Introduction

Nurses are essential for safe healthcare delivery, including patient care. In order to support the health of others, nurses need to be well themselves. It is well established that improvements in healthcare workplaces, such as increased staffing, improve patient outcomes [2, 3]. However, assessments of nurses’ health have frequently focused on individual level factors, including areas like diet and exercise [18, 22, 39]. Fewer studies have examined how nurses’ work environments impact upon their health. This study examined the impact of work environment on nurses’ Health Related Quality of Life (HRQOL).

Background

Work environment is known to influence workers’ health in many sectors, and nurses are no exception. Research shows that nurses are at increased risk for musculoskeletal injuries, particularly back injuries [15, 16, 19, 38]. Additionally, nurses have higher injury rates than workers in agriculture, mining, or construction [47]. Problems arising from shift work, such as sleep disturbances, fatigue, and depression can increase nurses’ intentions to leave their jobs [26]. The long hours and physical demands of nursing can cause irregular and abnormal menstruation [29] and difficulty becoming pregnant [20]. A 2015 nurses’ health study found that nurses experience many negative health consequences [8, 12, 20, 34, 45]. While it is evident that nursing is a physically taxing profession, there have been fewer studies exploring the impact of the work environment characteristics on nurses, rather than the work itself. This study assessed the self-reported health of Licensed Practical Nurses (LPN) and the impact of their work environments on their health. Licensed Practical Nurses are a regulated nursing profession in Canada and work in collaboration with members of health care teams [10].

Research questions

The questions addressed in this study were:

  1. 1.

    How do LPNs rate their Health Related Quality of Life?

  2. 2.

    What is the impact of work environment on self-reported Health Related Quality of Life for LPNs, when controlling for resilience and demographic factors?

Methodology

A cross-sectional survey of the population of LPNs in the Canadian province of Alberta was conducted in August 2018. The online survey was designed to take a well-rounded look at the work-life of Alberta LPNs. This article presents the study of the relationship between work environment, measured by the Practice Environment Scale of the Nursing Work Index (PES-NWI) (participation, leadership, resources, and relationships), personal resilience, and health-related quality of life, as measured by the Short Form 36 item Health Survey (SF-36).

Procedure

The survey was distributed by email to the total population of Alberta’s LPNs (n = 15,860). LPNs were informed that their participation was voluntary, anonymous, and that completion of the survey implied consent. The survey was conducted by the Alberta LPN regulatory college and university partners. These data for this article were a secondary use of this original data. After discussion with the Community Ethics Research Board, it was determined that this secondary data analysis study was a low ethical risk and did not require a formal ethics review.

Measures

Health-related quality of life

The SF-36 [48] examines general health, including both physical and emotional well-being. The SF-36 examines 8 dimensions of health, with each dimension is scored from 0–100. In general, higher scores indicate better health and functioning. The 8 dimensions can be collapsed into two summary scores referred to as the Physical Component Summary (PCS) and the Mental Component Summary (MCS). In studies of the general population, internal consistency reliability coefficients ranged from 0.92 to 0.94 for the PCS score and 0.87 to 0.89 for the MCS score [48].

Perceptions of work environment

Nursing practice environment is defined as the “organizational characteristics of a work setting that facilitate or constrain professional nursing practice” [27, p. 178]. The PES-NWI examines nurses’ perceptions of the various practice environments at their current job. The ‘Nurse Participation’ subscale refers to perceived opportunities for nurses to meaningfully participate in workplace activities, including at the broader organizational context (e.g., participate in policy decisions, serve on organizational committees, etc.). ‘Leadership’ refers to nurses’ perceptions of effective management and how well they are supported by management. The ‘Resources’ subscale measures their perception on staffing and resources available to do their work. The ‘Relationships’ subscale measures their perceptions of the quality of their professional relationships. The ‘Nursing Foundations’ quality of care’ subscale was not used in the present study. Lake [27] notes that the ‘Nursing Foundations’ subscale measures facility-level phenomena while the other subscales measure unit or ward-level phenomena. The four PES-NWI sub-scales used in this study assess how nurses feel about their immediate work environment.

The PES-NWI is a 4-point Likert scale. Respondents are asked to indicate the degree to which they feel each item is present in their current work environment. Items are scored from 1 (strongly disagree) to 4 (strongly agree). All the items in a subscale are summed and then the mean is calculated to produce a subscale score. A subscale score greater than 2.5 indicates a more positive perception of the work environment. The Cronbach’s alpha for the PES-NWI subscales have been reported as 0.89 for participation, 0.71 for leadership/manager, 0.77 for resources, and 0.85 for relationships [37].

Resilience

Resilience is defined as “the personal qualities that enable one to thrive in the face of adversity” [13], p. 76). The Connor-Davidson Resilience Scale (CD-RISC) is a widely used measures of resilience. The CD-RISC 10 is used in the present study. The CD-RISC 10 is a Likert scale with each item scored from 0 to 4. A respondent is asked to indicate how much they agree that each statement applied to them over the past month. Possible responses range from ‘not true at all’ (0) to ‘true nearly all the time’ (4). The scores for the 10 items are summed to produce an overall resilience score with the highest possible score of 40. Higher scores indicate greater resilience. The CD-RISC 10 is reported to have strong internal consistency with a Cronbach’s alpha of 0.87 in a study of hospital employees [28], and 0.81 in a study of student nurses [4].

Data analysis

The SF-36 has 8 dimensions examining health where each dimension scored from 0 to 100. Therefore, for intuitive interpretations, scores for the sub-scales were first converted to proportions. This was to align our analysis with studies that have examined HRQoL together with corresponding sub scales, where outcomes are usually bounded between 0 and 1 [14]. Here, individuals with a score of zero would be perceived to have poor health, while those with a score of one would be assumed to be in optimal health. Scores between 0 and 1 have a natural beta distribution [14]. The outcome is linked to covariates in a regression model which has two structures: (i) the mean and (ii) dispersion. The mean structure models the average HRQoL, while the dispersion structure adjusts for non-constance of variability in the data. In our analysis, the beta distribution was used to model each of the following HRQoL outcomes mental composite score, body pain, emotional well-being, general health, and energy fatigue.

In instances where majority in a sample are of optimal health (score of ‘1’), then an extended version of beta distribution (called one—inflated beta) is deemed appropriate to describe these data [14]. One inflated regression partitions the population into two: those with optimal health and those with suboptimal health. By extension, this results into a three-part model: (i) a logistic regression comparing the outcomes of those with optimal health versus those with suboptimal health (HRQoL < 100%), (ii) beta regression model examining HRQoL outcome in those with suboptimal health (HRQoL < 100%), and lastly (iii) the dispersion part. The one—inflated beta distribution was used to model each of the following HRQoL outcomes: role physical, social functioning, physical functioning, role emotional and physical composite score. For both models, logit link function was preferred to use as this would allow for the interpretation of parameter estimates in terms of odds ratios [14]. To assess model, we compared observed and fitted values for each model (see Additional file 1).

In both beta and one inflated regression model, the mean structure of model was selected as follows: first, we explored if any non-linearity existed between the HRQoL subscale outcomes and each of the continuous covariates, i.e., leadership, resource, relationship, resilience, and participant age. This was mainly examining whether higher order polynomials would be suitable (see Additional file 1: figure X & Y). The following variables leadership, resource, relationship, resilience, gender, area of work (acute, continuing, community and primary care), years of practice (0–5,6–15 and 16 plus years), employment status (full time, part time and casual) and participant age were used as predictors to examine the association with each outcome variable.

Secondly, for each outcome, given the relationship visualized (see Additional file 1: figure X & Y) linear, quadratic, and cubic terms were explored. The models were compared using AICs and loglikelihood ratio test. If the difference in AIC values was less than 1%, we proceeded with linear relationship with the outcome. In this analysis there was no variable selection as all covariates were predetermined bested on literature review and expert opinion and were not removed from the model if not found to be significant.

All the analysis a performed using R package ‘gamlss’ [43] and betareg [14].

Results

A total of 4,425 LPNs participated in the study. The average age of respondents was 39 years, most self-identified as female (92%) female, had full/part time employment (76%), worked primary in acute and continued care (63%) and most had work experience of 6 + years (51.5%). Demographic data are presented in Table 1.

Table 1 Socio-demographic characteristics and descriptive variables for the survey
Full size table

The responses to the SF-36 varied greatly across each subscale. Figure 1 illustrates the distribution of responses, with many subscales demonstrating a high degree of skewedness.

Fig. 1
figure 1figure 1

HRQoL response distribution

Full size image

The regression estimates from both beta and one inflated regression are shown in Tables 2 and 3. In both models, resources, relationships, resilience, gender, and health status were all significantly associated with increased odds of mental component summary. Overall, resilience and resource are associated with higher odds of both physical and mental component summary for nurses. Employment status and relationships predicted increased mental functioning in nurses. Furthermore, resilience and participant gender were associated with increased physical component summaries for LPNs with optimal health.

Table 2 Predictors of mental component summary using a beta regression model
Full size table
Table 3 Predictors of physical component summary using a one inflated beta regression model
Full size table

Discussion

In this study, we assessed how LPNs rate their HRQoL, and what impact work environment had upon their HRQoL. We had a high proportion of female participants. This was expected, as nursing populations generally include > 90% female employees [49].

Overall, LPNs in this study (mean age: 40) rated their health as being lower than Canadian norms for people 35–44 [23] across every SF-36 subscale. However, this sample reported higher mean scores than some published examples of nurses’ SF-36 scores [6, 11, 42], and similar scores to those of Omrani and Talebi [36]. These results indicate that LPNs report their health as lower than the general population, but higher than other groups of nurses. Self-reported health is a key factor in healthcare worker retention, with eldercare employees with lower psychological wellbeing scores leaving their jobs more often than those with higher scores [21]. Nurses reporting increased health problems also reported increased rates of turnover [26].

LPNs reported that their workplace resources and employment status significantly impacted their HRQoL. It is well known that improvements in healthcare work environments improve patient outcomes [1, 9]. This study illustrates that improvements in the work environment can also impact nurse health outcomes. Quality work environments are associated with lower levels of nurse burnout [3]. The results of this study add that sufficient equipment, staffing, and other resources is positively related to all the dimensions of LPN health. This result could indicate that heaver work loads with fewer resources could be detrimental to the health of nurses, thereby putting patients at risk. Improvements in work environment will likely result in improved nurses’ health as well.

It is also important to note that work environments make a difference to nurses’ health, even after controlling for resilience. There have been a plethora of studies on nurses’ resilience [5, 7, 17, 24, 25, 30,31,32,33, 40, 41, 46]. These studies have been criticized for placing more emphasis on the individual, rather than considering nursing workforce issues [35, 44]. While personal resilience is a predictor of LPNs’ HRQoL, resources such as staffing, and equipment have a larger effect. The results of this study highlight the need to focus the work environment, rather than individual factors.

Implications for practice

The results of this study indicate that a key area for intervention is resource adequacy. LPNs reported that having enough staff, equipment, and resources improved their self reported health. This result adds to the evidence supporting suitable nursing staffing level as being essential for nurses’ health, as well as that of patients. Additionally, this result indicates that nurses’ health is impacted by their work environments, not only their own actions. While interventions like exercise and a healthy diet are important, these individual level interventions would fail to address the quality of the work environment. Nurse leaders can be encouraged to secure safe staffing to promote the health of nursing staff at a population level.

Limitations

Causal inferences cannot be made due to the cross-sectional design of the study. Nevertheless, the findings provide valuable insights into a part of the nursing workforce for which there is limited research. It is acknowledged that other factors like chronic illness have accounted for larger amounts of the variance in other studies. It was beyond the scope of this study to include questions in this area, and it is an acknowledged limitation.

Conclusion

In conclusion, work environment could play a role in promoting a healthy nursing workforce. With the potential for a global nursing shortage, and negative impacts on nurses during the COVID-19 pandemic, it will be imperative that health care organizations address the health of their workforce. This study identifies areas of the work environment that can have a significant effect on a nurse’s physical and emotional wellbeing. An organization can promote nurse health by ensuring their nursing staff can participate in decision making, have adequate staffing and equipment to perform their work, and encouraging positive team collaborations.