) working on medical (n = 417) or surgical units (n = 343) included medication administration, patient assessment, writing/updating of care plans, patient monitoring, and patient education (Al-Kandari & Thomas, 2008). In addition to clinical setting providing specific dimensions of workload, type of patient has also been cited as a contributing factor for nurses’ perception of workload (Mittmann, Seung, Pisterzi, Isogai, & Michaels, 2008). Patients after myocardial infarction required the most care in an acute care setting (2.5 T 2.3 hours of care per patient day [ppd]) followed by patients treated for pneumonia (2.4 T 2.0 hours ppd), patients with diabetes (2.0 T 1.3 hours ppd), patients with a schizophrenia diagnosis (1.72 T 2.0 hours ppd), and patients with stroke (1.7 T 1.5 hours ppd; Mittmann et al., 2008). Job-related tasks have also been shown to contribute to workload. For example, in a neuro-rehabilitation setting, 8883 nursing activities were examined, which
The primary intent of this study is to create a definition of workload that reflects the experience of frontline staff who provide care in a neurosurgical unit. revealed that 46% of the tasks were for direct patient care, 25% were for indirect care, 10% were unrelated to patient care, and 19% were personal-related time (Williams, Harris, & Turner-Stokes, 2009). Specific tasks can add to workload such as medically induced hypothermia (Olson, Kelly, Washam, & Thoyre, 2008). While working with patients with hypothermia, nurses identified the need to know how to maintain a desired body temperature (Olson et al., 2008). Each intervention to induce hypothermia was shown to have statistically significant differences regarding efficacy of effecting body temperature. The 10 different interventions to maintain a specified body temperature during induced hypothermia included removing bed covers, administering acetaminophen of ibuprofen, using a circulating fan, placing ice bags in patients’ axilla and/or groin area, sponge bathing, adjusting the room thermostat, using warm intravenous fluids, using warmed blankets, adjusting sedation, and circulating of hot air (Olson et al., 2008). Learning the methods and association efficacy of each intervention during induced hypothermia takes time as does the application of one or many of the interventions (Olson et al., 2008). In a study of 539 nurses in Finland, nonpatient care tasks shown to increase workload included organizing the work to be done, staff resources, and mental stress and disharmony with other units/departments (Rauhala & Fagrstrom, 2007). Gurses, Carayon, and Wall (2009) identified several obstacles for nurses completing their work, including precepting new nurses, family-related issues, poor working environment, poor handoff, poor working relationships, poor resources, inability to find patient charts, and poorly stocked rooms. Transporting patients has also been reported to impede the ability for nurses to care for patients (Nuikka, Paunomen, Hanninen, & Lansimies, 2001). In an analysis of 453 staff nurse tasks on a day shift, transporting patients to tests and procedures consumed 4.4% of the staff nurses time during the morning shift (Nuikka et al., 2001). Transportation was identified as a physically strenuous task that contributed to their perception of heavy workload (Nuikka et al., 2001). Considering Spector’s (1997) definition of workload, as cited earlier, the variance of number of tasks
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
and associated intensity of work for each task and the perceived level of workload from day to day are likely to vary dramatically. This is compounded by the ability of other professionals surrounding the nurse who may or may not be able to provide support to accomplish the work (Jenaro et al., 2011).
Outcomes of Workload Perceived unmanageable workload has been associated with higher levels of complications for patients 24 hours after ICU discharge (Apostopoulou, Georgoudi, Tsaras, & Veldekis, 2006). Specific post-ICU discharge complications examined include pneumonia, bloodstream infection, urinary tract infection, surgical infections, pressure sores, central venous system infection, hydrothorax, arrhythmia, atelectasis, bowel obstruction, renal failure, and pneumothorax (Apostopoulou et al., 2006). Moderate-to-extreme stress was reported by staff to relate to workload (n = 398, 77.8%; Lu, While, & Barriball, 2007). Rauhala et al. (2007) found a statistically significant relationship between workload and sick time (p G .002). Nurses who reported the heaviest workload were 71%Y92% more likely to experience burnout and job dissatisfaction (Rafferty et al., 2007). Workload also had the strongest correlations with work environment satisfaction (r = .53; Lin & Cai, 2009). A recent study revealed that workload was the most influential reason for intent to stay with perception of increased workload decreasing the intent to stay (Mrayyan, 2007). The relationship between workload and job satisfaction is inconsistent (Kovner, Brewer, Wu, Cheng, & Suzuki, 2006). The inconsistent findings related to nurse outcomes of job satisfaction may be secondary to insufficient methods of measuring workload.
Measuring Workload Illness and acuity of the patient was the predominant method found in the literature to measure workload. The Acute Physiology and Chronic Health Evaluation Classification System, designed to measure illness severity and predict mortality in the intensive care setting, was used to evaluate hospital resources and nurse work time for each patient, but it is only one aspect of workload (Kiekkas et al., 2007). Another method to measure workload is the Nottingham Patient Dependency System, which scores nine areas of patient status including respiratory, cardiovascular, neurological, pain control, elimination, psychosocial, hygiene, feeding, and mobility needs (Kiekkas et al., 2007). These clinical severity measures were shown to relate to workload of nurses (Kiekkas et al., 2007). Acuity systems have been proposed as a way to make judgments regarding how much staffing is needed for care (Graf, Millar, Feilteau, Coakley, & Erickson, 2003). However,
&
Number 3
&
June 2015
patient illness and acuity has been found to be insufficient because it does not integrate dimensions like individual patient coping with illness that can make care more complex and thus increase workload (Spence et al., 2006). The Workload Management System for Critical Care Nurses was developed to measure workload in critical care (Cho, Shin, Chio, Jung, & Lee, 2006). The Professional Assessment of Optimal Nursing Care Intensity Level is another method to measure nurse intensity (Frilund & Fagerstrom, 2009). The Workload Intensity Measurement System uses nursing diagnosis to measure workload of nurses (Hoi, Ismail, Ong, & Kang, 2010). The Community Client Need Classification System was developed to measure workload of community nurses (Brady, Byrne, Horan, Macgregor, & Begley, 2008). The Workload Measurement and Reporting System was developed to measure workload in the ambulatory setting (Dickson, Cramer, & Peckham, 2010). The primary challenge with measures, like the ones listed herein, is either projected or retrospective and does not evaluate workload in real time (Lin & Cai, 2009). In addition, identified models may be misspecified because they were not designed with the specificities of neuroscience care units in mind.
Method Considering there were no existing measures of workload developed for a neurosurgical care unit, validity of the measure was dependent on careful identification, articulation, and verification of the dimensions and intensity of each item of work conducted by nurses on this neurosurgical care unit. The Unit Practice Council (UPC) for this neurosurgical care unit was deemed as the source that would be able to provide the desired information. In 2007, the nurses on a neurosurgical inpatient unit initiated a UPC as an outgrowth of a shared governance model. The UPC included staff nurses, nurse support staff, and the unit manager who were the primary drivers of this research project to develop a workload model. Input from all three levels of staff was consistent with the sociotechnical systems framework that proposes that both staff and management should have input into refinement of work operations (Maxwell, Ziegenfuss, & Chisholm, 1993). Collaboratively, the UPC identified 43 elements they perceived contributed to workload of staff nurses on the unit. All four members of the UPC were in 100% agreement that all 43 items were essential to include in the measure if their workload was to be accurately measured. The 43 items were categorized into nine categories, including patient load (eight items); charting (one item); computer system (three items); process of care (one item) ; delegation (two items); admissions,
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E11
E12
Journal of Neuroscience Nursing
discharges, and transfers (six items); total number of supply carts and supply rooms on the unit (one item); and support staff (five items). There were also 16 specific care tasks unique for the neurosurgical care unit. The first eight categories were rank ordered from the most contributory to workload to the least contributory and then numbered 1Y8 (see Figure 1). The 43 items were used to develop a survey for staff nurses and a survey for charge nurses. Staff nurses would be asked to report on each activity each day regarding the amount of time spent, measured in minutes, for each of the 43 items. Charge nurses also completed a survey for staff nurses because charge nurses not only are responsible for the overall operations of the unit but also take a number of patients to care for during their shift of work. In addition to completing a staff nurse survey, charge nurses were also asked to report the cumulative activity specific to the 43 items for the overall unit. For example, a staff nurse would report the time in minutes spent on a ventilator for their patient, whereas the charge nurse would report the number of total vents for the unit. It was proposed by the members of the UPC that overall activity on the unit impacted their work because it determined the level of resources available to them to help if help was needed to carry out patient care. The final dimension of workload that would be measured was time, which has been a neglected component of social science (McGrath & Tschan, 2004). Time for this study is defined operationally as the progression of days within the 30-day period this study took place. Every nurse on the neuroscience unit was asked to respond to the workload measure every day of work, within the 30-day study. According to McGrath and Tschan (2004), social change occurs at microlevel, mesolevel, and macrolevel. Thus, the monitoring of the day of study would be included in the regression equation to understand how time itself related to the perception of workload. This descriptive study used correlations and stepwise regression equations to interpret the data. Stepwise regression was selected for this study because of the current lack of data about satisfaction with workload on a neurosurgical care unit, lack of theory testing, and lack of understanding of independent variables that correlate with satisfaction with workload. Stepwise regression is the most appropriate regression procedure to use when the researcher wants to do model building, not model testing (Tabachnick & Fidell, 2007). Correlates that were statistically significant were entered into the stepwise regression in order of strength of correlation with the strongest correlation being entered first. Diagnostic residuals were used to evaluate if the model fit improved or worsened upon entry of each variable. If the model fit worsened or the explained
variance declined, the variable was excluded from the final equation. An alpha of .05 was used for the level of significance. Power analysis was also conducted for both correlation procedures and stepwise regression. A sample size of 300 survey responses was essential for a power of 0.95 and effect size of 0.1 for the regression equation.
Setting The study took place on a 36-bed neurosurgical unit. The unit has six step-down beds, five ventilator beds, and one close observation room with a 1:4 ratio and utilizes a nursing assistant to monitor the patients for behavioral changes and assist with personal needs such as toileting, feeding, and mobility. The unit is the neurosurgical care unit in a large tertiary care center in New York City.
Sample Participant selection in the study was based on a convenience sample. Inclusion criteria included being a registered nurse (RN) on the unit, identified by the unit manager as regular staff, working part or full time. Exclusion criteria included temporary staff, orientees, or any regular staff on leave of any sort (e.g., family leave, sick leave, etc.) or on vacation during a portion of the study. There were 38 RNs who met the inclusion criteria and were at various levels of clinical development, which included 12 staff nurses, 21 clinical I nurses, 2 clinical II nurses, and 3 clinical III nurses. Each nurse was asked to respond and give their perception of workload and tasks performed during their shift over the 28-day period. The authors felt that a 28-day period would offer a realistic representation of nurses’ workload, reflecting high and low unit census and changes in patient acuity. All eligible nurses were surveyed over the 28-day period. All nurses on this nursing unit worked either the 12-hour day shifts or the 12-hour night shifts. There were 349 surveys sent out over this 28-day period.
Instrumentation The literature review did not reveal an instrument that would adequately represent the variables of workload identified by the study group; thus, an instrument was developed for examining the workload items specific to the unit. The consultant and expert in survey development developed a first draft of the survey that would query all the variables in the workload model. The survey was revised four times until all participants in the research study group reached full agreement that the survey accurately reflected the unit workload. The workload survey consisted of 47 variables and 1 openended question for comments on aspects of workload during the shift not identified in the survey. There
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
&
Number 3
&
June 2015
FIGURE 1 Facets of Workload Identified by Staff on Neurosurgical Care Unit
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E13
E14
Journal of Neuroscience Nursing
were six items for the dependent variable of workload, which is the subscale from the Nursing Work Index developed by Stamps and Piedmonte (1976). The Cronbach’s alpha has consistently been above .80, which was considered adequate for this study. The index developed by Stamps and Piedmonte has been used to measure workload (Lin & Cai, 2009).
Procedures After internal review board approval, surveys were sent electronically to participants’ hospital email addresses. The survey was sent out every day, every shift, to every nurse on the unit from September 8, 2009, to October 5, 2009. Staff were instructed to respond to the survey only on the days worked. Final results were compared with the responses in the data set to confirm that the employee was working that day. Responders could ‘‘save and return’’ to the survey as many times as necessary, depending on time available during the shift to complete the survey. Most nurses choose to complete the survey at the end of their shift when they had time to reflect on their patient assignment, workload, and shift events. Nurses received overtime pay if they stayed after their assigned shift to complete the survey. Staff hypothesized that workload was less on shifts when there was available staff (e.g., low nurseYpatient ratios or sufficient ancillary staff) compared with shifts where staffing was suboptimal. To validate the unit activity for the shift, the charge nurse was also given a survey to report on the number of admissions; transfers in and out; discharges; additional RNs added during the shift like float nurses or ‘‘call backs’’ (e.g., nurses called to work in the middle of the shift); number of support staff; total number of supply carts on the unit; number of patients at the start of the shift; number of available computers; number of ventilator patients; number of total care patients; number of patients with spinal drains; number of patients requiring point-ofcare testing (e.g., urine-specific gravity or blood glucose testing); number of patients receiving blood products; number of patients with neurological deficits; number of confused patients; number of combative patients; number of postoperative neurosurgical patients; number of patients on a urinary catheterization protocol; number of patients who required a 1:1 observation for delirium, agitation, or confusion; and number of patients at high risk for falls who required a 1:4 observation. Charge nurses also had the ability to comment on aspects that added to workload and what eased workload. These items were validated by unit leaders as contributing to overall workload. As the number of these items increased on the unit, the demand on staff resources increased. This added demand increased not only the unit workload but also the individual workload.
Results Response Rate There were 349 surveys sent out over this 28-day period, and 298 surveys were returned, representing an 85.4% response rate. Upon closer examination of the nonresponders, it was noted that there were only a few nurses who did not respond to the daily survey of workload, which accounted for most of the nonresponses. The workload scale revealed a low Chronbach’s alpha of the six-item workload scale, the dependent variable of this study (Chronbach’s alpha = .46). Three items were deleted, and Chronbach’s alpha increased to .71, which was considered acceptable for this study. The three final items inquired about staff nurses’ perception that level of workload facilitated good job performance, their current workload allowed enough time to discuss patient problems with colleagues, and the types of activities in the job were satisfying. Before running regression analysis, the relationship satisfaction with workload had with every independent variable was examined. Only independent variables that had statistically significant correlations with the dependent variable of satisfaction with workload would be entered into the final regression equation. Correlation analysis revealed 6 of 47 variables that related to workload and were statistically significant. Time had the most significant relationship with workload revealing a positive correlation of .41. Percentage of work that could be delegated to support staff but remained undelegated also correlated with workload, but this was found to be a negative relationship (i.e., work that could/ should be delegated but was not). The number of patients, including patients admitted, discharged, or transferred, was found to have a negative correlation (r = j.15). Satisfaction with the amount of paperwork had a positive correlation with workload. Finally, the more patients each nurse had with a neurosurgical diagnosis within their patient load, the more satisfied with workload the nurses felt. Pearson’s correlation value and p value for level of significance for each of the correlates of workload are noted below. b Time (day of the study): r = .408, p G .001 b Percent of work that could have been delegated but was not: r = j.261, p G .001; Spearman’s r = j.273, p G .001 b Number of patients, including ADT: r = j.148, p = .012; Spearman’s r = j.149, p = .012 b Number of isolation patients: r = j.137, p = .021; Spearman’s r = j.127, p = .032 b Satisfied with amount of paperwork: r = .128, p = .029; Spearman’s r = .174, p = .003 b Number of neuro postoperative patients: r = .122, p = .039; Spearman’s r = .116, p = .050
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
All independent variables that were found to have a statistically significant relationship with the dependent variable of satisfaction with workload were entered into a stepwise regression equation. Results revealed that time was the most predictive, explaining 16.2% of the variance of workload, followed by percentage of work that could have been delegated but was not, which explained 3.9% of the variance. The third predictor of workload in this study was the number of isolation patients, which explained 2.8% of the variance. Fourth was the number of patients, including the number of admissions, discharges, and transfers, which predicted 1.4% of the variance of workload. The final variable entered into the equation was number of postoperative neurosurgical patients, which predicted 1.5% of the variance of satisfaction with workload. In total, 25.8% of the variance of satisfaction with workload was explained by the work of the individual staff nurse (see Table 1). This second part of analysis included the examination of workload and satisfaction of workload reported by charge nurses. This study examined the unit-wide activity such as total number of ventilator patients, total number of isolation patients, and any other activity that utilized unit resources that could potentially impact the perception of nurse workload. The charge nurses provided the unit-specific data 91% of the time or 51 of the 56 surveys completed. There were six different nurses who worked as charge nurse on the day shift and nine different nurses who worked as charge nurse on the night shift. The correlates for unit data were entered into a stepwise regression. Forty-four of the 51 shifts where data were reported had adequate data to be included in the regression equation. Data were controlled for day/night shift as well as time. After each of the correlates of unit data were entered into a stepwise regression equation, it was noted that time, which is referred to as ‘‘wave,’’ predicted 42.5% of the variance. The number of nurses predicted 7.7% of the variance. It is obvious that what influenced the
TABLE 1.
&
Number 3
&
June 2015
perception of workload over the 28-day study needs to be better understood. Further inquiry is warranted (see Table 2). The investigators wanted to understand if there were differences in the activity on the unit when different charge nurses worked. It was the experience of the co-investigators that some charge nurses managed the units very efficiently and collaboratively with staff, whereas other charge nurses were not as effective. It was desired to know if the perception of workload of staff nurses differed when comparing shifts for each of the 15 charge nurses, and significant differences were found on the day shift using an analysis of variance procedure, using an alpha of .05. Significant differences were also found on the night shift when comparing charge nurse shifts using the least significant differences procedure. Again, an alpha of .05 was used. The box plot graph (Figure 2) reveals the dummy codes for each of the night-shift charge nurses and the variance of staff nurses’ report of workload when working with each charge nurse. It is noted that, for the night shift, the top two box plots scored over 5.0 on the 1Y7 Likert scale, with higher scores indicating more nurse satisfaction with workload. The statistical significance is noted by the asterisk to the right of the dummy code. The box plot graph (Figure 3) depicts data for nurse workload satisfaction by charge nurse for the day shift. It is noted that nurses reported more satisfaction when working with only one of the five assigned charge nurses. A score greater than 5.0 indicates a high degree of satisfaction. Again, the difference was statistically significant at the .05 level and is noted by an asterisk placed only by the top box plot dummy code. The two final variables that were found to correlate with workload were the total number of RNs working each shift (r = .30, p = .025) and the number of patients who were on 1:1 observation by support staff. These variables had a negative impact on perception of workload (r = j.34, p = .013). This is because of the frustration of taking support staff out of the staff
Regression Equation Workload: Staff Nurses Statistic
Step
Variable
Total R2
R2
Change
Beta
Significance
.402
.162
.162
52.144
.000
1
Time
2
Work not delegated
.449
.201
.039
13.262
.000
3
Number of isolation patients
.479
.230
.028
9.889
.002
4
Number of patients
.493
.243
.014
4.826
.029
5
Number of neuro postoperative patients
.508
.258
.015
5.334
.022
Note. Residual diagnostics were run, and no violations of assumptions were found.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E15
E16
Journal of Neuroscience Nursing
TABLE 2.
Regression Equation Workload: Charge Nurses Statistic
Step
Variable
Total R
2
R
2
Change
Beta
Significance
1
Time
.652
.425
.425
31.060
.000
2
Number of RNs
.709
.502
.077
6.370
.016
Note. Residual diagnostics were run, and no violations of assumptions were found.
count so a nurse attendant can sit with one patient. Further inquiry is warranted in this area. The final unit-level variable that was found to have a positive relationship with workload was the number of RNs on duty. This excluded nurses who were called in to work during the shift. This final variable had an r value of .286. In layman’s terms, this would mean 28.6% of the time that, when the number of RNs increased, satisfaction with workload also increased. That is, however, those two variables alone being considered and not everything else that is factored into the final regression equation. This is the final model that integrated the individual nurse data, unit-level data, and aggregate (charge nurse) data from other studies conducted within New York Presbyterian Hospital. It is feasible to consider that up to 50% or 60% of the variance could be explained for workload on this neurosurgical unit (Figure 4).
Discussion The study revealed a statistically significant difference in staff satisfaction with workload when comparing shifts led by different charge nurses. Charge nurses need to be accomplished in team building, prioritization, communication, and delegation skills. Furthermore, charge nurses need to possess a high degree of
clinical knowledge making them proficient or expert in their specialty. There must be a concerted effort to have a strong charge nurse with insight and situational awareness to assess and anticipate the unit’s activities for each shift. The primary tasks of the RN must be taken into consideration by the charge nurses when making patient assignments. Primary RN tasks within the study institution at the time of this study included catheterization protocol, vented patients, or spinal drains. In addition, every effort must be made for continuity of care, to ensure nurses who are working 2 or 3 days in a row keep their same patient assignment. In turn, charge nurses make every effort to see that neurosurgical patients are admitted to 8HN because this is 8HN’s specialty. The staff undergo lengthy and advanced training to care for the neurosurgical population and take great pride in their specialty. Nurses working in highly specialized areas are recognized by their peers and nursing leadership for their expert clinical knowledge and take immense pride and ownership in their patient population. This substantiates the value of geographic localization for very specialized patient populations. The data revealed that perception of workload decreased when nurses were caring for their specialty population compared with when they were caring for overflow
FIGURE 2 Box Plots Revealing Differences of Workload by Night Charge Nurse
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
&
Number 3
&
June 2015
FIGURE 3 Box Plots Revealing Differences of Workload by Day Charge Nurse
patients from other specialties such as medicine or general surgery. In today’s interdependent healthcare environment, there is a shared responsibility for getting things done. It is no longer reasonable to expect nurses to do it all. Nurses need training and direction in how and what to delegate and to whom. It is also important that nurses understand the breadth, licensure, and certification of each role, including their own. Effective delegation is about sharing workload, with the added bonus of developing skills and responsibility in others. Nurses who learn to effectively delegate not only to help themselves but also to help their patients get better care. Time as a predictor of workload was explained by staff who listened to the study results. First was the impact of study itself. Initially, staff felt that the survey
was frustrating and added to their perception of workload. However, the staff reported they wanted to communicate to the researchers as accurately as possible what was happening within each day they worked. One nurse even called the outside consultant working with this study to inquire how to submit her survey. She had worked for 12 hours and 2 hours of overtime but now wanted to complete the survey. When it was explained to her by the consultant that the survey was voluntary and not mandatory and, thus if too tired, she could forego responding, she replied: ‘‘I had a really bad day and I want to take time to tell you about it in the survey.’’ Despite the frustration of responding to the survey at the end of the day, the staff nurses informed the investigators that, eventually, it became part of the day and quite easy to respond to. This may,
FIGURE 4 Final Model of Workload
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E17
E18
Journal of Neuroscience Nursing
in part, explain why workload improved over time as the staff became used to responding to the survey. The second reason time played an important role in the study is secondary to the survey itself providing insight into how they could work better with staff. After data collection was completed, staff informed study investigators that the survey itself helped them understand that they were not using support staff adequately to delegate nonclinical tasks. The survey itself was not instructive, but the staff reported that the questions made them ponder how they use the support staff. Thus, over the period of the study, they began to use support staff differently by discussing and delegating tasks that support staff could do. The result of the discussion between RNs and support about tasks resulted in more effective use of the support staff role, which lightened the workload of the RN. The findings revealed what staff suspected regarding 1:1 patient observation. Satisfaction rates were lower when support staff were required to 1:1 sit with patients as it takes away from the unit staffing. Workload is a concept that is real yet hard to quantify because of the many variables that impact the design. To further study the variable effects on workload, it is suggested that each of these processes be automatically monitored in real time using a data management/computer interface system. Each time a patient is admitted or transferred, the system would collect data surrounding the event in real time. It would include the number of patients per nurse, number of admits per nurse, and number of transfers on and off the unit. The process could also do patient tracking to determine if a nurse has cared for a patient in the past, whether it be during the patient’s current hospitalization or a past admission. The data management system can also monitor of how much time the nurse spends documenting in the electronic medical record. An automated model would include the following variables: 1. 2. 3. 4. 5. 6. 7. 8. 9.
Number of patients per nurse Number of admits per nurse Number of transfers per nurse on the unit Number of transfers per nurse off the unit Number of discharges per nurse Number of support staff per nurse Number of isolation patients Number of neuro postoperative patients Number of days nurse took care of the same patient, both consecutive and sporadic 10. Documentation by nurse about patient (including notes about patient to other disciplines, number of referrals, number of documented conversations with family, and time charting on each patient).
Staff can continue to propose other variables that they hypothesize to relate to this model. This will facilitate simulation modeling to evaluate how changes in an independent variable impact an outcome variable under study.
Conclusion The workload study has helped this unit by allowing the nurses to identify key components impacting their workday. Since the completion of this study, many factors have been implemented to reduce the variables negatively affecting workload. For example, an additional nurse attendant was hired during the 11 P.M.Y 7 A.M. shift. The hiring of another nurse attendant has allowed nurses to receive more help with their daily tasks and assistance with patients who require total care. The unit also added an ICU technician to the staff. The skill set of ICU technicians allows for delegation of point-of-care testing and 12 lead electrocardiograms. The ICU technician is able to help with many tasks in the step-down unit where the patients are of higher acuity. In addition to increased staff, the unit also added an additional supply cart in the 1:4 observation room and has partnered with the transport department to have more wheelchairs available on site. These changes afford less travel back and forth for supplies and equipment and more time dedicated to patient care. The designation of computers during specific time intervals for nursing has also allowed more time for nurses to document during peak usage hours. The workload study has allowed management and staff to collaborate and bring forth positive unit changes. This study has allowed an opportunity not only to confront and solve workload challenges but also to make improvements leading to patient and staff satisfaction.
References Al-Kandari, F., & Thomas, D. (2008). Perceived adverse patient outcomes correlated to nurses’ workload in medical and surgical wards of selected hospitals in Kuwait. Journal of Clinical Nursing, 18(4), 581Y590. Apostolopoulou, E., Georgoudi, E., Tsaras, K., & Veldekis, D. (2006). Post-ICU complications in critically ill patients: The impact of nursing workload the first day on the ward. ICUs & Nursing Web Journal, AugYOct(27), 6 pp. Brady, A., Byrne, G., Horan, P., Macgregor, C., & Begley, C. (2008). Reliability and validity of the CCNCS: A dependency workload measurement system. Journal of Clinical Nursing, 17(10), 1351Y1360. Cho, Y., Shin, H., Chio, J., Jung, M., & Lee, B. (2006). Development and application of the Workload Management System for Critical Care Nurses (WMSCN) using the Workload Management System for Nurses (WMSN). American Journal of Critical Care, 15(3), 325Y325. Dickson, K. L., Cramer, A. M., & Peckham, C. M. (2010). Nursing workload measurement in ambulatory care. Nursing Economic$, 28(1), 37Y43.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
Frilund, M., & Fagerstrom, L. (2009). Managing the optimal workload by the PAONCIL methodVA challenge for nursing leadership in care of older people. Journal of Nursing Management, 17(4), 426Y434. Graf, C. M., Millar, S., Feilteau, C., Coakley, P. J., & Erickson, J. I. (2003). Patients’ needs for nursing care. Journal of Nursing Administration, 33(2), 76Y81. Gurses, A. P., Carayon, P., & Wall, M. (2009). Impact of performance obstacles on intensive care nurses’ workload, perceived quality and safety of care, and quality of working life. Health Services Research, 44(2 Pt 1), 422Y443. Henderson, J., Willis, E., Walter, B., & Toffoli, L. (2008). Measuring the workload of community mental health nurses: A review of the literature. Contemporary Nurse: A Journal for the Australian Nursing Profession, 29(1), 32Y42. Hoi, S. Y., Ismail, N., Ong, L. C., & Kang, J. (2010). Determining nurse staffing needs: The workload intensity measurement system. Journal of Nursing Management, 18(1), 44Y53. Jenaro, C., Flores, N., Orgaz, M. B., & Cruz, M. (2011). Vigour and dedication in nursing professionals: Towards a better understanding of work engagement. Journal of Advanced Nursing, 67(4), 865Y875. doi:10.1111/j.1365-2648.2010.05526.x Kiekkas, P., Brokalaki, H., Manolis, E., Samios, A., Skartsani, C., & Baltopoulos, G. (2007). Patient severity as an indicator of nursing workload in the intensive care unit. Nursing in Critical Care, 12(1), 34Y41. Kovner, C., Brewer, C., Wu, Y., Cheng, Y., & Suzuki, M. (2006). Factors associated with work satisfaction of registered nurses. Journal of Nursing Scholarship, 38(1), 71Y79. Lin, Y., & Cai, H. (2009). A method for building a real-time cluster-based continuous mental workload scale. Theoretical Issues in Ergonomics Science, 10(6), 531Y543. Lu, H., While, A. E., & Barriball, K. L. (2007). A model of job satisfaction of nurses: A reflection of nurses’ working lives in Mainland China Journal of Advanced Nursing, 58(5), 468Y479. Maxwell, C. I., Ziegenfuss, J. T., & Chisholm, R. F. (1993). Beyond quality improvement teams: Sociotechnical systems theory and self-directed work teams. Quality Management in Health Care, 1(2), 59Y67. McGrath, J. E., & Tschan, F. (2004). Temporal matters in social psychology. Washington DC: American Psychological Association. Mittmann, N., Seung, S. J., Pisterzi, L. F., Isogai, P. K., & Michaels, D. (2008). Nursing workload associated with hospital patient care. Disease Management and Health Outcomes, 16(1), 53Y61.
&
Number 3
&
June 2015
Morris, R., MacNeela, P., Scott, A., Treacy, P., & Hyde, A. (2007). Reconsidering the conceptualization of nursing workload: Literature review. Journal of Advanced Nursing, 57(5), 463Y471. Mrayyan, M. T. (2007). Jordanian nurses’ job satisfaction and intent to stay: Comparing teaching and non-teaching hospitals. Journal of Professional Nursing, 23(3), 125Y136. Nuikka, M., Paunonen, M., Hanninen O., & Lansimies, E. (2001). The nurse’s workload in care situations. Journal of Advanced Nursing, 33(3), 406Y415. Olson, D. M., Kelly, A. P., Washam, N. C., & Thoyre, S. M. (2008). Critical care nurses’ workload estimates for managing patients during induced hypothermia. Nursing in Critical Care, 13(6), 305Y309. Rafferty, A. M., Clarke, S. P., Coles, J., Ball, J., James, P., McKee, M., & Aiken, L. H. (2007). Outcomes of variation in hospital nurse staffing in English hospitals: Cross-sectional analysis of survey data and discharge records. International Journal of Nursing Studies, 44(2), 175Y182. Rauhala, A., & Fagerstrom, L. (2007). Are nurses’ assessments of their workload affected by non-patient factors? An analysis of the RAFAELA system. Journal of Nursing Management, 15(5), 490Y499. Rauhala, A., Kivimaki, M., Fagerstrom, L., Elovainio, M., Virtanen, M., Vahtera, J., I Kinnunen, J. (2007). What degree of work overload is likely to cause increased sickness absenteeism among nurses? Evidence from the RAFAELA patient classification system. Journal of Advanced Nursing, 57(3), 286Y295. Sawatzky-Dickson, D., & Bodnaryk, K. (2009). Validation of a tool to measure neonatal nursing workload. Journal of Nursing Management, 17(1), 84Y91. Spector, P. E. (1997). Job satisfaction: Application, assessment, cause, and consequences. London, UK: Sage. Spence, K., Tarnow-Mordi, W., Duncan, G., Jayasuryia, N., Elliott, J., King, J., & Kite, F. (2006). Measuring nursing workload in neonatal intensive care. Journal of Nursing Management, 14(3), 227Y234. Stamps, P. L., & Piedmonte, E. B. (1986). Nurses and work satisfaction: An index for measurement. Ann Arbor, MI: Health Administration Press Perspectives. Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics. Boston, MA: Pearson. Williams, H., Harris, R., & Turner-Stokes, L. (2009). Work sampling: A quantitative analysis of nursing activity in a neurorehabilitation setting. Journal of Advanced Nursing, 65(10), 2097Y2107.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E19
&
Number 3
&
June 2015
Measuring Workload of Nurses on a Neurosurgical Care Unit John Nelson, Linda Valentino, Laura Iacono, Peter Ropollo, Natalia Cineas, Stephanie Stuart
ABSTRACT Aim of the Study: The aim of this study was to create a model of workload that could be used to manage workload and increase satisfaction of workload for nurses on a neuroscience care unit. Background: No study was found that delineated a model of workload that could be used to manage or improve satisfaction with workload for a neuroscience care unit at either the individual nurse or unit level. Methods: Staff, management, and a researcher collaboratively developed a model to examine workload on a neuroscience care unit. Forty-three independent variables of workload and the dependent variable of satisfaction with workload were studied over 28 days using stepwise regression. Stepwise regression is appropriate for model building. Criteria to enter any independent variable into a regression equation included correlating with the dependent variable of satisfaction with workload, validation of central tendency assumptions, and good data fit using residual diagnostics. Results: Independent variables of workload that explained the variance of satisfaction with workload included time (15.9%), undelegated work (4.0%), number of isolation patients (2.9%), individual employees (2.1%), number of patients (1.3%), and number of postoperative neurosurgical patients (1.1%). On the unit level, satisfaction with workload was predicted by time (42.5%) and the number of nurses on duty (7.7%). Conclusions: Satisfaction with workload as reported by staff nurses is predicted by both individual- and unit-level factors of workload. Staff input is crucial to the development of a model of workload on clinical specialty units like neuroscience care. Staff nurses identify key variables, otherwise overlooked, affecting workload and satisfaction and satisfaction with workload. Implications for Nursing Management: It is vital to develop unit-specific models of workload and consider both individual- and unit-level factors. Such models have potential for deeper research into both management and increasing satisfaction of workload at the level of clinical specialty/unit. Keywords: action research, nurse satisfaction, specified models, staff nurses, workload
Questions or comments about this article may be directed to John Nelson, PhD, at [email protected]. He is the President, Healthcare Environment, Inc., New Brighton, MN. Linda Valentino, MSN RN, is the Executive Vice President, Visiting Nurse Service of New York, New York, NY. Laura Iacono, MSN RN, is the Nurse Manager Neurosurgical Intensive Care Unit, Department of Nursing, North Shore University Hospital, Manhasset, NY. Peter Ropollo, BA DC BSN MPA, is a Nursing Administrator, New York Presbyterian Hospital, Columbia University Medical Center, New York, NY. Natalia Cineas, MS RN, is the Senior Director of Nursing, Mount Sinai St. Luke’s Hospital and Mount Sinai Roosevelt, New York, NY. Stephanie Stuart, BSN RN, is a Staff Nurse, Neurosurgery Care Unit, New York Presbyterian Hospital, Columbia Medical Center, New York, NY. No funding was provided for this study outside of investigators making their resources, including time, available to conduct this study. Study facility allowed nurses to complete their surveys over the 28-day study period. No conflict of interest was identified for authors of this study outside any recognition for conducting this study. No company supplied any additional resource to conduct this study. Copyright B 2015 American Association of Neuroscience Nurses DOI: 10.1097/JNN.0000000000000136
W
orkload for this study is defined as ‘‘the perceived amount of work in terms of pace and volume’’ (Spector, 1997, p. 358). Pace and volume of workload varies between professional roles (Jenaro, Flores, Orgaz, & Cruz, 2011) and from day to day (Mittman, Seung, Pisterzi, Isogoi, & Michaels, 2008) making measurement an amorphous process. Workload has been depicted using several models that configure nurses’ tasks, intensity of tasks, staffing/ resources, and external contributing factors (Morris, MacNeela, Scott, Treacy, & Hyde, 2007). The construct of workload is complex and context dependent, thus requiring careful consideration if workload is to be measured properly, delegated appropriately, and reimbursed correctly (Morris et al., 2007). The primary intent of this study was to show how workload was defined by frontline care providers and use the data to develop a model of workload specified within the context of a neurosurgical unit. Authors of this study used sociotechnical systems theory as a guide for staff, management, and researchers to develop a model of workload specified for the neurosurgical inpatient unit. It is hoped that creating models of workload specified at the unit/context
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E9
E10
Journal of Neuroscience Nursing
level will aid with both management and improvement of satisfaction of workload. Staff were the primary resource in identifying the pace and elements of work within the process of developing the model of workload for nurses on a neurosurgical care unit.
Background Several contributors to workload of nurses have been identified (Al-Kandari & Thomas, 2008; Henderson, Willis, Walter, & Toffoli, 2008; Kiekkas et al., 2007; Sawatzky-Dickson & Bodnaryk, 2009). Al-Kandari and Thomas (2008) found that several workload items had a statistically significant relationship with adverse patient events. Workload items included nurseYpatient load, number of unstable patients, occupancy rate of the unit/ward, and critical patient events (Al-Kandari & Thomas, 2008). Workload has been found to vary based on clinical specialty, as identified in a study in community health workload (Henderson et al., 2008). Henderson et al. found that community care workload included complexity of patient care, duration of the home visit, client support within the home, level of dependency, geographic location of the home, transfer to hospital for care if needed, case management availability, and number of activities carried out within home during each visit (Henderson et al., 2008). Workload in the intensive care unit (ICU) setting has been linked with acuity of patient, specifically because it relates to respiratory status of the patient as well as the diagnostics and treatments specific to the patient in the ICU (Kiekkas et al., 2007). Workload specific to the neonatal setting has been identified to include time for nursing care, documentation, communication, and observation of the patient (Sawatzky-Dickson & Bodnaryk, 2009). The most common elements of workload identified by nurses in Kuwait (n = 780) working on medical (n = 417) or surgical units (n = 343) included medication administration, patient assessment, writing/updating of care plans, patient monitoring, and patient education (Al-Kandari & Thomas, 2008). In addition to clinical setting providing specific dimensions of workload, type of patient has also been cited as a contributing factor for nurses’ perception of workload (Mittmann, Seung, Pisterzi, Isogai, & Michaels, 2008). Patients after myocardial infarction required the most care in an acute care setting (2.5 T 2.3 hours of care per patient day [ppd]) followed by patients treated for pneumonia (2.4 T 2.0 hours ppd), patients with diabetes (2.0 T 1.3 hours ppd), patients with a schizophrenia diagnosis (1.72 T 2.0 hours ppd), and patients with stroke (1.7 T 1.5 hours ppd; Mittmann et al., 2008). Job-related tasks have also been shown to contribute to workload. For example, in a neuro-rehabilitation setting, 8883 nursing activities were examined, which
The primary intent of this study is to create a definition of workload that reflects the experience of frontline staff who provide care in a neurosurgical unit. revealed that 46% of the tasks were for direct patient care, 25% were for indirect care, 10% were unrelated to patient care, and 19% were personal-related time (Williams, Harris, & Turner-Stokes, 2009). Specific tasks can add to workload such as medically induced hypothermia (Olson, Kelly, Washam, & Thoyre, 2008). While working with patients with hypothermia, nurses identified the need to know how to maintain a desired body temperature (Olson et al., 2008). Each intervention to induce hypothermia was shown to have statistically significant differences regarding efficacy of effecting body temperature. The 10 different interventions to maintain a specified body temperature during induced hypothermia included removing bed covers, administering acetaminophen of ibuprofen, using a circulating fan, placing ice bags in patients’ axilla and/or groin area, sponge bathing, adjusting the room thermostat, using warm intravenous fluids, using warmed blankets, adjusting sedation, and circulating of hot air (Olson et al., 2008). Learning the methods and association efficacy of each intervention during induced hypothermia takes time as does the application of one or many of the interventions (Olson et al., 2008). In a study of 539 nurses in Finland, nonpatient care tasks shown to increase workload included organizing the work to be done, staff resources, and mental stress and disharmony with other units/departments (Rauhala & Fagrstrom, 2007). Gurses, Carayon, and Wall (2009) identified several obstacles for nurses completing their work, including precepting new nurses, family-related issues, poor working environment, poor handoff, poor working relationships, poor resources, inability to find patient charts, and poorly stocked rooms. Transporting patients has also been reported to impede the ability for nurses to care for patients (Nuikka, Paunomen, Hanninen, & Lansimies, 2001). In an analysis of 453 staff nurse tasks on a day shift, transporting patients to tests and procedures consumed 4.4% of the staff nurses time during the morning shift (Nuikka et al., 2001). Transportation was identified as a physically strenuous task that contributed to their perception of heavy workload (Nuikka et al., 2001). Considering Spector’s (1997) definition of workload, as cited earlier, the variance of number of tasks
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
and associated intensity of work for each task and the perceived level of workload from day to day are likely to vary dramatically. This is compounded by the ability of other professionals surrounding the nurse who may or may not be able to provide support to accomplish the work (Jenaro et al., 2011).
Outcomes of Workload Perceived unmanageable workload has been associated with higher levels of complications for patients 24 hours after ICU discharge (Apostopoulou, Georgoudi, Tsaras, & Veldekis, 2006). Specific post-ICU discharge complications examined include pneumonia, bloodstream infection, urinary tract infection, surgical infections, pressure sores, central venous system infection, hydrothorax, arrhythmia, atelectasis, bowel obstruction, renal failure, and pneumothorax (Apostopoulou et al., 2006). Moderate-to-extreme stress was reported by staff to relate to workload (n = 398, 77.8%; Lu, While, & Barriball, 2007). Rauhala et al. (2007) found a statistically significant relationship between workload and sick time (p G .002). Nurses who reported the heaviest workload were 71%Y92% more likely to experience burnout and job dissatisfaction (Rafferty et al., 2007). Workload also had the strongest correlations with work environment satisfaction (r = .53; Lin & Cai, 2009). A recent study revealed that workload was the most influential reason for intent to stay with perception of increased workload decreasing the intent to stay (Mrayyan, 2007). The relationship between workload and job satisfaction is inconsistent (Kovner, Brewer, Wu, Cheng, & Suzuki, 2006). The inconsistent findings related to nurse outcomes of job satisfaction may be secondary to insufficient methods of measuring workload.
Measuring Workload Illness and acuity of the patient was the predominant method found in the literature to measure workload. The Acute Physiology and Chronic Health Evaluation Classification System, designed to measure illness severity and predict mortality in the intensive care setting, was used to evaluate hospital resources and nurse work time for each patient, but it is only one aspect of workload (Kiekkas et al., 2007). Another method to measure workload is the Nottingham Patient Dependency System, which scores nine areas of patient status including respiratory, cardiovascular, neurological, pain control, elimination, psychosocial, hygiene, feeding, and mobility needs (Kiekkas et al., 2007). These clinical severity measures were shown to relate to workload of nurses (Kiekkas et al., 2007). Acuity systems have been proposed as a way to make judgments regarding how much staffing is needed for care (Graf, Millar, Feilteau, Coakley, & Erickson, 2003). However,
&
Number 3
&
June 2015
patient illness and acuity has been found to be insufficient because it does not integrate dimensions like individual patient coping with illness that can make care more complex and thus increase workload (Spence et al., 2006). The Workload Management System for Critical Care Nurses was developed to measure workload in critical care (Cho, Shin, Chio, Jung, & Lee, 2006). The Professional Assessment of Optimal Nursing Care Intensity Level is another method to measure nurse intensity (Frilund & Fagerstrom, 2009). The Workload Intensity Measurement System uses nursing diagnosis to measure workload of nurses (Hoi, Ismail, Ong, & Kang, 2010). The Community Client Need Classification System was developed to measure workload of community nurses (Brady, Byrne, Horan, Macgregor, & Begley, 2008). The Workload Measurement and Reporting System was developed to measure workload in the ambulatory setting (Dickson, Cramer, & Peckham, 2010). The primary challenge with measures, like the ones listed herein, is either projected or retrospective and does not evaluate workload in real time (Lin & Cai, 2009). In addition, identified models may be misspecified because they were not designed with the specificities of neuroscience care units in mind.
Method Considering there were no existing measures of workload developed for a neurosurgical care unit, validity of the measure was dependent on careful identification, articulation, and verification of the dimensions and intensity of each item of work conducted by nurses on this neurosurgical care unit. The Unit Practice Council (UPC) for this neurosurgical care unit was deemed as the source that would be able to provide the desired information. In 2007, the nurses on a neurosurgical inpatient unit initiated a UPC as an outgrowth of a shared governance model. The UPC included staff nurses, nurse support staff, and the unit manager who were the primary drivers of this research project to develop a workload model. Input from all three levels of staff was consistent with the sociotechnical systems framework that proposes that both staff and management should have input into refinement of work operations (Maxwell, Ziegenfuss, & Chisholm, 1993). Collaboratively, the UPC identified 43 elements they perceived contributed to workload of staff nurses on the unit. All four members of the UPC were in 100% agreement that all 43 items were essential to include in the measure if their workload was to be accurately measured. The 43 items were categorized into nine categories, including patient load (eight items); charting (one item); computer system (three items); process of care (one item) ; delegation (two items); admissions,
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E11
E12
Journal of Neuroscience Nursing
discharges, and transfers (six items); total number of supply carts and supply rooms on the unit (one item); and support staff (five items). There were also 16 specific care tasks unique for the neurosurgical care unit. The first eight categories were rank ordered from the most contributory to workload to the least contributory and then numbered 1Y8 (see Figure 1). The 43 items were used to develop a survey for staff nurses and a survey for charge nurses. Staff nurses would be asked to report on each activity each day regarding the amount of time spent, measured in minutes, for each of the 43 items. Charge nurses also completed a survey for staff nurses because charge nurses not only are responsible for the overall operations of the unit but also take a number of patients to care for during their shift of work. In addition to completing a staff nurse survey, charge nurses were also asked to report the cumulative activity specific to the 43 items for the overall unit. For example, a staff nurse would report the time in minutes spent on a ventilator for their patient, whereas the charge nurse would report the number of total vents for the unit. It was proposed by the members of the UPC that overall activity on the unit impacted their work because it determined the level of resources available to them to help if help was needed to carry out patient care. The final dimension of workload that would be measured was time, which has been a neglected component of social science (McGrath & Tschan, 2004). Time for this study is defined operationally as the progression of days within the 30-day period this study took place. Every nurse on the neuroscience unit was asked to respond to the workload measure every day of work, within the 30-day study. According to McGrath and Tschan (2004), social change occurs at microlevel, mesolevel, and macrolevel. Thus, the monitoring of the day of study would be included in the regression equation to understand how time itself related to the perception of workload. This descriptive study used correlations and stepwise regression equations to interpret the data. Stepwise regression was selected for this study because of the current lack of data about satisfaction with workload on a neurosurgical care unit, lack of theory testing, and lack of understanding of independent variables that correlate with satisfaction with workload. Stepwise regression is the most appropriate regression procedure to use when the researcher wants to do model building, not model testing (Tabachnick & Fidell, 2007). Correlates that were statistically significant were entered into the stepwise regression in order of strength of correlation with the strongest correlation being entered first. Diagnostic residuals were used to evaluate if the model fit improved or worsened upon entry of each variable. If the model fit worsened or the explained
variance declined, the variable was excluded from the final equation. An alpha of .05 was used for the level of significance. Power analysis was also conducted for both correlation procedures and stepwise regression. A sample size of 300 survey responses was essential for a power of 0.95 and effect size of 0.1 for the regression equation.
Setting The study took place on a 36-bed neurosurgical unit. The unit has six step-down beds, five ventilator beds, and one close observation room with a 1:4 ratio and utilizes a nursing assistant to monitor the patients for behavioral changes and assist with personal needs such as toileting, feeding, and mobility. The unit is the neurosurgical care unit in a large tertiary care center in New York City.
Sample Participant selection in the study was based on a convenience sample. Inclusion criteria included being a registered nurse (RN) on the unit, identified by the unit manager as regular staff, working part or full time. Exclusion criteria included temporary staff, orientees, or any regular staff on leave of any sort (e.g., family leave, sick leave, etc.) or on vacation during a portion of the study. There were 38 RNs who met the inclusion criteria and were at various levels of clinical development, which included 12 staff nurses, 21 clinical I nurses, 2 clinical II nurses, and 3 clinical III nurses. Each nurse was asked to respond and give their perception of workload and tasks performed during their shift over the 28-day period. The authors felt that a 28-day period would offer a realistic representation of nurses’ workload, reflecting high and low unit census and changes in patient acuity. All eligible nurses were surveyed over the 28-day period. All nurses on this nursing unit worked either the 12-hour day shifts or the 12-hour night shifts. There were 349 surveys sent out over this 28-day period.
Instrumentation The literature review did not reveal an instrument that would adequately represent the variables of workload identified by the study group; thus, an instrument was developed for examining the workload items specific to the unit. The consultant and expert in survey development developed a first draft of the survey that would query all the variables in the workload model. The survey was revised four times until all participants in the research study group reached full agreement that the survey accurately reflected the unit workload. The workload survey consisted of 47 variables and 1 openended question for comments on aspects of workload during the shift not identified in the survey. There
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
&
Number 3
&
June 2015
FIGURE 1 Facets of Workload Identified by Staff on Neurosurgical Care Unit
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E13
E14
Journal of Neuroscience Nursing
were six items for the dependent variable of workload, which is the subscale from the Nursing Work Index developed by Stamps and Piedmonte (1976). The Cronbach’s alpha has consistently been above .80, which was considered adequate for this study. The index developed by Stamps and Piedmonte has been used to measure workload (Lin & Cai, 2009).
Procedures After internal review board approval, surveys were sent electronically to participants’ hospital email addresses. The survey was sent out every day, every shift, to every nurse on the unit from September 8, 2009, to October 5, 2009. Staff were instructed to respond to the survey only on the days worked. Final results were compared with the responses in the data set to confirm that the employee was working that day. Responders could ‘‘save and return’’ to the survey as many times as necessary, depending on time available during the shift to complete the survey. Most nurses choose to complete the survey at the end of their shift when they had time to reflect on their patient assignment, workload, and shift events. Nurses received overtime pay if they stayed after their assigned shift to complete the survey. Staff hypothesized that workload was less on shifts when there was available staff (e.g., low nurseYpatient ratios or sufficient ancillary staff) compared with shifts where staffing was suboptimal. To validate the unit activity for the shift, the charge nurse was also given a survey to report on the number of admissions; transfers in and out; discharges; additional RNs added during the shift like float nurses or ‘‘call backs’’ (e.g., nurses called to work in the middle of the shift); number of support staff; total number of supply carts on the unit; number of patients at the start of the shift; number of available computers; number of ventilator patients; number of total care patients; number of patients with spinal drains; number of patients requiring point-ofcare testing (e.g., urine-specific gravity or blood glucose testing); number of patients receiving blood products; number of patients with neurological deficits; number of confused patients; number of combative patients; number of postoperative neurosurgical patients; number of patients on a urinary catheterization protocol; number of patients who required a 1:1 observation for delirium, agitation, or confusion; and number of patients at high risk for falls who required a 1:4 observation. Charge nurses also had the ability to comment on aspects that added to workload and what eased workload. These items were validated by unit leaders as contributing to overall workload. As the number of these items increased on the unit, the demand on staff resources increased. This added demand increased not only the unit workload but also the individual workload.
Results Response Rate There were 349 surveys sent out over this 28-day period, and 298 surveys were returned, representing an 85.4% response rate. Upon closer examination of the nonresponders, it was noted that there were only a few nurses who did not respond to the daily survey of workload, which accounted for most of the nonresponses. The workload scale revealed a low Chronbach’s alpha of the six-item workload scale, the dependent variable of this study (Chronbach’s alpha = .46). Three items were deleted, and Chronbach’s alpha increased to .71, which was considered acceptable for this study. The three final items inquired about staff nurses’ perception that level of workload facilitated good job performance, their current workload allowed enough time to discuss patient problems with colleagues, and the types of activities in the job were satisfying. Before running regression analysis, the relationship satisfaction with workload had with every independent variable was examined. Only independent variables that had statistically significant correlations with the dependent variable of satisfaction with workload would be entered into the final regression equation. Correlation analysis revealed 6 of 47 variables that related to workload and were statistically significant. Time had the most significant relationship with workload revealing a positive correlation of .41. Percentage of work that could be delegated to support staff but remained undelegated also correlated with workload, but this was found to be a negative relationship (i.e., work that could/ should be delegated but was not). The number of patients, including patients admitted, discharged, or transferred, was found to have a negative correlation (r = j.15). Satisfaction with the amount of paperwork had a positive correlation with workload. Finally, the more patients each nurse had with a neurosurgical diagnosis within their patient load, the more satisfied with workload the nurses felt. Pearson’s correlation value and p value for level of significance for each of the correlates of workload are noted below. b Time (day of the study): r = .408, p G .001 b Percent of work that could have been delegated but was not: r = j.261, p G .001; Spearman’s r = j.273, p G .001 b Number of patients, including ADT: r = j.148, p = .012; Spearman’s r = j.149, p = .012 b Number of isolation patients: r = j.137, p = .021; Spearman’s r = j.127, p = .032 b Satisfied with amount of paperwork: r = .128, p = .029; Spearman’s r = .174, p = .003 b Number of neuro postoperative patients: r = .122, p = .039; Spearman’s r = .116, p = .050
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
All independent variables that were found to have a statistically significant relationship with the dependent variable of satisfaction with workload were entered into a stepwise regression equation. Results revealed that time was the most predictive, explaining 16.2% of the variance of workload, followed by percentage of work that could have been delegated but was not, which explained 3.9% of the variance. The third predictor of workload in this study was the number of isolation patients, which explained 2.8% of the variance. Fourth was the number of patients, including the number of admissions, discharges, and transfers, which predicted 1.4% of the variance of workload. The final variable entered into the equation was number of postoperative neurosurgical patients, which predicted 1.5% of the variance of satisfaction with workload. In total, 25.8% of the variance of satisfaction with workload was explained by the work of the individual staff nurse (see Table 1). This second part of analysis included the examination of workload and satisfaction of workload reported by charge nurses. This study examined the unit-wide activity such as total number of ventilator patients, total number of isolation patients, and any other activity that utilized unit resources that could potentially impact the perception of nurse workload. The charge nurses provided the unit-specific data 91% of the time or 51 of the 56 surveys completed. There were six different nurses who worked as charge nurse on the day shift and nine different nurses who worked as charge nurse on the night shift. The correlates for unit data were entered into a stepwise regression. Forty-four of the 51 shifts where data were reported had adequate data to be included in the regression equation. Data were controlled for day/night shift as well as time. After each of the correlates of unit data were entered into a stepwise regression equation, it was noted that time, which is referred to as ‘‘wave,’’ predicted 42.5% of the variance. The number of nurses predicted 7.7% of the variance. It is obvious that what influenced the
TABLE 1.
&
Number 3
&
June 2015
perception of workload over the 28-day study needs to be better understood. Further inquiry is warranted (see Table 2). The investigators wanted to understand if there were differences in the activity on the unit when different charge nurses worked. It was the experience of the co-investigators that some charge nurses managed the units very efficiently and collaboratively with staff, whereas other charge nurses were not as effective. It was desired to know if the perception of workload of staff nurses differed when comparing shifts for each of the 15 charge nurses, and significant differences were found on the day shift using an analysis of variance procedure, using an alpha of .05. Significant differences were also found on the night shift when comparing charge nurse shifts using the least significant differences procedure. Again, an alpha of .05 was used. The box plot graph (Figure 2) reveals the dummy codes for each of the night-shift charge nurses and the variance of staff nurses’ report of workload when working with each charge nurse. It is noted that, for the night shift, the top two box plots scored over 5.0 on the 1Y7 Likert scale, with higher scores indicating more nurse satisfaction with workload. The statistical significance is noted by the asterisk to the right of the dummy code. The box plot graph (Figure 3) depicts data for nurse workload satisfaction by charge nurse for the day shift. It is noted that nurses reported more satisfaction when working with only one of the five assigned charge nurses. A score greater than 5.0 indicates a high degree of satisfaction. Again, the difference was statistically significant at the .05 level and is noted by an asterisk placed only by the top box plot dummy code. The two final variables that were found to correlate with workload were the total number of RNs working each shift (r = .30, p = .025) and the number of patients who were on 1:1 observation by support staff. These variables had a negative impact on perception of workload (r = j.34, p = .013). This is because of the frustration of taking support staff out of the staff
Regression Equation Workload: Staff Nurses Statistic
Step
Variable
Total R2
R2
Change
Beta
Significance
.402
.162
.162
52.144
.000
1
Time
2
Work not delegated
.449
.201
.039
13.262
.000
3
Number of isolation patients
.479
.230
.028
9.889
.002
4
Number of patients
.493
.243
.014
4.826
.029
5
Number of neuro postoperative patients
.508
.258
.015
5.334
.022
Note. Residual diagnostics were run, and no violations of assumptions were found.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E15
E16
Journal of Neuroscience Nursing
TABLE 2.
Regression Equation Workload: Charge Nurses Statistic
Step
Variable
Total R
2
R
2
Change
Beta
Significance
1
Time
.652
.425
.425
31.060
.000
2
Number of RNs
.709
.502
.077
6.370
.016
Note. Residual diagnostics were run, and no violations of assumptions were found.
count so a nurse attendant can sit with one patient. Further inquiry is warranted in this area. The final unit-level variable that was found to have a positive relationship with workload was the number of RNs on duty. This excluded nurses who were called in to work during the shift. This final variable had an r value of .286. In layman’s terms, this would mean 28.6% of the time that, when the number of RNs increased, satisfaction with workload also increased. That is, however, those two variables alone being considered and not everything else that is factored into the final regression equation. This is the final model that integrated the individual nurse data, unit-level data, and aggregate (charge nurse) data from other studies conducted within New York Presbyterian Hospital. It is feasible to consider that up to 50% or 60% of the variance could be explained for workload on this neurosurgical unit (Figure 4).
Discussion The study revealed a statistically significant difference in staff satisfaction with workload when comparing shifts led by different charge nurses. Charge nurses need to be accomplished in team building, prioritization, communication, and delegation skills. Furthermore, charge nurses need to possess a high degree of
clinical knowledge making them proficient or expert in their specialty. There must be a concerted effort to have a strong charge nurse with insight and situational awareness to assess and anticipate the unit’s activities for each shift. The primary tasks of the RN must be taken into consideration by the charge nurses when making patient assignments. Primary RN tasks within the study institution at the time of this study included catheterization protocol, vented patients, or spinal drains. In addition, every effort must be made for continuity of care, to ensure nurses who are working 2 or 3 days in a row keep their same patient assignment. In turn, charge nurses make every effort to see that neurosurgical patients are admitted to 8HN because this is 8HN’s specialty. The staff undergo lengthy and advanced training to care for the neurosurgical population and take great pride in their specialty. Nurses working in highly specialized areas are recognized by their peers and nursing leadership for their expert clinical knowledge and take immense pride and ownership in their patient population. This substantiates the value of geographic localization for very specialized patient populations. The data revealed that perception of workload decreased when nurses were caring for their specialty population compared with when they were caring for overflow
FIGURE 2 Box Plots Revealing Differences of Workload by Night Charge Nurse
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
&
Number 3
&
June 2015
FIGURE 3 Box Plots Revealing Differences of Workload by Day Charge Nurse
patients from other specialties such as medicine or general surgery. In today’s interdependent healthcare environment, there is a shared responsibility for getting things done. It is no longer reasonable to expect nurses to do it all. Nurses need training and direction in how and what to delegate and to whom. It is also important that nurses understand the breadth, licensure, and certification of each role, including their own. Effective delegation is about sharing workload, with the added bonus of developing skills and responsibility in others. Nurses who learn to effectively delegate not only to help themselves but also to help their patients get better care. Time as a predictor of workload was explained by staff who listened to the study results. First was the impact of study itself. Initially, staff felt that the survey
was frustrating and added to their perception of workload. However, the staff reported they wanted to communicate to the researchers as accurately as possible what was happening within each day they worked. One nurse even called the outside consultant working with this study to inquire how to submit her survey. She had worked for 12 hours and 2 hours of overtime but now wanted to complete the survey. When it was explained to her by the consultant that the survey was voluntary and not mandatory and, thus if too tired, she could forego responding, she replied: ‘‘I had a really bad day and I want to take time to tell you about it in the survey.’’ Despite the frustration of responding to the survey at the end of the day, the staff nurses informed the investigators that, eventually, it became part of the day and quite easy to respond to. This may,
FIGURE 4 Final Model of Workload
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E17
E18
Journal of Neuroscience Nursing
in part, explain why workload improved over time as the staff became used to responding to the survey. The second reason time played an important role in the study is secondary to the survey itself providing insight into how they could work better with staff. After data collection was completed, staff informed study investigators that the survey itself helped them understand that they were not using support staff adequately to delegate nonclinical tasks. The survey itself was not instructive, but the staff reported that the questions made them ponder how they use the support staff. Thus, over the period of the study, they began to use support staff differently by discussing and delegating tasks that support staff could do. The result of the discussion between RNs and support about tasks resulted in more effective use of the support staff role, which lightened the workload of the RN. The findings revealed what staff suspected regarding 1:1 patient observation. Satisfaction rates were lower when support staff were required to 1:1 sit with patients as it takes away from the unit staffing. Workload is a concept that is real yet hard to quantify because of the many variables that impact the design. To further study the variable effects on workload, it is suggested that each of these processes be automatically monitored in real time using a data management/computer interface system. Each time a patient is admitted or transferred, the system would collect data surrounding the event in real time. It would include the number of patients per nurse, number of admits per nurse, and number of transfers on and off the unit. The process could also do patient tracking to determine if a nurse has cared for a patient in the past, whether it be during the patient’s current hospitalization or a past admission. The data management system can also monitor of how much time the nurse spends documenting in the electronic medical record. An automated model would include the following variables: 1. 2. 3. 4. 5. 6. 7. 8. 9.
Number of patients per nurse Number of admits per nurse Number of transfers per nurse on the unit Number of transfers per nurse off the unit Number of discharges per nurse Number of support staff per nurse Number of isolation patients Number of neuro postoperative patients Number of days nurse took care of the same patient, both consecutive and sporadic 10. Documentation by nurse about patient (including notes about patient to other disciplines, number of referrals, number of documented conversations with family, and time charting on each patient).
Staff can continue to propose other variables that they hypothesize to relate to this model. This will facilitate simulation modeling to evaluate how changes in an independent variable impact an outcome variable under study.
Conclusion The workload study has helped this unit by allowing the nurses to identify key components impacting their workday. Since the completion of this study, many factors have been implemented to reduce the variables negatively affecting workload. For example, an additional nurse attendant was hired during the 11 P.M.Y 7 A.M. shift. The hiring of another nurse attendant has allowed nurses to receive more help with their daily tasks and assistance with patients who require total care. The unit also added an ICU technician to the staff. The skill set of ICU technicians allows for delegation of point-of-care testing and 12 lead electrocardiograms. The ICU technician is able to help with many tasks in the step-down unit where the patients are of higher acuity. In addition to increased staff, the unit also added an additional supply cart in the 1:4 observation room and has partnered with the transport department to have more wheelchairs available on site. These changes afford less travel back and forth for supplies and equipment and more time dedicated to patient care. The designation of computers during specific time intervals for nursing has also allowed more time for nurses to document during peak usage hours. The workload study has allowed management and staff to collaborate and bring forth positive unit changes. This study has allowed an opportunity not only to confront and solve workload challenges but also to make improvements leading to patient and staff satisfaction.
References Al-Kandari, F., & Thomas, D. (2008). Perceived adverse patient outcomes correlated to nurses’ workload in medical and surgical wards of selected hospitals in Kuwait. Journal of Clinical Nursing, 18(4), 581Y590. Apostolopoulou, E., Georgoudi, E., Tsaras, K., & Veldekis, D. (2006). Post-ICU complications in critically ill patients: The impact of nursing workload the first day on the ward. ICUs & Nursing Web Journal, AugYOct(27), 6 pp. Brady, A., Byrne, G., Horan, P., Macgregor, C., & Begley, C. (2008). Reliability and validity of the CCNCS: A dependency workload measurement system. Journal of Clinical Nursing, 17(10), 1351Y1360. Cho, Y., Shin, H., Chio, J., Jung, M., & Lee, B. (2006). Development and application of the Workload Management System for Critical Care Nurses (WMSCN) using the Workload Management System for Nurses (WMSN). American Journal of Critical Care, 15(3), 325Y325. Dickson, K. L., Cramer, A. M., & Peckham, C. M. (2010). Nursing workload measurement in ambulatory care. Nursing Economic$, 28(1), 37Y43.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
Volume 47
Frilund, M., & Fagerstrom, L. (2009). Managing the optimal workload by the PAONCIL methodVA challenge for nursing leadership in care of older people. Journal of Nursing Management, 17(4), 426Y434. Graf, C. M., Millar, S., Feilteau, C., Coakley, P. J., & Erickson, J. I. (2003). Patients’ needs for nursing care. Journal of Nursing Administration, 33(2), 76Y81. Gurses, A. P., Carayon, P., & Wall, M. (2009). Impact of performance obstacles on intensive care nurses’ workload, perceived quality and safety of care, and quality of working life. Health Services Research, 44(2 Pt 1), 422Y443. Henderson, J., Willis, E., Walter, B., & Toffoli, L. (2008). Measuring the workload of community mental health nurses: A review of the literature. Contemporary Nurse: A Journal for the Australian Nursing Profession, 29(1), 32Y42. Hoi, S. Y., Ismail, N., Ong, L. C., & Kang, J. (2010). Determining nurse staffing needs: The workload intensity measurement system. Journal of Nursing Management, 18(1), 44Y53. Jenaro, C., Flores, N., Orgaz, M. B., & Cruz, M. (2011). Vigour and dedication in nursing professionals: Towards a better understanding of work engagement. Journal of Advanced Nursing, 67(4), 865Y875. doi:10.1111/j.1365-2648.2010.05526.x Kiekkas, P., Brokalaki, H., Manolis, E., Samios, A., Skartsani, C., & Baltopoulos, G. (2007). Patient severity as an indicator of nursing workload in the intensive care unit. Nursing in Critical Care, 12(1), 34Y41. Kovner, C., Brewer, C., Wu, Y., Cheng, Y., & Suzuki, M. (2006). Factors associated with work satisfaction of registered nurses. Journal of Nursing Scholarship, 38(1), 71Y79. Lin, Y., & Cai, H. (2009). A method for building a real-time cluster-based continuous mental workload scale. Theoretical Issues in Ergonomics Science, 10(6), 531Y543. Lu, H., While, A. E., & Barriball, K. L. (2007). A model of job satisfaction of nurses: A reflection of nurses’ working lives in Mainland China Journal of Advanced Nursing, 58(5), 468Y479. Maxwell, C. I., Ziegenfuss, J. T., & Chisholm, R. F. (1993). Beyond quality improvement teams: Sociotechnical systems theory and self-directed work teams. Quality Management in Health Care, 1(2), 59Y67. McGrath, J. E., & Tschan, F. (2004). Temporal matters in social psychology. Washington DC: American Psychological Association. Mittmann, N., Seung, S. J., Pisterzi, L. F., Isogai, P. K., & Michaels, D. (2008). Nursing workload associated with hospital patient care. Disease Management and Health Outcomes, 16(1), 53Y61.
&
Number 3
&
June 2015
Morris, R., MacNeela, P., Scott, A., Treacy, P., & Hyde, A. (2007). Reconsidering the conceptualization of nursing workload: Literature review. Journal of Advanced Nursing, 57(5), 463Y471. Mrayyan, M. T. (2007). Jordanian nurses’ job satisfaction and intent to stay: Comparing teaching and non-teaching hospitals. Journal of Professional Nursing, 23(3), 125Y136. Nuikka, M., Paunonen, M., Hanninen O., & Lansimies, E. (2001). The nurse’s workload in care situations. Journal of Advanced Nursing, 33(3), 406Y415. Olson, D. M., Kelly, A. P., Washam, N. C., & Thoyre, S. M. (2008). Critical care nurses’ workload estimates for managing patients during induced hypothermia. Nursing in Critical Care, 13(6), 305Y309. Rafferty, A. M., Clarke, S. P., Coles, J., Ball, J., James, P., McKee, M., & Aiken, L. H. (2007). Outcomes of variation in hospital nurse staffing in English hospitals: Cross-sectional analysis of survey data and discharge records. International Journal of Nursing Studies, 44(2), 175Y182. Rauhala, A., & Fagerstrom, L. (2007). Are nurses’ assessments of their workload affected by non-patient factors? An analysis of the RAFAELA system. Journal of Nursing Management, 15(5), 490Y499. Rauhala, A., Kivimaki, M., Fagerstrom, L., Elovainio, M., Virtanen, M., Vahtera, J., I Kinnunen, J. (2007). What degree of work overload is likely to cause increased sickness absenteeism among nurses? Evidence from the RAFAELA patient classification system. Journal of Advanced Nursing, 57(3), 286Y295. Sawatzky-Dickson, D., & Bodnaryk, K. (2009). Validation of a tool to measure neonatal nursing workload. Journal of Nursing Management, 17(1), 84Y91. Spector, P. E. (1997). Job satisfaction: Application, assessment, cause, and consequences. London, UK: Sage. Spence, K., Tarnow-Mordi, W., Duncan, G., Jayasuryia, N., Elliott, J., King, J., & Kite, F. (2006). Measuring nursing workload in neonatal intensive care. Journal of Nursing Management, 14(3), 227Y234. Stamps, P. L., & Piedmonte, E. B. (1986). Nurses and work satisfaction: An index for measurement. Ann Arbor, MI: Health Administration Press Perspectives. Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics. Boston, MA: Pearson. Williams, H., Harris, R., & Turner-Stokes, L. (2009). Work sampling: A quantitative analysis of nursing activity in a neurorehabilitation setting. Journal of Advanced Nursing, 65(10), 2097Y2107.
Copyright © 2015 American Association of Neuroscience Nurses. Unauthorized reproduction of this article is prohibited.
E19
