Registered nurses' self-rated research utilization in relation to their work climate: Using cluster analysis to search for patterns

1.1 Research utilization

One aspect of evidence-based nursing is the use of research in clinical nursing (Hörberg et al., 2011; Scott & McSherry, 2009). It is important that RNs have a reflective and active approach to research in order to achieve evidence-based nursing, and that they engage in research utilization (RU) (Squires et al., 2011). The body of research on how nursing should be performed is growing. It is a time-consuming and sometimes difficult task to search for, read and evaluate research and its relevance for the daily clinical work performed by RNs.

Research utilization in nursing can be understood as using recent research that is relevant for nursing to the benefit of the patient (Estabrooks, 1999). It can be conceptualized as instrumental, conceptual and persuasive (Estabrooks et al., 2011). With instrumental RU, the research findings are used in a specific and concrete way; implementation of guidelines is an example of this. Conceptual RU is a more general use of research, leading to new insights and use of concepts or theories in a less specific manner. The term ‘persuasive RU’, finally, refers to the use of research findings to legitimate, justify, or mobilize support for actions or decisions (Estabrooks et al., 2011). These three types of RU have been previously studied in a Swedish context as part of the national Longitudinal Analysis of Nursing Education (LANE) project (Rudman et al., 2010). Its findings showed that newly graduated nurses' self-reported RU was relatively low during the first 3 years post-graduation (Forsman et al., 2009; Forsman et al., 2010).

1.2 Work climate

‘Work climate’ is a broad term that needs clarification in the context of nursing. In this study, the concept of a creative work climate, as presented by Ekvall (1996), comprises attitudes, feelings and behaviours in an organization, in a nested way. (Ekvall's term ‘the creative work climate’ (1996) is herein referred to, simply, as ‘the work climate’.)

Work climate entails the interaction between the members of an organization and can be seen as a part of the organization and the organizational climate (Ekvall, 1996), where the organization also includes, for example, money, the staff and guidelines that influence processes and the official lead of the organization (Ekvall, 1996). Regarding organizational climate, it may, for example, be asked: Is there a certain degree of openness and trust, together with commitment and motivation? Other parameters, besides trust/openness, that can have an impact on the work climate, according to Ekvall (1996), are challenge, freedom, idea support, dynamism/liveness, playfulness/humour, debate, conflicts, risk taking and idea time. When measured, these different dimensions are parameters of how the creative work climate can be evaluated (Ekvall, 1996). To this end, Ekvall (1996) developed the Creative Climate Questionnaire (CCQ) instrument, which has been used in previous studies in the health care sector (Carlfjord et al., 2010; Söderlund et al., 2014).

Work climate is an important factor in work satisfaction for RNs (Caricati et al., 2014). Work climate has, among other things, been proposed to have an impact on nurses' work satisfaction and commitment (Caricati et al., 2014), which in turn is linked to the quality of nursing and patient outcome and satisfaction. The work climate can either support or undermine the feeling of being able to provide care of good quality, including being able to work evidence-based (Melnyk et al., 2010; Morténius et al., 2013; Rytterstrom et al., 2009).

The possibility to learn and improve knowledge at work is important for nurses to remain in work (Ahlstedt et al., 2019). Previous qualitative research has indicated that there is a connection between leadership, work climate and the use of research among RNs (Jansson & Forsberg, 2016; Karlberg Traav et al., 2018), which warrants further investigation. Other contextual structures, such as nursing research culture, have been studied and the use of research has been found to be a significant attribute to staying in work (Berthelsen & Hølge-Hazelton, 2016). Therefore, it is of importance to investigate how RNs report on the association between the work climate, their use of research and their possibilities to work in an evidence-based way.

2.1 Design

This cross-sectional survey study was conducted from March to June 2015. The first part of the survey measured how RNs perceived the work climate in the ward where they worked. The items in the second part of the survey were designed to measure three types of RU among RNs: instrumental, conceptual and persuasive.

2.2 Setting, participants and data collection

The study was performed in Sweden at a university hospital with approximately 550 beds. The eligible participants were RNs working in a clinical setting at the university hospital. To qualify for inclusion, the RNs had to have an active email address at work, as the questionnaire was distributed by email. Nurses' email addresses were obtained from the Human Resources Department at the hospital. A link to the web-based survey was distributed to the RNs' email addresses at work with the option to resend the link to a private email address if preferred. Two reminders after 2 and 4 weeks were sent out via email to all registered nurses and a paper questionnaire was thereafter distributed a further 2 weeks later through the internal mail system of the hospital. The paper questionnaires were sent to the first line nurse manager at the different wards, as a final reminder to distribute to RNs who had not responded.

2.3 Instruments

To measure the work climate, the CCQ developed by Ekvall (1996) was used. The purpose of the CCQ is to describe the climate in an organization as either innovative or stagnated, through 10 dimensions considered to either underpin or undermine a creative work climate: The dimension challenge is understood as involvement and commitment to the organization's goals; while freedom is the feeling of independence in decisions regarding the work; idea support is the experience that space is given for new ideas and suggestions; and trust/openness is a trustful, honest, and open climate in the discussion of different opinions. Further, a workplace has: dynamism/liveness if it can be described as eventful and dynamic; playfulness/humour, if the atmosphere is relaxed; debate, if there is space for different opinions, together with respect for knowledge and experience; conflicts, if there is undesirable tension and a poor climate for handling conflicts; risk taking, if decisions, uncertainties, and initiatives are handled with confidence; and, finally, idea time, if time and space are given for implementing and elaborating ideas (Ekvall, 1996). For each dimension, five different statements are given, adding up to a total of 50 statements, and the respondent is required to answer these on a four-point Likert-type scale, from 0 = ‘not at all’, to 3 = ‘to a high degree’. The mean value for each dimension is calculated and a high value indicates a creative climate at the workplace except for the conflicts dimension, where a high value indicates the opposite. The CCQ questionnaire has been validated and tested for reliability in different settings, including health care organizations (Carlfjord et al., 2010; Ekvall, 1996; Söderlund et al., 2014).

In addition to the CCQ, three single-item measures of RU, designed for the LANE project (Forsman et al., 2012) and originally developed by Estabrooks (1999), were used. As previously mentioned, RU in nursing means using recent nursing research to the benefit of the patient. Estabrooks et al. (2011) identified three types of RU in nursing, namely, instrumental, conceptual, and persuasive RU (for a definition, see ‘Research utilization’ above). The single-item measures about the three different types of RU contained definitions as well as examples of instrumental, conceptual, and persuasive RU. The respondents answered how often they had used RU during their work shifts in the last 4 weeks, on a five-point scale where 1 = ‘never’, 2 = ‘occasionally on a work shift’, 3 = ‘during approximately half of the work shifts’, 4 = ‘on more than half of the work shifts’, and 5 = ‘during almost all work shifts’. Respondents put 6 to indicate ‘I don't know’.

2.4 Analysis

Data are presented as number and percentage (%), with means and standard deviation or medians and interquartile range (IQR) indicated. Differences between groups were analysed with the chi-square test for nominal data, Student's t test or one-way analysis of variance (ANOVA) for continuous data, and the Mann–Whitney U test or Kruskal–Wallis test for ordinal data and continuous data that were non-normally distributed. The significance level was set to 0.05 and multiple significance tests were conducted, with Bonferroni correction applied. The analyses were performed using IBM SPSS Statistics version 24 (IBM Corp., Armonk, NY, USA).

A cluster analysis was performed to study different RU profiles (clusters) among the RNs. Each cluster consisted of a group of RNs with a similar individual response profile across all three types of RU. Calculations of clusters were made in SPSS using Ward's hierarchical agglomerative method, including only respondents with responses on the three RU variables. Single linkage was used to identify outliers (n = 2), resulting in a cluster sample of 453 nurses. The decision on the number of clusters was guided by the agglomeration schedule and a scree plot showing the change in the distance coefficient. Practical considerations concerning meaningfulness and interpretability of the cluster solutions also guided the decision on the number of clusters (Mooi & Sarstedt, 2011).

To assess the stability of the cluster solution, K-means analysis was conducted using cluster centroids from Ward's method as seed points. Furthermore, cases in the dataset were randomly ordered and the cluster solution was repeated three times on the rearranged data (Mooi & Sarstedt, 2011).

2.5 Ethical Considerations

Ethical permission for the study was granted by the Regional Ethical Review Board in Uppsala (Reg. No. 2015/049). Participants were given written information about the study. Voluntary participation was emphasized, as was the possibility to withdraw from the study at any time without having to give a reason. Participants were assured of confidentiality.

4.1 Study limitations and strengths

The results of this study should be viewed in the context of the low response rate, which indicates either low interest in the questions asked, or too little time to participate, or indeed that we did not reach all potential respondents. The age and gender distribution, however, is in agreement with that among clinical RNs working at the university hospital at the time the study was completed (personal communication with the Human Resources Department, January 2018). Regarding gender, the distribution is in line with the national statistics for RNs in the health care sector (The National Board of Health and Welfare, 2018). The decision to use a web-based survey was made to avoid socially desirable bias.

The variable- and pattern-oriented approaches used in this study complemented each other well and provided a broader illustration of data, since data based on each RU variable was reflected against the picture of research users and their profiles, across all three types of RU. Stability tests of the cluster solution showed relatively high agreement. Respondents in the ‘low RU’ cluster were stable over all three repeated tests, which completely supports the next step of the analyses, that is, the division into two RU groups, of low research users versus the rest of the sample. The cluster profiles identified in the present study were similar to those identified in previous studies using the same RU measures among RNs (Forsman et al., 2009, 2010), which supports the cluster solution and the evidence base for the RU cluster profiles that seem to occur naturally.