Diffusion of Innovations-Informed Knowledge Translation Strategy to Implement Optimal Safe Nursing Workforce Policy in Practice.

AIM: This study aims to develop a free, limited-edition workshop as an effective knowledge translation strategy to enhance nurse leader-perceived self-efficacy for competence using Park's Sweet Spot Theory and to evaluate its effectiveness over time.
METHOD: This is a study showing the process of developing a study protocol and its details.
RESULTS: A 2-day workshop was developed for innovators and early adopters among nurse leaders with a macro-level influence based on Rogers's diffusion of innovations theory, which consists of an introduction of Park's Sweet Spot Theory, hands-on experience, a summary session, and a presentation of a certificate of completion. The workshop will be held at the University of Alberta Faculty of Nursing, using the "enabling blends" mode. A hybrid design of comparative effectiveness research and analysis of change will be utilized to assess nurse leader-perceived self-efficacy.
CONCLUSION: This protocol is significant as the first step in providing scientific rationales on how to effectively implement new knowledge-optimal safe nurse staffing levels derived from Park's Sweet Spot Theory-into the right (safe yet efficient) nursing workforce policy-making to alleviate global nursing shortages.

Introduction

Coupled with a rapidly aging population, a low global birth rate is cutting the available workforce worldwide (Szabo et al., 2020). This is very likely to result in insufficient and unsafe nursing care for the elderly as well as climbing welfare costs in the future (Aghakhani & Park, 2019). Even more concerning is that its socioeconomic impact would be more than just undermined economic growth potential; health disparities represented by conflicts, inequity, and inequality among generations and social classes would further worsen (Vollset et al., 2020). Such problems are not limited by borders, and their effects may be experienced worldwide in the near future (Vollset et al., 2020). To make matters worse, the unprecedented coronavirus disease 2019 (COVID-19) pandemic has limited the birthrate and accordingly hastened population aging and decline across the globe, both of which are highly likely to bring out serious policy concerns in terms of the social and economic fallouts in the near future (Aassve et al., 2020).

One of those fallouts is already occurring; the global nursing shortage ( Park, 2017 , 2018a,b) is a worldwide problem that has only intensified due to the COVID-19 pandemic. There is a wealth of research over the past 40 years stating that more (or higher educated) nurses lead to better patient outcomes ( Park, 2018 a,b). However, its implementation in practice leaves a lot to be desired. One of the critical reasons for such a bottleneck maybe the failure to fully address contextual factors in the nursing workforce policy-making, so-called “decision-making context,” while disproportionately focusing on “fragmentary evidence-based decision-making” (which is usually based on “more nurses and better patient outcomes”) (Dobrow et al., 2004; Park, 2018b ).

Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Estimation Theory (Copyright © 2016 Park, Claire Su-Yeon. All Rights Reserved. https://onlinelibrary.wiley.com/doi/abs/10.1111/jan.13284) may help solve the riddle by providing optimal safe staffing levels fitted to each practice setting, satisfying patients, stakeholders, and nurses. Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Estimation Theory was developed through an interdisciplinary consilience-driven theory synthesis among nursing science, micro-/mathematical economics, health economics, and operations research and is currently being extended to computer science and decision psychology. Its philosophical stance is rooted in moderate realism linking post-positivism and contextualism (Fraser & Strang, 2004). Such distinctive qualities lead to identify clear threshold values of the optimal safe nurse staffing levels—specifically, the Optimized Nurse Staffing (Sweet Spot) and the Optimum Nurse Staffing Zone—maximizing “patient outcomes relative to attendant costs” (i.e., patient-perceived value) per each practice setting ( Park, 2017 , 2018a,b). The leverage points function as evidence-based, informed shared decision-making rationales on the optimal safe staffing levels satisfying all interested parties—patients, nurses, and healthcare organizations (or stakeholders)—in nursing workforce policy/decision-making fitted to each practice setting ( Park, 2017 , 2018a,b). Thereby, a valuable balance of efficiency and effectiveness in the nursing workforce can be achieved, contributing to lowering the bill of the next generations’ care, plus lowering the tax burden for the elderly by eliminating unnecessary costs while securing a reasonable level of quality of care ( Park, 2017 ; Aghakhani & Park, 2019).

However, this solution can only be navigated by an interdisciplinary consilience-driven multi-stage procedure that has never been addressed in nursing research until now ( Park, 2017 , 2018b), specifically, nursing workforce science (data collection and management based on the Conceptual–Theoretical–Empirical linkages of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory), data science (single imputation by votingForest and missForest) (Choi, 2019); health economics (cost-benefit analysis), statistics and operations research (a theory-driven, data integrated deterministic approach: statistics and/or optimization), computer science (a theory-integrated, data-driven probabilistic approach: change detection techniques and Bayesian neural network), and statistics (a hybrid of results obtained from the two different approaches: meta-analysis). Nurse leaders may thus perceive the solution as being too complex, which may hinder the translation into nursing practice.

A workshop-format training program may be effective in addressing this issue because it can promote the understanding of the contents and facilitate knowledge translation (KT) into practice. The workshop-format training program functions as a system that is composed of three parts: (a) a mini lesson, (b) a student-centered work portion, and (c) a debriefing to enhance meta-cognition of the lesson (Bennett, 2007). Each part is “orchestrated with purposeful reasons in a purposeful manner in order to ‘serve a common purpose’” (Bennett, 2007, p. 14). Such characteristics bring out participants’ increased engagement in learning, which leads a positive attitude toward the practice that can greatly affect the success of this particular education (Friedrich, 2019).

However, Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory currently resides in the early stages of the innovation lifecycle according to Rogers’s (2003) diffusion of innovations theory. Furthermore, very little is known about the effects of the workshop-format training program for Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory-integrated KT in nursing workforce policy-making. To overcome (or minimize) resistance to innovation that early or late majority may feel (Roger, 1995) and get closer to the scientific veracity of the workshop-format training program, falsifiability needs to be obtained (LeBel et al., 2017).

This protocol thus aims to (1) develop the workshop-format training program to enhance nurse leaders’ perceived level of self-efficacy for competence in the use of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting) and (2) evaluate its effectiveness over time regarding its utility.

Primary and Secondary Outcomes

This paper limits its scope to the nurse leaders’ perceived level of self-efficacy as the primary, proximal outcome of the study. The secondary, distal outcome of the study is Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory-driven safer staffing levels in practice, which requires policy analysis with a greater sharing of research resources and is out of the scope of this paper.

Methods

Study Design

The study protocol consists of two steps: (1) development of the workshop-style KT intervention based on Rogers’s (2003) diffusion of innovations theory and (2) evaluation of the KT intervention using a hybrid design of comparative effectiveness research (i.e., a single-blind, randomized cross-over study design) and analysis of change (i.e., a growth curve modeling [GCM] and an intra-individual variability index [IVI]).

Stage 1 Development of Theory-Driven KT Intervention

A free, limited-edition workshop-format training program for nurse leaders with a macro-level influence—that is, nurse executives, nursing directors, or nurse presidents/CEOs—based on Rogers’s (2003) diffusion of innovations theory (Figure 1) will be developed. A limited-edition product will deliver participants a conspicuous message of the scarcity of the workshop (limited time and quantity), enhancing its perceived value and evaluation of the training (Jang et al., 2015). Additionally, a partnership with such influential nurse leaders will enable them to generate a reinvention process (Roger, 1995), leading to a continuous virtuous cycle for better KT into practice (Valente & Vega Yon, 2020). The target audience will additionally be focused on innovators and early adopters among nurse leaders, characterized by their willingness to accept innovation and who thus have an advantage over their peers (Roger, 1995; Valente & Vega Yon, 2020). The promotion strategies for the workshop-format training program will be separately developed following Rogers’s (2003) five stages of adoption (Figure 2).

Figure 1.

Workshop Configuration Based on Rogers’s (2003) Theory of the Diffusion of Innovations.

Figure 2.

Promotion Strategies Using Rogers’s (2003) Five Stages of Adoption.

Specifically, the workshop-format training program will comprise: (1) an explanation of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory and steps to identify the optimal safe staffing levels (being matched with “a mini lesson”) (Bennett, 2007), (2) hands-on experience (being matched with “a student-centered work portion”) (Bennett, 2007), (3) a summary session (being matched with “a debriefing to enhance meta-cognition of the lesson”) (Bennett, 2007), and (4) a presentation of certificate of completion, considering that early adopters prefer to be seen as leaders and such social prestige is one of their biggest drivers (Roger, 1995; Figure 1). The main themes of the workshop-format training program are described in Table 1.

Table 1.

Main Themes for the Workshop-Format Training Program, Leading to Enhance Nurse Leader-Perceived Level of Self-Efficacy for Competence in Use of Park’s (2017a) Optimized Nurse Staffing (Sweet Spot) Theory

Contents	Step 1	Step 2	Step 3	Step 4
Mini Lesson	Hands-on Experience	Summary Session	Certification
I. Introduction of Park’s (2017a) Optimized Nurse Staffing (Sweet Spot) theory and procedures to identify the optimal safe staffing levels per each practice setting	√	√	√
I. CTE linkages		√	√
With a sample data or a participant’s own data
A. Single imputation		√	√
B. Cost-benefit analysis		√	√
C. Statistics, optimization, and sensitivity analysis		√	√
D. Change detection technique, Bayesian neural network, and eXplainable AI (XAI): SHapley additive explanation (SHAP)		√	√
E. Synthesizing results obtained from the two different approaches: Meta-analysis		√	√

Note: The educational contents and their order are based on the Park’s Optimized Nurse Staffing (Sweet Spot) Theory Extended: Copyright © 2021 Park, Claire Su-Yeon. All Rights Reserved. (private to the public). The copyright holder of the Park’s Optimized Nurse Staffing (Sweet Spot) Theory Extended and the first author of this article are one and the same. This derivative work was performed with the copyright holder’s prior written permission. Anyone may share and adapt the material of Table 1 for only non-commercial purposes by giving the appropriate credit to the original work. However, the use of the original contents, illustrations, or ideas (including an order of analysis procedure) in “Park’s Optimized Nurse Staffing (Sweet Spot) Estimation Theory (Extended)” to someone’s work, either in whole or in part, must require prior written permission from the copyright holder. This is based on the copyright guideline that prior written permission is required from both the copyright holder for the original work and the copyright holder for its derivative work(s), in a case that a third party would like to reuse the derivative work(s) (Choi, 2014, p. 71). For inquiries, please contact the copyright holder: clairesuyeonpark@gmail.com.

The mini lesson will be delivered online based on an “enabling blends” mode developed from three types of blended learning systems. The “enabling blends” mode refers to an educational modality that provides the same experience as face-to-face learning but through more user-friendly channels (Lieser & Taff, 2013). Thereby, the “enabling blends” mode-based free workshop will help mitigate participant-perceived risk and uncertainty about a new practice through online modality, creating easier access, better convenience, more flexibility, and a cost-free delivery of educational content (Lieser & Taff, 2013; Roger, 1995). The hands-on experience will also help this demographic adopt the new content without aversion by guiding them to try it out (trialability) using the basic level example (simplicity and easy use) and check the results with the naked eye (observable results) (Roger, 1995). Lastly, the summary session and the presentation of a certificate of completion will help the target group feel a sense of accomplishment in the task and lead them to disseminate their positive experience with the workshop through peer–peer conversations and peer networks (Roger, 1995). Thereby, the workshop will meet all of the core characteristics that cause innovations to spread (Roger, 1995).

Workshops originally aim to simultaneously promote the acquisition of knowledge and skills and changes in attitudes and behaviors but are characterized by simple, flexible, and timely application (Grossman & Salas, 2011). Thus, (a) limited time investments (usually a half to two days) from both participants and organizations and (b) small groups of active participants (typically less than 20) are common (Grossman & Salas, 2011). However, in contrast with such one-off events, longitudinal interventions may be more effective in achieving behavioral change (Grossman & Salas, 2011). In particular, when attitude changes or new pedagogical approaches are involved, such as the study protocol, more time and attention are required for group dynamics to take effect (Grossman & Salas, 2011). Nonetheless, considering that the study protocol is the first initiation to develop and validate a new workshop-format training program, the duration of the workshop will be set up as 2 days.

Based on the above conditions, the workshop-format training program can be considered as one of the effective KT strategies, particularly because the facilitation effect of the workshop is an active ingredient of KT that promotes successful implementation of evidence into practice (Kitson & Harvey, 2016). It can further address the causes of poor performance—that is, “a lack of shared responsibility for outcomes, lack of cooperation and collaboration, and limited understanding of what works”—and facilitate effecting change and improvement for collaborative public health action, such as “analyzing information, establishing a vision and mission, using strategic and action plans, developing effective leadership, documenting progress and using feedback, and making outcomes matter” (Fawcett et al., 2010).

Stage 2. Evaluation of the KT Intervention

A head-to-head comparison of the workshop-format training program versus the mini lesson (online education only) will be conducted, using a single-blind, randomized cross-over study design, at the University of Alberta Faculty of Nursing, Edmonton, Canada (Figure 3). The primary aim is to determine the utility of the workshop over time. Its specific evaluation indicator is nurse leaders’ perceived level of self-efficacy for competence in use of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting).

Figure 3.

Study Design.

Eligibility Criteria

The nurse leaders will include, but will not be limited to, nurse executives, nursing directors, or senior administrators working in Edmonton, Canada. The first strategy to recruit this target group will be media promotion on the popular webpages or social network services (Roger, 1995), such as the Twitter, Facebook, or Instagram of the Canadian Nurses Association, the Canadian Association of Schools of Nursing, and nursing (leadership) conferences which will be held in Edmonton, Canada. The second strategy to recruit the target demographic will be individually customized approaches (Roger, 1995), including, but not be limited to, a personalized mail, email, or phone call.

Randomization

After screening, the study participants will be randomized 1:1 to one of two educational intervention arms: the mini lesson (online education only) to the workshop (A arm) and the workshop to the mini lesson (online education only) (B arm) (Figure 3). The random allocation will be performed by independent researchers, according to a randomization table using a random permuted block of at least 80 participants. However, the allocation will not be concealed from the researchers enrolling, assessing, and educating the study participants until the end of the study. This is because, due to the nature of this study, researchers can sufficiently predict which arm each participant belongs to.

Interphase, the interval between the two stages will be 3 months (Figure 3). Since the period was verified to be an effective duration for the retention of educational effects, it is supported as a valid washout period between the two stages of the study (Ahn et al., 2017).

Sample

The effective sample size for the comparative effectiveness research design is relatively small (Lee et al., 2014; Portney, 2020). However, at least a total of 30 cases are required for GCM (Portney, 2020), which is in line with the Central Limit Theorem (CLT) that states sample sizes equal to or greater than 30 are considered sufficient for the CLT to hold (Islam, 2018). Nonetheless, to enroll at least more than 75 participants (more than 30 for each arm, with a total of 60 over) will be pursued to detect a significant difference between arms, expecting a dropout rate of up to 20%.

Classification of Events

The specifics of the workshop-format training program versus the mini lesson (online education only) have been discussed in detail above (see Stage 1: Development of Theory-Driven KT Intervention).

Data Collection

A total of four data collections will be performed immediately before and immediately after the workshop and the mini lesson (online education only). A consistent application of Scholz and colleagues’ (2002) General Perceived Self-Efficacy Scale (GPSS) will be administered throughout the study, increasing the internal validity of the study. To obtain uncontaminated data from the Hawthorne effect and the John Henry effect that may threaten the external validity of the study findings, a research assistant who will not participate in the workshop as an instructor or an operational staff will perform the data collection (Portney, 2020).

Statistical Analyses

Outcome Measure

The GPSS validated by Scholz and colleagues (2002) will be administered during the study period. It consists of 10 questions with a four-point Likert scale: 1 = Not at all true, 2 = Hardly true, 3 = Moderately true, 4 = Exactly true. Sum scores range from 10 to 40 accordingly (Scholz et al., 2002). A higher score indicates a better-perceived level of self-efficacy (Scholz et al., 2002). Cronbach’s alpha coefficient for GPSS ranged from .75 (Indian people) to .91 (Japanese) (Scholz et al., 2002), supporting satisfactory reliability of the instrument. Its construct validity was also verified to be satisfactory by Scholz and colleagues’ (2002) rigorous psychometric testing in 25 countries.

Data Management

Data cleansing, multiple imputation for handling missing data, and data analyses will be performed using R software program ver. 4.0.2 and IBM Statistical Package for the Social Sciences Statistics 26 (IBM SPSS Corp., Armonk, NY, USA) for Windows. Multivariate imputation via chained equation package of R software program ver. 4.0.2 will be applied for multiple imputation, specifically, mice.impute.2l.pan function for level-1 data and mice.impute.2l.pmm function for level-2 data.

Task I

The mixed-method analysis of covariance (ANCOVA) to determine the significance of the potential carry-out effect between two arms will be performed (Tabachnick & Fidell, 2013). If significant, doubly multivariate repeated ANCOVA (RMANCOVA) and if not significant, the doubly multivariate repeated analysis of variance (RMANOVA) will be conducted (Tabachnick & Fidell, 2013). The variable of time-order and the initial values will be controlled as covariates (Tabachnick & Fidell, 2013).

Task II

First, GCM will be executed. Second, the generation of a detrended standard deviation for IVI will be performed. Finally, Pearson’s correlation analysis between Person-Level Variables and IVIs will be conducted. GCM and IVI will utilize the Hierarchical Linear Modeling, adapted as the Mixed-Model Modeling approach, to model intercept, linear, and quadratic growth overall in the estimation of nurse leaders’ perceived level of self-efficacy for competence in the use of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting).

Ethical Consideration

An official approval from the Institutional Review Board of the University of Alberta will be obtained to ensure that the research is done in consideration of ethical principles of voluntariness, confidentiality, beneficence, and non-maleficence.

Reporting

Study findings will be articulated following the “CONSORT 2010 statement: extension to randomized crossover trials” (Dwan et al., 2019) to ensure scientific integrity, rigor, and clarity of the study.

Discussion

Considering that “evidence is a multidimensional construct embedded within innovation” (Kitson & Harvey, 2016) and it may take longer than 1 year for policy change to bring out behavioral changes in an individual (Saunders et al., 2019), in order to ameliorate global nursing shortages in a timely manner, effective KT strategies are urgently needed to induce stakeholders to increase nursing workforce without innovation resistance (Sochalski & Weiner, 2011). In this regard, this protocol has significant implications as the first initiation in providing scientific rationales on how to implement new knowledge into nursing workforce policy-making beyond stakeholders’ innovation resistance more effectively. This protocol is particularly important because it is based on the two well-developed theories—Rogers’s (2003) diffusion of innovations theory and Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory. “There is nothing so practical as a good theory for implementation” (Lynch et al., 2018).

Noteworthy, this study protocol utilizes a two-stage rigorous study design in order to develop the effective KT strategy, a so-called “free, limited-edition, theory-driven workshop-format training program,” which may significantly lead to the implementation of optimal safe nurse staffing policy practice. The two-stage rigorous study design consist of (1) development of the workshop-style KT intervention based on Rogers’s (2003) diffusion of innovations theory, and (2) evaluation of the KT intervention using a hybrid design of comparative effectiveness research (i.e., a single-blind, randomized cross-over study design) and analysis of change (i.e., a GCM and an IVI) (Figure 3). The objective and the evaluation indicator of the newly developed KT strategy are to improve nurse leaders’ perceived level of self-efficacy for competence in the use of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting).

Specifically, the evaluation study design can control the main threats to internal validity such as history effect, maturation effect, learning effect, instrumentation-variation effect, regression effect, selection bias, casual time-out, compensation, and demand artifact (Park et al., 2016). Even though the recruitment process of the participants is convenient sampling, making it difficult to control sampling errors and obtain a more representative subset, this study design can control the interaction between the intervention and selection through random assignment, complementing generalizability of the study’s findings (Portney, 2020). Random assignment further guarantees no systemic bias between comparisons through equal distribution of participants to each arm, enhancing the reliability of the study’s findings (Portney, 2020). Statistical corrections can also be performed to control the possible systemic bias between comparisons. Above all, the evaluation study design utilizes a cross-over structure that leads the participants to avoid the risk of being assigned to the study group that receives less benefit, which ensures “equipoise” of the comparative effectiveness research as well as guarantees the participants’ educational equivalency (Park et al., 2016).

GCM along with IVI illuminates a change in the effectiveness of the workshop over time. GCM presents overall change for a sample in the effectiveness of the workshop, describing interindividual differences in change and IVI illustrates a reliable trait-like indicator of a within-person consistency in the effectiveness of the workshop ( Park et al., 2017 ). IVIs are purified residual T scores, which will be obtained from standardizing residuals generated by the best fitting model in the estimation of nurse leaders’ perceived level of self-efficacy for competence using Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting). The analytic procedure strengthens the significance of this study because its calculation process makes IVIs uncontaminated by the learning trend, which is one of the main threats to internal validity as well as external validity ( Park et al., 2017 ; Portney, 2020).

GCM and IVI uncover (a) who shows better outcomes than others over time, (b) why some people show better outcomes than others over time, (c) which contextual factors explain some of those interindividual differences, both in level and in the rate of change, and (d) which contextual factors are correlated to the increase or decrease of the within-person fluctuations ( Park et al., 2017 ). Both of them have an important implication in that they reveal new knowledge, which existing traditional statistical analyses cannot, and shed light on possible solutions to unanswered questions in nursing research ( Park et al., 2017 ).

Study Limitations

Such explicit, prospective implementation research design will evaluate the effectiveness of the workshop over time, that is, its utility can fill evidence gaps and enable course correction in real time (Zambruni et al., 2017). Thereby, the reliability of the follow-up research results through this protocol may be secured. Nonetheless, there are the following limitations: (1) since a single-blind, randomized cross-over study will be conducted, perception bias of the instructors or operational staff who participate in the workshop may, directly or indirectly, influence the study results and (2) if communication between the two arms is possible, participants can know in advance the contents of the workshop, which can influence the research results (LoBiondo-Wood & Haber, 2014).

To address the two limitations, preliminary training for the instructors and operational staff will be provided to control their perception bias as much as possible. Also, participants will be asked to sign a voluntary confidentiality pledge, separate from a consent form, to increase the reliability of the study results. However, even if these preventive measures are applied, further potential drawbacks may still remain because they purely rely on voluntary endeavors of all participating members to control such perception biases. To address the potential drawbacks, a multi-site (physically separating participants between arms), two different education teams (separating A team for the workshop and B team for the mini lesson (online education only) and further separating instructors and operational staffs per arm) can be considered for future study.

Conclusion

This article provides a step-by-step detailed study protocol for (1) the development of the workshop-style KT intervention based on Rogers’s (2003) diffusion of innovations theory and (2) the evaluation of its effectiveness over time with regards to its utility. The protocol will serve as a guide to inform readers of the Florence Nightingale Journal of Nursing “who/how should do what” in a step-by-step manner in their own research/practice setting. Thereby, the protocol will contribute to enhancing nurse leaders’ perceived level of self-efficacy for competence in the use of Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory (providing optimal safe staffing levels fitted to each practice setting). The protocol will accordingly lead the nurse leaders to effectively implement new knowledge—that is, optimal safe nurse staffing levels derived from Park’s (2017) Optimized Nurse Staffing (Sweet Spot) Theory—into the right (safe yet efficient) nursing workforce policy-making within their own context. Such a sequence of processes may consequently contribute to alleviating global nursing shortages.