What can organisational theory offer knowledge translation in healthcare? A thematic and lexical analysis

Agency theory

Agency theory awards primacy to the politic of the individual or agent, highlighting the role of enticements and personal agendas – in brief, ‘It reminds us that much of organizational life, whether we like it or not, is based on self-interest’ ([8], p. 64). This understanding can help to determine the most appropriate type of agent-principal relationship or contract to optimise performance and manage risk – more specifically, when might a contract be used to efficiently shape behaviour, rather than a contact (partly) based on outcomes [22]? For instance, agency theory has been used to understand how the pharmacist role might be expanded and sustained to one of prescriber [23] and how governing hospital boards engage in quality improvement practices [24]. Agency theory holds particular relevance to research on healthcare organisations because of the strong professional identity within medicine [25].

Professional identity is an awareness of the role and functions that one performs or is expected to perform in a social context as a member of a particular profession. It essentially concerns the self and the self in relation with others – be they individuals or organisations [26]. Professional identity is an ongoing concern of the professional involving work practices, social interactions with colleagues and clients, and sense of place within the professional institution and the professional discourse. Professional identity is thus socially bestowed, socially sustained, and socially transformed. Given such reinforcement, professional identity is typically enduring [27]. Within medicine, professional identity appears particularly robust, as demonstrated by the value placed on professional autonomy and efforts to retain it. Given the strong professional identity within medicine, the unit of analysis within this study is the clinician (as agent), and their perceptions of their obligations (or contract) with those they are accountable – namely, their organisation and their patients (as principals; [8]).

However, agency theory is not without criticism. Its capacity to theorise performance has been questioned [28]. It is faulted for its limited accuracy and generalisability, partly due to a perceived reliance on contract design to reduce uncertainty, with limited recognition of social dimensions. This is indicated by references to the ‘problem of unilateralism’ [29] and ‘boundary conditions that limit the theory’s explanatory power of current business phenomena’ – notably, strategic and social entrepreneurship, family businesses, and modern stakeholder relations ([30], p. 182). However, developments that redirect the theoretic gaze from the agent to the organisation and institution serve to address (some of) these critiques – these include: an argument to recognise the complete structure of agency [31]; a repositioning of the firm as a body of agents who endeavour to harmonise their efforts towards a shared goal [32]; an empirical expansion from narrow to demarcated self-interest [33]; and recognition of a bilateral relationship between agents and principals [29]. This (re)interpretation suggests agency theory, ‘is particularly suited to [help to] bridge the gap between theory and application’ ([31], p. 94). Towards this aim, and guided by a rehabilitated theory [32], this study centres on the human agents.

Institutional theory

Institutional theory awards greater primacy to the institution – its rules, requirements, customs, and conventions; that is, its ‘rules of thumb’ ([22], p. 489). These standardisations and assumptions are beneficial for two reasons. First, they foster efficiencies – stock-standard practices enable individuals to rely on experience when attending to mundane or low-priority matters, and direct their energies to novel or high-priority issues [34]. Second (and perhaps more importantly), they offer legitimacy. Rules, requirements, customs, and conventions help organisations to operate, orchestrate, and evaluate [35]. These activities can suggest responsible management [36] and thus shape the beliefs and values of those within the organisation as well as those external to it.

Despite these seeming benefits, an organisation’s ‘rules of thumb’ can also represent a constraint. In the absence of a systemic overhaul, change can be difficult to introduce and sustain [37]. Individuals who are tightly wedded to, or embedded within their social context are largely said to resist change [38]. Although personal agency can help to transform embeddedness into an opportunity for change, albeit incrementally [39], such change still requires negotiation with prevailing organisational norms.

Following the institutional turn in organisational research, reinforcing that institutions are worthy of scholarship [40], there have been calls to refocus institutional theory on its primary focus [41] – namely, to understand how, ‘social structures and processes… acquire meaning and stability in their own right rather than as instrumental tools for the achievement of specialized ends’ ([42], p. 1147). The scholarly scales have tipped towards accounts of typically formalised institutions and the processes therein, much to the neglect of how an organisation operates [43]. Consequently, institutional theory has become diluted, explaining much about legal entities with limited critical insight [44].

To redress this imbalance, there is a need to consider institutional stories [45] – that is, sensemaking processes, particularly those at the individual level, to capture different forms of organisations, and of organising [46]. Recognising a fluid relationship between institutions and individuals [47], this is likely to involve interpretivist methods to examine how organisational actors experience institutions, and how they adopt and justify non-rational practices that seem at odds with economic, isomorphic motivations. As such, this study involved the collection and analysis of qualitative material to elucidate participant accounts of knowledge translation processes and how they interpreted or made sense of these processes.

Situated change theory

As its name suggests, the situated change theory suggests that change is situated in an evolving context, in which individuals adapt or metamorphosise accordingly. Given its strong links to the environment, change is gradual and meaningful, rather than planned and deliberate. Change involves subtle deviations from the paths most travelled and most controlled to pursue a flexible and self-organised path [48]. And only by peering over one’s shoulder, do the paths and patterns become apparent [49]. The metamorphosis of practice change reflects the ways in which individuals make sense of their world [9]. Consider for instance, the ways that situated change theory has helped to understand how individuals improvise with, and adapt the technologies they work with [50]. Research reveals technology-use as an emergent activity, shaped by previous practices, the properties of the technology, political and social dynamics, as well as the task at hand [51]. For two reasons, unscheduled, incremental change can be advantageous, particularly when confronted by demands for change. First, the respite between periods of change can help to reduce stress [52] – it offers a breathing space. Second, respite also represents a time to master new skills [53], which is likely to be easier with reduced stress [54].

Given significant change within many national health systems – including an ageing population, the rise of chronic and preventable diseases, technological developments that enable medical breakthroughs, and the rise of healthcare costs [55] – the situated change theory is likely to hold particular relevance to the study reported in this article. In the wake of the aforesaid changes, governments are vying for different ways to improve the organisation, management, and delivery of healthcare, notably via healthcare reforms. Given the rate of reform, it might be argued that reform is now routine [56]. However, reform often demands considerable change. It requires organisations to restructure, downsize, merge, close, or (perhaps more to the point), do more with less. These sizeable demands suggest that change at the individual level might be eased by opportunities for respite – when there is opportunity to reclaim a degree of autonomy and remain unchanged, like that suggested by the situated change theory.

Relevance to knowledge translation

In addition to representing ‘adjuvant’ and ‘necessary’ theories to further knowledge translation research ([12], p. 33), the agency, institutional, and situated change theories each direct the scholarly gaze to interrelated aspects of an organisation. Akin to the layers of an onion (see Fig. 1), institutional theory clarifies (some of) the parameters of practice – the field in which an organisation operates and the forces that shape what it, and those therein, can and cannot do. Despite these parameters, agency theory highlights the influential role of self-interest. Like an inner layer of the metaphorical onion, it helps to moderate the influence of these forces by underscoring personal agency. Yet this is not to suggest causal relationships between institutional or personal factors and behavioural (or attitudinal) change – for situated change theory suggests that change is an evolution, which occurs through dynamic relationships between factors, structures, and processes; furthermore, this evolution only becomes apparent over time.

Despite their complementarity and relevance to knowledge translation research, the use of the agency, institutional, and situated change theories in this study is not intended to suggest the irrelevance or insignificance of other theories (or frameworks) – organisational or otherwise. Indeed, although beyond the scope of this study, the relevance of theories that direct scholarship to the antecedents that trigger innovation, the processes through which innovation is introduced, and/or the outcomes associated with innovation, have been acknowledged – consider for instance, the diffusion of innovation theory [20]. Also recognised is the value of social theories, including (but not limited to) social capital theory and network theory [12, 57, 58]. And like these, the agency, institutional, and situated change theories have received limited scholarly attention in the context of knowledge translation. As such, and heeding calls for the proactive use of organisational theories [7], a study was conducted, guided by this tri-focal lens, to examine how GPs translate clinical knowledge – particularly evidence-based sexual healthcare – into practice. This knowledge was presented to the GPs in the form of a clinical aide – namely, the STI Testing Tool (withheld for blind-review [2, 59, 60]).

Thematic analysis

The thematic phase involved three overlapping processes: (1) constant comparison analysis [65]; (2) the optimisation of variance [66]; and (3) triangulation through researcher and member checks [67]. Aided by QSR NVivo, constant comparison analysis involved methodically coding the material and constructing themes from the codes [68]; themes were identified in relation to the study aim and objectives, and were compared between GPs. To protect GP identity, names and identifying characteristics are not disclosed. The optimisation of variance involved ensuring that descriptions and explanations about the research material contained both typical and atypical elements [66] – that is, similar themes (e.g., the types and sources of clinical knowledge) as well as those that were unique (e.g., organisational strategies that promote sexual healthcare). Triangulation involved the participation of three individuals in the analysis [67] – two researchers (one of whom is relatively unfamiliar with relevant research) and one Division representative. Discussion of their constructed themes helped to increase the rigour and trustworthiness – or dependability [69–72] – of the findings. These three interrelated processes helped to ‘decrease, negate, or counterbalance the deficiency of a single strategy, thereby increasing the ability to interpret the findings’ ([73], p. 253).

Lexical analysis

The lexical phase involved conceptual and relational content analysis [74]. To optimise the likelihood of a methodical approach [75], this process was aided by Leximancer – data-mining software that uses Bayesian reasoning to detect key concepts and reveal their relationships [76]. Using algorithms, Leximancer identifies frequently occurring and co-occurring words and amalgamates these to form and visually map concepts that reflect themes within the text [77]. The maps convey three types of information – ‘the main concepts in the text and their relative importance; the strengths of links between concepts (how often they co-occur); and similarities in contexts where links occur’ ([78], p. 1735). Clusters of concepts within a map suggest contextual similarity. The components of these concepts are ordered within a thesaurus, comprised of relevant words and weightings to indicate relative importance. By default, two-sentence blocks of text are then assessed to identify evidence of conceptual relevance – however, given the concise nature of some responses, one-sentence blocks of text were assessed. For further detail on Leximancer and its validation, relative to other coding practices, see Smith and Humphreys [79].

Following Angus-Leppan and colleagues [74], the recommended Leximancer analysis procedure was used. This involved aggregating the transcripts from both stages of data collection and using the discovery to determine the concepts automatically generated by Leximancer. This helped to determine how the data might be refined and how relevant concepts were profiled. For instance, it was evident that interviewer comments dominated certain concepts. For this reason, all interviewer quotes were suppressed to give voice to the GPs. Although this approach potentially discounts the interviewer-interviewee exchange and context, this was addressed by sourcing the full transcript when analysing relevant concepts.

Preparing the interview transcripts for analysis involved two steps. First, given the focus of this study on knowledge translation, the concepts automatically generated by Leximancer were suppressed and relevant concepts were created to represent clinical knowledge and clinical practices. This involved reviewing excerpts that represented relevant concepts and merging those that were conceptually similar. For instance, the concept, ‘clinical knowledge’, was created by merging, ‘antibiotic guidelines’; ‘card’ and ‘placard’, denoting the STI Testing Tool; ‘electronic therapeutic guidelines’ and ‘ETG’; ‘guidelines’; ‘Harrison’s’, as in the principles of internal medicine [80]; ‘Murtagh’s’, as in the reference guide on evidence-based primary care [81]; ‘protocol’; ‘Red book’, which is the Royal Australian College of General Practitioners guidelines for preventive activities in general practice [82]; ‘STI Testing Tool’; ‘TGA’ (or the Therapeutic Goods Administration); and ‘therapeutic guidelines’. Similarly, the concept, ‘clinical practices’, was created by merging, ‘screen’; ‘screening’; ‘test’; ‘tested’; ‘testing’; ‘tests’; ‘treat’; and ‘treatment’. Guided by previous research [78], thirty new user-defined concepts were profiled to examine the concept maps of ‘clinical knowledge’ and ‘clinical practices’ – this number was selected to avoid diluting the focus of the created concepts, ‘clinical knowledge’ and ‘clinical practices’. Despite the demonstrated value of concept profiling, Young and Denize [77] warned that it moves the analysis away from the overall concepts and themes revealed by the material, to one focused on a smaller part of an already relatively small dataset. Although concept profiling helps to focus researcher attention to pertinent concepts and themes, the approach also risks leaving more general concepts and themes unexplored. Given this study focuses on the translation of knowledge into practice, concept profiling was deemed appropriate.

Together, concept profiling and tagging the transcripts helped to view the research material through the tri-focal theoretical lens. This involved two techniques. First, as per previous, thirty concepts were profiled using the themed discovery setting, ‘Concepts in EACH’, to ‘discover concepts that distinguish… categories from one another’ ([86], p. 154). For instance, reflecting agency theory, which ‘reminds us that much of organizational life, whether we like it or not, is based on self-interest’ ([8], p. 64), the setting – ‘Concepts in EACH’ – was used to compare GPs who would increase STI testing if subsidised by the government as a Medicare service, with those who would not increase STI testing (see Table 2). Similarly, this setting was used to compare GPs who would increase STI testing if provided with an incentive payment for each STI test performed, with those who would not increase STI testing. Second, concepts close to a tag (e.g., the tags to identify those GPs with less than, or more than ten years in general practice) were examined to determine whether they reflected a shared or an individual experience – this was aided by sourcing excerpts that epitomised the concept under investigation. This helped to identify the concepts that were relevant to the study and as such, should be included within the map. Conversely, concepts deemed irrelevant were suppressed.

Relevant concepts were then analysed in two ways. First, the proximal distances between tags and concepts were considered. Second, excerpts associated with the created concepts, ‘clinical knowledge’ and ‘clinical practices’, were sourced and examined in relation to each tag. Reflecting previous research [79], because the two created concepts were common to all maps, there was considerable similarity between them. Yet, of value here is the ability to unveil similarities and differences between the maps, as well as between the thematic and lexical analyses.

Thematic analysis

Lexical analysis

Furthermore, incentive payments for sexual healthcare were associated with discourse around, ‘access’. For instance, some GPs commented on the need for information that was obtainable and user-friendly; while others considered how an increase in sexual health testing might increase consultation times and affect patient throughput:

that’s actually quite extensive. If you really do the HIV [human immunodeficiency virus] pre-test counselling properly, or as it’s done in the books, that alone could be a whole consultation.

Key findings

Thematic analysis

Lexical analysis

Complementing this thematic analysis, a lexical analysis was conducted by developing three concept maps. Within the first, the 21 GPs were categorised according to their years of experience (see Fig. 4); within the second, they were categorised according to their year of graduation (see Fig. 5); and within the third, the GPs were categorised according to whether or not they were a Fellow of the RACGP (see Fig. 6). Each map is interpreted in turn.

Together, Figs. 4 and 5 suggest a form of cognitive inertia is associated with discourse on sexual healthcare, when considering the GPs according to their years of experience and year of graduation. Those with over ten years of experiences and (related to this) those who graduated before 1982 seemed to own the language on ‘clinical practices’, relative to their inexperienced counterparts. This might indicate a greater preference for institutional ‘rules of thumb’ ([22], p. 489) than different forms of ‘clinical knowledge’. What these findings reveal is that experience is a powerful predictor in terms of engagement with clinical guidelines and resources. Although inexperienced GPs might frequently draw on external resources to supplement their knowledge-base, this tendency perhaps for some, fades over time.

Key findings

The thematic and lexical analyses reveal the relevance of institutional theory by highlighting the ways in which embeddedness within an organisation and a profession can shape clinical practices. Although the thematic analysis largely suggests the GPs were somewhat disconnected from institutional life, it also revealed the value of institutional savviness. Extending this finding, the lexical analysis illustrated how experience can be both an asset and a limitation when attempting to engage with, and keep abreast of evidence-based practices.

Thematic analysis

Lexical analysis

Reflecting situated change theory, this concept map illustrates the dynamic, gradual, vacillating, and faltering nature of change. The mere exposure to, and comprehension of knowledge artefacts is unlikely to simply align individual behaviours with evidence, as defined by accredited organisations. But rather, knowledge artefacts might help to ignite (even if temporarily): reflectivity and reflexivity [89] around whether and how change might happen, as well as marginal improvisations, with their associated slippages. This reveals the situatedness of knowledge translation – a process of mini defeats and victories, shaped by contextual nuances. The GPs, as registered professionals who largely worked behind closed doors, performed knowledge work within a context that simultaneously structured and freed their thinking and their practices to determine, ‘what things get done and how they are done’ ([90], p. x). This in turn can open additional opportunities for change – as Weick indicated:

The advantages of emergent change include its capability to increase readiness for and receptiveness to… change and to institutionalize whatever sticks… sensitivity to local contingencies… experimentation, learning, and sensemaking; comprehensibility and manageability; likelihood of satisfying needs for autonomy, control, and expression; proneness to swift implementation; resistance to unraveling; ability to exploit existing tacit knowledge; and tightened and shortened feedback loops from results to action ([91], p. 227).

As such, seemingly trivial change in patient consultations, the consideration of resources, or recognition of awkwardness around sexual healthcare might prime the GP towards evidence-based care – not necessarily as defined by accredited organisations, but rather, as a practice that reflects the situated blend of clinician expertise; patient (and potentially carer) preferences; available resources; and the context of care. This is supported by the relative infrequency of those exceptional strategies used by a small number of GPs to explicitly change their sexual healthcare practices. The challenges that GPs face in familiarising with, synthesising, and adhering to ‘evidence’, while managing their workload and accommodating patient expectations, might suggest that these exceptional strategies are precisely that; the exception. Most of their counterparts appeared to inhabit a space that either: primed them to varying degrees of (less exceptional) change; or pushed them back towards the familiar.

Key findings

Although the thematic analysis elucidated the factors that influenced change among the GPs, the lexical analysis clarified the dynamic nature of change. This confirms the complementary value of the two approaches.