Implementation Science and Employer Disability Practices: Embedding Implementation Factors in Research Designs

Implementation is inextricably linked with management. Organizational science has adopted evidence-based practice principles, and as a result, Imp Sci has tended to take what has been described as a predominantly ‘rational’ (or structured problem-solving) approach. Thus the decision maker should first identify the problem, then search for and generate alternative courses of action, implement the option selected, and then evaluate the outcome [32]. It has been assumed that acknowledgement of evidence in the ‘science-informed practice of management’ [33], is not only desirable and perhaps necessary, but also sufficient. However, numerous authors have argued that this view requires qualification. For example, Baba and Hakem Zadeh [34] described a model in which evidence based on judgment, education and experience affects the decision-makers’ options as well as their actual decision, and in which such decisions are moderated by context, management preferences and values, as well as by stakeholders’ preferences and values. It is important also to consider the decision process, characteristics of the decision-maker and the context when implementing evidence-based management practices [35]. It has also been shown that decision makers may deviate from a strictly ‘rational’ approach in weighting the current status more strongly or seek the minimum requirements needed to satisfy choices rather than necessarily optimizing outcomes [36]. Further, in Wright et al.’s analysis [35], decisions are often made in a social environment allowing for political bargaining [37] and other communications between participants [38]. Based on their study of a program implemented in an emergency department in Australia, Wright et al. [35] found that the implementation process began with problem recognition and then proceeded to assembling evidence before exchanging evidence across disciplines and decision-makers. This resulted in reformulating the problem, engaging stakeholders and generating alternatives. This process resulted in commitment to the evidence-based solution and implementation. Wright et al. [35] conclude that it is important to recognize “situated expertise” among decision-makers and people who are actually implementing a new program or policy. (By situated expertise they are referring to the proficiency and judgment that individuals have as a result of their experiences, education and practice.)

Viewing implementation as a decision process emphasizes that the decision-making process and actions occur in the context of the involvement of multiple individuals within the organizations, and possibly, stakeholders outside of the organization. The widespread use of teams within organizations suggests that the project management literature may be informative in understanding the process of implementation, especially in the initial stages, although, project teams tend to have a specific goal and a limited life as opposed to the apparently unlimited time line of regular organizational staff [39]. In order to improve the implementation processes, implementation teams need to seek feedback, experiment, and discuss errors that are made [40]. Finally, Horwitz [41] proposes that functional diversity in project teams, and we would argue, in implementation teams, improves the likelihood of successful implementation due to likely greater flexibility. In the next section we will consider the contribution of knowledge translation.

The advancement of the science of knowledge translation or how to most effectively promote and support the use of evidence in health and healthcare policy and practice, is challenged by the plethora of terms, models, frameworks, and heterogeneous interventions employed in the field [42]. Broadly, knowledge translation is the synthesis, dissemination, exchange and ethically sound application of knowledge to improve health and well-being [43]. It offers a “technology” for change which is potentially applicable to the management of work disability, but it is necessary to ask how this might be achieved.

There have been many attempts to systematise interventions. The Cochrane Effective Practice and Organization of Care (EPOC) Group [44], for example, is a widely used classification scheme, but recently, in a scoping review of interventions, Lokker et al. [42] identified 51 diverse classification schemes, and described them in terms of content, focus and methodology. The content areas include dissemination and implementation, knowledge translation, quality improvement, knowledge transfer and research utilization. Lokker et al. [42] updated an earlier review [45] by including policy articles and adding search terms related to knowledge translation to capture broader classification schemes. There have been attempts to simplify the field of Imp Sci by developing some broad, over-arching frameworks. One of the better known is the Consolidated Framework for Implementation Research (CFIR) [46].

The phases of implementation (EPIS) in the CFIR and possible contributions aligned with inner and outer contexts, are illustrated in Fig. 2. Phase 1, described as Exploration, is characterized by developing an awareness of an issue requiring attention (e.g. a desire for an improved approach to a problem) and should include consideration of the question in terms of the possible inner and outer contexts.

Phase 3 comprises the Active Implementation of the plan, which is expected to require engagement at inner and outer contexts, as well as fostering supporting linkages between them. Aarons et al. [47] suggest that the scale of the implementation project is also likely to have important implications for specific issues in both outer and inner contexts. For example, at the outer level there may be large system issues, like funding availability, while at the inner level there may be issues that concern the fit with the productivity and other work demands of individual workers. Other inner context issues likely to require addressing might include: readiness to change (which may vary within an organization), the receptivity of the organizational culture to change, the current ‘Organizational climate’ (e.g. employees’ perceptions of their work environment), and how well the implementation plan fits with the existing values of the organization and its workers.

Phase 4 is concerned with Sustainment of the intervention or the continued use of the intervention (or innovation) as standard practice within an organization. Aarons et al. [47] acknowledge that this aspect of their model has the least systematic supporting knowledge, with little empirical work on which to draw. Nevertheless, if the implementation is to move beyond mere demonstration, it is essential to consider sustainment. The problem of maintenance of change has long been recognized as a challenge in the clinical literature [48] and is reflected in the organizational sphere in the recognition that sustained return to work after injury may be of more relevance than speed of return to work (further discussion of outcome measures is offered in the companion paper [1]) and this may involve a series of changes as in the Organizational Readiness for Change (ORC) model [49], and the “self-regulation” model [50]. However, further empirical work on the determinants of change in the context of sustainment would seem to be merited.

To test the application of the Aarons et al. [47] classification model to disability prevention practices, we conducted a brief keyword search [“disability” and “accommodation” and “implementation”; “return to work” and “implementation”; (“presenteeism¹” or “stay at work”) and “implementation”] that identified nine articles [25, 53‐60], describing implementation factors related to absence management² and RTW programs (one systematic review, four process evaluations, and four conceptual/theoretical summaries). Collectively, these 9 articles made mention of 89 factors influencing implementation, and we found these factors could be organized within the four-phased EPIS conceptual framework [47] without difficulty (Table 1). Examination of these factors shows similarity in the implementation issues raised by various researchers, and this sorting does suggest actionable problems that could be addressed.

Lack of fidelity during initial implementation may lead to underestimation of efficacy of the intervention and weaken the strength of conclusions [66]. Assessment of fidelity to behavioural interventions that require direct human observation and judgment has proved a challenge [67, 68]. However, technical solutions have been recommended to scale-up the evaluation and quantification of such behavioural interventions. For example, Atkins et al. [69] and Balsubramanian et al. [70] propose blending quality improvement and Implementation Research. This approach has been termed Learning Evaluation, in which qualitative and quantitative data are collected to conduct real-time assessment of implementation processes while also assessing changes in context, facilitating quality improvement using run charts, audit and feedback, and generating transportable lessons. If these principles could be applied across organizations they would merit consideration for adaptation to organisational settings.

Organizational culture has been defined as the shared values, assumptions, and beliefs that are communicated in the behaviors that the organization uses to overcome prior problems, thereby validating the importance of these actions [71, 72]. More specifically the distinction has been made between artefacts (or the most visible or easily accessed layer, such as people’s dress or the physical environment; the meaning or significance of which may vary from organization to organization); the espoused values (which may or may not be consistent with how the organization actually operates) and finally, the deepest layer comprising the underlying assumptions that are typically shared throughout the organization and that drive how employees interact and behave, Thus the concept of organizational culture is generally quite broad, encompassing all of these layers and almost all aspects of organizational life. Research has suggested the importance of a number of dimensions of organizational culture for implementation success across a variety of settings [73‐78].

Organizational climate has been defined as the shared perception of the work environment including the policies, practices, and procedures that guide the expected, supported, and rewarded behaviours [71, 79, 80]. Although some climate researchers examine the general work environment that employees experience (or the molar climate [81]), when specific strategic outcomes are of interest), it may be helpful to adopt a more specific focus [80, 82]). In recent years, there has been a growing interest in the concept of implementation climate. Implementation climate [83, 84] is a global construct consisting of items related to expectations, support and rewards and has been suggested as an integrative framework linking the organisation and the worker [46]. It has been defined as “employees’ shared perceptions of the importance of innovation implementation within the organization” [84] (p. 813) and captures the expectations, support, and rewards associated with implementation [85]. Multiple measures of implementation climate have been developed [81, 84, 85], with generally supportive evidence for their reliability and validity.

In summary, implementation climate involves employee perceptions of what happens in the organization and implementation culture focuses on why it happens [86]. Unfortunately, while these may have explanatory utility, to date, outcomes of attempts to change organisational culture in health care, have been disappointing [87], and it has been argued that a focus on changing organizational climate may be more fruitful [81].

Although leadership, in general, has been shown to be important for effective implementation [106, 107], efforts that do not consider both contextual and individual factors likely to facilitate or hinder implementation are likely to result in sub-optimal outcomes. In particular, Aarons et al. [108] argue that strategies that involve assessment, intervention, and support for implementation at multiple organizational levels should have a greater likelihood of success. Organizational readiness has been described in terms of its theoretical basis [109], its conceptualisation and measurement [110, 111] in terms of organizational members’ change commitment and change efficacy to implement organizational change [111], and in its utility in investigating influences on the implementation of worksite healthcare promotion programs [112]. More recent developments include a decision tool for assessing organizational readiness [113] and a detailed protocol for an organizational readiness intervention [114].

Franche et al. [25], however, have cautioned that in the area of work disability prevention there may be significant differences in the perception of roles and tasks between stakeholders (e.g. workers, management, health care providers). This suggests that such potential barriers should be identified and, if possible, addressed prior to implementation of the intervention [115]. Importantly, Franche et al. [25] also cautioned that perfect agreement between stakeholders may not be possible, and stakeholders may need to find ways to accommodate these differences.

In addition to acknowledging the link between the organizational and individual perspectives [104], recognition of both the positive and negative aspects of the work environment may also be important [116, 117]. Recognition of the importance of the facilitation of positive adaptation to problems of ill-health and health-related work compromise has led to the view that well-being is not only an outcome of intervention, but also is a potential mediator of improved adjustment and performance. However, worksite wellness programs are characterized by complex pathways [112, 118, 119] and to date have shown only modest treatment effects [120]. Martins [121] has identified five major threats to the success of organizational change efforts: (a) lack of an adequate framework for implementing organizational change; (b) failure to accurately identify the problem; (c) inaccurate diagnosis of the problem and its root causes; (d) lack of fidelity in the implementation of a planned intervention; and (e) inadequate measurement of the resulting effect or insufficient time given. Although such challenges are not specific to worksite wellness programs, it would seem sensible at this time to defer further comment on their utility as an intervention for work disability until there is a clearer understanding of the mechanisms of change and in the implementation of interventions.

Traditionally, disability management has been built on the three pillars of prevention, work accommodation and support for recovery, although how these have been implemented in different contexts has depended on policy at both a national/agency level and negotiated conditions of service and entitlements. There are also differences across jurisdictions in terms of legal responsibility of employers for sickness management in general and for specific injuries in particular [122]. In tackling work disability, there has been a major focus on primary prevention, (with worker centered education and instruction) and the minimization of danger, whether in the design of environment or in job design (in terms of its physical and psychological demands). Much of this can be developed of course at the level of the workplace and workforce. Secondary prevention, in terms of the interventions described in other chapters, may also be tackled at a workforce level, but at early stages typically will have involved clinical healthcare. In secondary prevention the primary focus has been on the amelioration of symptoms and re-integration into work, sometimes with phased return-to work or work modifications [19].

However, as mentioned above, there has been a change in understanding of the impact of work on health [2] and the introduction of wellness initiatives [104] has become a feature within many organizations. Implicit in many of the traditional approaches has been an assumption that “restoration” of physical and mental health will be sufficient to ensure return to work, but as mentioned earlier, interventions aimed at these targets have often failed when workplace factors have not also been addressed [22]. The process of work engagement/re-engagement may have multiple consequences for employees’ performance [123] and in organizational research there has been a broadening in perspective from attendance management to sickness management and a concomitant shift in emphasis from symptom management to the enhancement of well-being. It would seem that attention to facilitating engagement and re-engagement in work has the potential to assist those returning to work after illness or injury.

Engagement in work traditionally has not fallen within the purview of occupational health or the rehabilitation literature, but perhaps merits consideration in the context both of the enhancement of well-being and in pain management [18]. Schaufelli and Baker [124] acknowledge that the term “work engagement” has been understood somewhat differently in business and academia, but offer an integrative model linking characteristics of work (resourceful and challenging) and positive affect, with work engagement (characterized by job satisfaction and job involvement) with organizational commitment and with enhanced performance (evidenced, for example, in discretionary effort).

In summary, there is a case for reconceptualizing the challenge of work disability as one of sustained work re-engagement, often in the context of ongoing symptomatology, rather than one primarily of clinical cure or job redesign/accommodation. For such a shift in emphasis to gain any traction however, it is necessary to consider the challenge both from an employer’s and a worker’s perspective.