Article Text
PDF
Abstract
Introduction The failed or partial implementation of clinical practices negatively impacts patient safety and increases systemic inefficiencies. Implementation of sepsis screening guidelines has been undertaken in many settings with mixed results. Without a theoretical understanding of what leads to successful implementation, improving implementation will continue to be ad hoc or intuitive. This study proposes a programme theory for how and why the successful implementation of sepsis screening guidelines can occur.
Methods A rapid realist review was conducted to develop a focused programme theory for the implementation of sepsis screening guidelines. An independent two-reviewer approach was used to iteratively extract and synthesise context and mechanism data. Theoretical context-mechanism-outcome propositions were refined and validated by clinicians using a focus group and individual realist interviews. Implementation resources and clinical reasoning were differentiated in articulating mechanisms.
Results Eighteen articles were included in the rapid review. The theoretical domains framework was identified as the salient substantive theory informing the programme theory. The theory consisted of five main middle-range propositions. Three promoting mechanisms included positive belief about the benefits of the protocol, belief in the legitimacy of using the protocol and trust within the clinical team. Two inhibiting mechanisms included pessimism about the protocol being beneficial and pessimism about the team. Successful implementation was defined as achieving fidelity and sustained use of the intervention. Two intermediate outcomes, acceptability and feasibility of the intervention, and adoption, were necessary to achieve before successful implementation.
Conclusion This rapid realist review synthesised key information from the literature and clinician feedback to develop a theory-based approach to clinical implementation of sepsis screening. The programme theory presents knowledge users with an outline of how and why clinical interventions lead to successful implementation and could be applied in other clinical areas to improve quality and safety.
- Healthcare quality improvement
- Implementation science
- Infection control
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Introduction
Sepsis is a common and deadly condition that can occur in all in-patient populations,1 leads to significant long-term morbidity1–5 and contributes to significant financial healthcare costs.6 Caused by dysregulated physiological response to infection,7 sepsis is life-threatening because its rapid progression can necessitate urgent attention for less severe cases and emergent intervention as severity increases. Therefore, rapid recognition and treatment are paramount. Evidence-based clinical guidelines for the early recognition and treatment of sepsis have been developed and implemented in a variety of settings: however, successful implementation of these interventions has been mixed.8–17
Despite a multitude of theories that seek to explain behaviour change,18 19 a practical understanding of how to successfully introduce new evidence-based clinical practice guidelines, such as sepsis screening tools, is not well understood. Failed or partial implementation of clinical guidelines is common, negatively impacting patient safety, patient outcomes and systemic inefficiencies.20–22 The ongoing issue of partial or failed implementation has been attributed to the lack of a theoretical understanding of how successful implementation of clinical guidelines occurs.20–34 Change theory can explain how certain variables and processes create change for individuals or groups and can be derived from substantiated and research-based theory or an implicit understanding of how programmes work35 as well as how a programme theory can align with clinical quality improvement.
The implementation of clinical practice guidelines at the local level (ie, clinical area such as a surgical unit) is often ad hoc or through intuitive planning despite evidence that supports increased success using a theory-based approach.30 Barriers to theory-based approaches are dependent on local needs that often dictate that the there is rapid implementation of a clinical guideline,36 involve iterative evaluation37 and occur in a dynamic environment.25 A programme theory brings together theoretical explanatory constructs related to change and action from multiple sources, including theories and frameworks, which can translate into practical applications.35 A middle-range programme theory (MRT) attempts to frame these same explanatory constructs at a mid-level of abstraction in order to address social phenomenon, such as behaviour change, in different contexts.38 Therefore, the realist MRT seeks to define substantiated constructs from existing theories with applied supporting processes and actions related to real-world circumstances, thereby providing a basis for answering how, why, for whom and in what circumstances successful implementation of something like a sepsis screening protocol can occur.
Multiple approaches to screening, diagnosis and treatment of sepsis have been explored;39–41 however, what causes successful implementation is not well understood. The reasons for variation in adoption and compliance of sepsis interventions have been mostly speculative or suggestive at various levels of application. System-level rationales for successful adoption and compliance have included the presence of quality improvement or process improvement systems42–45 and an institutional execution strategy including leadership alignment, standardisation of evidence-based practices across clinical areas, project management, and collection and use of timely and actionable data available to practitioners.46
Rationales for successful local-level adoption and compliance are equally variable and vague, highlighting a need for a more theoretical approach to implementation. The many strategies associated with the adoption and improvements in sepsis mortality and hospital length-of-stay include a hospital-wide code sepsis and rapid response team,40 47 48 standardised order sets,40 47–50 decision-making algorithms,40 50 protocols allowing non-physicians to intervene early,47 targeted education48 50 and electronic health record trigger.48 Although this might be a helpful resource list to improve sepsis identification and treatment goals, there is little explanation of why and how these do or do not work. Implementation is often hampered by the multilevel system complexity that surrounds different contexts in which implementation achieves desired outcomes. Failure to successfully implement clinical interventions, such as those for the early recognition and treatment of sepsis, directly affects patients and the health system.1 3
Objectives and focus
The purpose of this research was to construct a programme theory that identifies and describes the salient resources (R), contexts (C) and mechanisms (M) that lead to an outcome of successful implementation of a sepsis screening intervention. Understanding why and how successful implementation of a sepsis screening protocol occurs in the form of an MRT, will help to inform successful implementation of this intervention and others in similar settings. This rapid realist review (RRR) was conducted concurrently with the implementation of a sepsis screening protocol at a tertiary hospital in Canada. Initiated by the knowledge-user group overseeing this implementation, the objective was to understand what leads to successful implementation. The protocol was implemented in medical, surgical and emergency departments; therefore, the scope of the review focused on similar clinical areas.
Methods
Rationale
An RRR36 seeks to derive a causative proposition using the realist premise that context+mechanism=outcome for phenomenon that have a narrow focus and are informed by knowledge users that can be rapidly synthesised for timely application in practice. An RRR works backwards from the intended intervention outcome(s), in this case successful implementation, to reveal families of interventions and the surrounding contextual factors that trigger key mechanisms.36 The RRR makes the merging of theory with practice more pragmatic as the results tend to be responsive to local policy or programme needs and are utilisation focused.36 The methods process map summary is captured in online supplemental appendix A.
Supplemental material
Search processes
Literature was searched using a focused, iterative approach to identify key informative peer-reviewed and grey literature sources (figure 1). Databases included MEDLINE, EMBASE, Scopus and Google Scholar. Using focused search criteria (table 1) generated in collaboration with the local knowledge-user group, the titles and abstracts of 1123 published articles were screened using broadly defined inclusion criteria (box 1). Following the RRR search strategy iterative searches included references provided by content experts and knowledge users.36
Document search flow diagram for realist synthesis adapted from Wong, Greenhalgh, Westhorp et al.75
Rapid search strategy
Reference inclusion and exclusion criteria
Inclusion: first screening
Sepsis programme, intervention related to sepsis screening, ‘protocol’, ‘bundle’ and ‘pathway’.
‘Implementation’ or factors related to implementation, ‘evaluation’ and ‘performance’.
Early goal-directed therapy, clinical practice guideline and ‘reduced mortality’.
Inclusion: second screening
Descriptions of implementation that are potentially relevant and rigorous.51
Exclusion
Reference to ‘paediatric’ or ‘neonatal’ patient populations as these areas were not subject to the former Saskatoon Health Region sepsis trigger tool implementation (online supplemental appendix B).
Supplemental material
The second screening at each step of the search was based on the criteria of relevance and rigour. Evidence is relevant if it addresses the MRT of interest.51 Evidence is rigorous if a particular inference is drawn by the original researcher in such a way that is has sufficient weight to make a methodologically credible contribution to the test of a particular intervention theory.51 Establishing relevance and rigour in the screening and appraisal process can be done if there is one instance in an article, regardless of the section, where the original researchers describe or reflect on the process and impact of the intervention.52 Table 2 illustrates articles excluded after the second screening for each step in the search strategy. Eighteen articles (online supplemental appendix C) were included that contained information related to successful implementation and were consistent with realist definitions of context or mechanism.36 53
Supplemental material
References included and excluded with second screening at each search step
Data extraction and synthesis
Data of interest included descriptions of realist configurations: context (C), mechanism (M) and outcome (O).54 Resource (r)55 was included to purposefully differentiate concepts of resources and reasoning within a mechanism. Resource refers to the mechanism component introduced into a context; reasoning is the hidden intrinsic quality of actors reflected through thinking and decision making.56
An independent two-reviewer approach was used to iteratively extract and group descriptive data relating to contextual factors and mechanisms.57 This involved working backwards from the predefined outcome, successful implementation, and identifying emerging patterns of construct pairs (CM, MO, CO) and CMO configurations. Abductive reasoning was applied at each step of data synthesis, including reconciling, consolidating and situating.52 58
Construct pairs and CMO configurations were organised into explanatory accounts (EA) followed by consolidated accounts (CA)59 to preserve relevance and rigour while facilitating synthesis in a timely manner.60 EAs were assembled if they suggested a possible contextual factor and/or mechanism and recorded in the form of ‘if and then’ statements. Any EA that was identified as being illogical or overly abstract (ie, not fitting a middle-range description) was excluded. Any discrepancies between the reviewers were discussed and refined until there was agreement on inclusion or exclusion.
Each reviewer independently grouped EAs by themes (online supplemental appendix D) and together they synthesised all EAs within each group into single CAs (online supplemental appendix E). CAs have the same form as EAs, but because they are a synthesis of multiple EAs, their meanings can be more precise and theoretically rich (ie, it accounts for ‘how’ and ‘why’ where the first element leads the second).60 CAs were then grouped into draft CMO configurations (online supplemental appendix F). Throughout the synthesis process, reviewers applied three basic principles to guide their abductive reasoning: keep CMO configurations parsimonious by identifying ‘key’ constructs36; ensure the degree of CMO abstraction was middle-range and ensure the constructs, particularly mechanisms, are aligned with realist definitions.
Supplemental material
Supplemental material
Supplemental material
Role of substantive theory
A substantive theory is one that exists in the reviewed literature on change and action and is related to the field that is the subject of the synthesis.61 Existing substantive theories help to make sense of patterns of findings in a realist synthesis and provide structure to the programme theory. Whether one or more substantive theories are included in the programme theory synthesis is determined by its presence in the literature on the phenomenon of interest. If there is no evidence of a substantive theory being applied in a real-world context, then there is none included in the synthesis.
Programme theory validation
Rigour and relevance were ensured by a two-step validation process involving clinician knowledge users because they know the most about the interventions that they are responsible for, the context within which an intervention functions and the agents through which any change occurs. Step 1 involved a focus group with knowledge users from a local sepsis quality improvement group to discuss the CMOs. In step 2, clinician end-users were interviewed to confirm, refute or refine CMOs. The need for step 2 became apparent during the initial synthesis because knowledge users found the programme theory to be overly linear and potentially missing pertinent linkages that reflected real-life complexity.
Knowledge user validation
The focus group involved a physician and three nurses with a range of experience in clinical roles, clinical areas and educational backgrounds (table 3) including experience in developing and trialling the Saskatoon Health Region trigger tool (online supplemental appendix B). The focus group, facilitated by JM, was conducted over an hour.
Demographics include those from participants in the focus group (FG) and interviews
Participants were selected by convenience sampling and recruitment was by invitation to all members of the working group with a copy of the consent agreement.
The principles and process of realist interviews62 were used to structure focus group questions and emphasised theory gleaning, refinement and consolidation. The focus group allowed knowledge users to explore contextual circumstances in relation to theoretical constructs from the literature by interacting and building on each other’s ideas. To maintain the theory as the focus, a series of semistructured questions, graphical representations of the literature-based programme theory propositions and text descriptions of the same propositions were used (online supplemental appendix G). New ideas or considerations from participants were explored in the discussion. Transcripts were analysed for whether participants confirmed, refined or refuted CMOs, and the introduction of new constructs (online supplemental appendix H) were incorporated into a revised programme theory (online supplemental appendix I).
Supplemental material
Supplemental material
Supplemental material
Clinician validation
To collect sufficient data, individual semistructured and audiorecorded interviews were conducted for further refinement of the theory. Interviews were completed with 11 nurse and physician participants from emergency, surgical and medical wards where the local sepsis screening intervention had been trialled. Sampling utilised convenience and snowballing, with four individuals declining to participate. Recruitment activities included emails to identify key contacts, posters displayed in clinical areas and participant referral. Nurses, medical residents or physicians who practised in the trial clinical areas were eligible. The interviewer provided revised statements and graphical representations summarising each proposition from those used in the focus group (online supplemental appendix J). Each proposition was read audibly, and the participant was asked to confirm, refute or refine it. Interviews lasted 30–60 min. Refinement involved the participant and interviewer discussing why a statement needed refinement, what was missing or needed to be added, and exploring ideas and concepts that were not necessarily directly related to the statement (online supplemental appendix K). If a response required further inquiry, the participants were engaged to reveal rationales and meaning behind their ideas and impressions.
Supplemental material
Supplemental material
Results
Substantive theory
The theoretical domains framework (TDF) emerged as the only theory-based structure resembling a substantive theory from the literature review.63–65 The TDF provides a theoretical structure for examining the influence of cognitive, affective, social and environmental factors on behaviour especially around the implementation of evidence-based recommendations.66 It is derived from 33 synthesised theories relating to behaviour and behaviour change, which are developed to make psychological theory more accessible to implementation.63 67 It is composed of 14 theoretical behaviour change domains (eg, belief about consequences, environmental context and resources, and behavioural regulation) on which interventions are designed to change practice and transfer knowledge, such as guiding quality improvement68 and targeting interventions.69 Matching elements of the theoretical determinants to CMO constructs resulted in the TDF being incorporated into propositions of this programme theory.
Final programme theory
Successful implementation
Successful implementation is achieved when the fidelity of the intervention is maintained and there has been sustained use. Fidelity indicates the effectiveness for behaviour change28 because it is a measure of whether a change is implemented as intended.70 Fidelity demonstrates both adoption and adherence, which are the defining endpoints in the programme theory developed in this research. Fidelity, an often under-reported outcome in quality improvement studies, is a central outcome in understanding if the overall success or failure of a practice change is related to the effectiveness of the new practice or the adherence to it.28 To achieve this final outcome, there must be two successive intermediate outcomes achieved (figure 2). First is where clinicians accept various aspects of the intervention including content, complexity, comfort and credibility and find it feasible, that is, usable, suitable and trialable in their setting.71 Second is when clinicians adopt and use the intervention.
Overview of the final program theory outcomes for the implementation of a sepsis screening protocol within which promoting and inhibiting mechanisms operate. Stop signs represent inhibitory outcomes. Caution signs represent where mechanisms diverge for the different actors at each stage of successful implementation.
Achieving acceptability and feasibility
Acceptability and feasibility occurred when clinicians believed that the sepsis screening intervention benefits their workflow and benefits their patients. This mid-level mechanism is achieved by using standardised evidence-based approaches to clinical practice and implementation in the context of a complex work environment (table 4).
Programme theory summary including the salient mid-range mechanisms, CMO propositions and key resources and functions of an implementation strategy
It is promoted when several salient lower-level mechanisms are triggered. Improved action planning and plan execution occurs when the perception that the intervention is practical and realistic is triggered. A sense of leadership among clinicians comes from the motivation inherent in engaging in practice change that, when triggered, leads to modelling and peer influence. When triggered, positive reinforcement of self-efficacy leads to confidence in clinical decision-making, especially among nurses as well as the perception of smooth interdisciplinary processes, which enables clinical judgement among physicians when intervention parameters do not match a patient’s clinical presentation.
Achieving adoption
Adoption of the intervention occurs once acceptability and feasibility are achieved, and several additional mechanisms are triggered. A belief that using the sepsis screening intervention is legitimate is formed when there is education that is aligned with a clinical team’s need for information sharing (table 4). Regardless of modality, education must include evidence for a need to change practice expressed in terms of the impact on patients and clinicians; knowledge of who, what, where, when and how the intervention will function; characteristics of intervention fidelity and what happens for the patient if fidelity is not maintained. Highly connected interdisciplinary teams benefit from interdisciplinary education through partnership reinforcement. Less connected teams can experience the same benefits through non-integrated education when they are stable (ie, consistent membership) and there is a sense of clear roles, responsibilities and processes. For nurses, both team arrangements create the perception of a sense of legitimate authority and increased confidence in critical thinking.
Achieving fidelity and sustained use
Fidelity and sustained use of the intervention occurs when it is adopted and when trust within the clinical team is achieved. This mid-level mechanism is triggered when strategic implementation and performance feedback is used as part of team problem solving (table 4). This is promoted when a theoretical and evidence-based implementation approach is introduced into a formal team structure by triggering trust in the knowledge and expertise of others and an enhanced sense of self-accountability. Among nurses, mature informal relationships with physicians leads to a belief that physicians understand how nurses see their own roles.
Unsuccessful implementation
Unsuccessful implementation occurs when certain mechanisms are unmitigated, leading to inhibited acceptability, feasibility, fidelity and intervention abandonment.
Inhibited acceptability and feasibility
The inhibition of acceptability and feasibility comes from pessimism about the benefits of a standardised intervention (table 4). Two theoretical resource and context construct pairs can lead to this pessimism: standardised clinical processes implemented across all areas in the context of individually adapted routines to function with heavy workload where standardisation is uncommon; and a clinical process that is adapted for local use across all clinical areas in the context of individually adapted routines to function with heavy workload where standardisation is common. A standardised clinical process differs from standardised evidence-based practice in that it may not be related to an evidence base but instead describes the standardisation of operational clinical processes for who does what, when and how.
For physicians in either context, the reasoning that is triggered involves how they view their practice relative to standardisation and arrive at clinical decisions that can differ from those of nurses. Individual physicians with highly individualised practice (ie, where standardised hospital practices are uncommon) can feel that a standardised sepsis screening protocol is redundant and non-value added. These individuals believe that identifying and responding to suspected sepsis is within their domain of expertise and will be considered in their critical thinking and clinical decision-making regarding a patient presenting with relevant signs and symptoms. These beliefs are amplified if the intervention does not reduce their workload or they perceive that their workload might increase through the generation of unnecessary testing. These beliefs are reinforced when the protocol is used inappropriately thereby compounding dissatisfaction.
Dissatisfaction with a standardised protocol will also occur if it fails to address or amplifies differences in how physicians and nurses from geographically separated teams (ie, when a medical physician cares for patient across multiple different hospital units) arrive at and communicate clinical decisions. In these instances, physicians rely on the communication of objective assessment findings to inform their decision-making when they cannot directly assess the patient. Nurses will often use subjective and objective findings because of the accumulation of multiple mini assessments to inform their clinical impression and decision-making which can be reflected and emphasised when communicating concerns about suspicion of sepsis.
Dissatisfaction with interventions adapted for local areas to the extent that they are significantly different in terms of roles, resources, timing and sequence of action leads to confusion in decision-making and the feeling that the intervention is non-value added. For clinicians who move between different hospital units, intervention design differences force them to critically think, make decisions and respond differently regarding the same phenomenon. For physicians who move between different clinical areas, overly adapted practices increase workload. For nurses, a lack of standardisation removes the reinforcement to develop clinical skills, reduces the sense of authority to follow the requisite actions and promotes infidelity in using the intervention.
Inhibited fidelity and abandonment
Inhibited fidelity and ultimately abandonment comes from pessimism among clinicians about the intervention (table 4). An intervention not aligned with existing workflow in the context of a disconnected team structure because of geography, staff turnover, or process isolation, creates pessimism. This promotes behaviours that decrease fidelity, including decisions to escalate care before prescribed actions are completed or completing actions out of sequence. It also promotes an inability to implement change which creates feelings of powerlessness among clinicians to help their patients. An inability to anticipate daily workflow impedes physicians and nurses from planning and mutual goal setting resulting in an inefficient intervention. Repeated experience of intervention infidelity creates doubt in the legitimacy of the intervention, devaluing it and eventually results in abandonment.
Discussion
A programme theory, establishing how local clinical teams can successfully implement a sepsis screening intervention, is valuable for understanding how to bring the evolving evidence base of screening and treating sepsis into consistent practice. It also translates into implementation approaches of similar clinical interventions. Greenhalgh72 suggested the most important determinant of adoption is a clinician’s belief that an intervention is workable in practice and that a workable intervention simply provides a relative advantage, decreased complexity and trialability. The final programme theory aligns with and expands on these basic characteristics to create parsimonious and purposeful hypotheses for implementation success and suggestions on the resources and functions that could translate into practice.
Actors in this programme theory are physician and nurse end-users of a sepsis screening intervention in medical, surgical and emergency clinical areas. Their beliefs about the positive benefit of an intervention, legitimacy in using the intervention and trust within the clinical team is how and why successful implementation occurs. These promoting mechanisms are tied to key resources including standardised evidence-based practices and implementation activities and education strategies that match team structure and theory-based implementation strategies, respectively. Conversely, their pessimism about the benefits of the intervention and about the team inhibit implementation. These are tied to atypical standardisation and adaptation relative to practice norms and disconnected clinical practice among actors.
Specific contextual circumstances trigger both promoting and inhibiting mechanisms. Actors’ beliefs and trust are triggered by a complex work environment, highly connected or stable teams and formal team structure. Those circumstances that trigger pessimism include individually adapted routines that vary between individuals, variation in the use of standardisation and a disconnected team structure.
Alignment with quality improvement
Successful implementation is essential to clinical quality improvement. Two elements of this programme theory reflect important aspects of the quality improvement approach: an emphasis on measuring fidelity28 70 and use of audit and feedback.73 Fidelity indicates the effectiveness for behaviour change28 because it measures what change is implemented as intended.70 Fidelity demonstrates both adoption and adherence, which are defining endpoints in this programme theory.
Audit and feedback are the means by which information about fidelity is communicated to clinicians. It establishes the meaning and credibility of performance data and prompts reflection, which stimulates cues for action planning.73 In a study of physicians receiving feedback from peers, individuals reached action planning by responding to feedback in a cycle of behaviours that included reaction to data, questioning and understanding data, justifying and contextualising, sharing and reflecting on data and the associated guidelines, and planning for change.54 The programme theory in this study reflects these salient quality improvement characteristics and reinforces its relevance for knowledge users because useable implementation theory has to be conducive to quality improvement.74
Limitations
Engaging with knowledge users using realist concepts and jargon different from typical clinical research terminology results in the loss of meaning when validating theoretical constructs. This was mitigated by using descriptions translated into more common terms as opposed to realist jargon, visual representations of constructs and connections and discussion in data gathering using realist interview techniques to resolve any ambiguity.
The rapid nature of an RRR represents another limitation because the focus is on identifying and referring to key sources of information as opposed to a comprehensive approach. This process could omit a potentially relevant reference, thereby limiting the relevance and robustness of the final programme theory. Although knowledge user involvement is intended to minimise any gaps by relating the evidence to real life, their bias or knowledge deficits might not only fail to address this gap but introduce other limitations. However, in attempting to derive theory-based approaches to local, non-research clinical implementation, stakeholder involvement, a focus on middle-range abstraction and timeliness of results are benefits in the drive to translate knowledge into action.
Conclusions
Describing how to achieve successful implementation of clinical interventions, such as sepsis screening, can be articulated for a particular setting in programme theory form. Such programme theory provides guidance to clinical teams in developing implementation strategies by focusing on salient characteristics that can lead to success. These characteristics include human and environmental factors that complicate implementation. By outlining a theory of causation in a parsimonious way, teams can understand the resources and context that make implementation activities purposeful.
While sepsis screening was the review’s focus, the resulting programme theory is intended to inform the implementation of sepsis screening interventions in other similar settings. The programme theory could also apply to other similar interventions in similar contexts with similar actors. In fact, many of the knowledge users drew from their experience with the implementation of the sepsis trigger tool as well as other interventions they had been involved with in their thinking and answers. Although the programme theory is not a generalisable hypothesis for all clinical implementation interventions, it can serve as a starting point for local teams to consider when planning and implementing clinical interventions. Mechanisms reflective of trust within teams, beliefs influenced by workload and perceived benefit and pessimism are not necessarily dependent on the nature of a specific clinical practice change or one clinical setting. These constructs could easily exist in the implementation of many other clinical practice changes. Similarly, the identified contexts were not necessarily unique to one organisation or hospital. Therefore, this programme theory is potentially applicable to and adaptable for implementation of multiple types of clinical interventions in other local contexts.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by University of Saskatchewan Research Ethics Board (U of S BEH 17-381). Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors This manuscript was prepared from the JM’s master’s thesis. JM was responsible for the concept, data collection, analysis and drafting of the manuscript. TC supported the data collection and analysis and contributed to drafting the manuscript. DG and NM were members of JM’s master’s thesis committee; they contributed to the intellectual development of the review and reviewed manuscript drafts. GG supervised all aspects of the research, including the final draft of the submission.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.