Supporting treatment adherence for children and young people living with Cystic Fibrosis by introducing an adherence protocol in the outpatient setting – A mixed methods implementation study.

Background: In cystic brosis care, patients are required to adhere to a burdensome daily treatment regimen. The outpatient cystic brosis clinic presents an opportunity for multidisciplinary healthcare teams to monitor disease progression, support daily treatment adherence, address adherence challenges and improve health outcomes. However, the complexity of adherence care, and inconsistencies in methods and delivery of adherence interventions may impact the effectiveness of healthcare teams changing adherence outcomes. Comprehensive adherence protocols have been proposed (CF My Way) to increase the visibility, consistency and ecacy of adherence work. Barriers in the clinic environment have been implicated in their unsuccessful translation into standard care. The aim of this study was to explore the feasibility, acceptability and appropriateness of implementing the ‘CF My Way’ protocol into the outpatient cystic brosis clinic of a tertiary, paediatric hospital Methods: This implementation study employed a pre-post observational study design using mixed methods. A four-phased implementation framework guided the implementation plan. Prior to the program introduction, qualitative data was collected from 23 clinicians, adolescents and parents using focus groups and interviews to identify barriers and enablers to implementation. Quantitative data was collected using purpose-designed audit tools and surveys throughout the implementation and analysed to inform modications and support translation to standard care. Discussion: Theoretically and conceptually-derived implementation science frameworks can inform the process of translating an adherence protocol into an existing CF healthcare setting. Results indicate that while using an adherence protocol is acceptable and appropriate according to the perspectives of clinicians and parents, ‘CF My Way’ is not yet an “off the shelf” solution to the adherence challenge. Modications were required to the original adherence protocol to support integration. Implementation strategies such as pilot testing, systems analysis and team cultural interventions are indicated to support effective and consistent adherence care pathways in the CF clinic.

organisational and behavioural adherence approaches and packaged these into a set of tools for use in a CF outpatient clinic environment [17]. The effectiveness of the 'CF My Way' protocol' was evaluated in a clustered 2-arm randomized controlled trial (RCT) involving 607 families across 18 CF health care centres in the United States [17].
The primary outcome measure used to evaluate 'CF My Way' intervention effectiveness was patient medication adherence. Medication adherence was measured according to composite pulmonary medication possession ratios (cMPR) based on pharmacy re lls and calculated from patient pharmacy records. Based on the cMPR results, no signi cant treatment effect was identi ed for medication adherence. However, the authors conceded limitations of the primary outcome measure, the medication performance ratios were inadequate to detect short term changes in medication adherence following intervention. Another limitation of the study was the large amount of missing data, which the authors surmised resulted in the study being underpowered to detect differences in medication adherence [17]. The large amount of missing data was reported as a result of participants being able to choose their adherence intervention target of choice, as per the 'CF My Way' protocol. Medication adherence represents a single aspect of the full CF treatment regimen and in the trial, only 18% of participants selected medication adherence as their intervention target [17]. The majority (58%) of 'CF My Way' participants utilised the sessions to target their adherence to airway clearance, exercise or nutritional treatments, none of which were evaluated in the trial due to the lack of reliable measurement tools [17].
An unanticipated nding of the study, identi ed following retrospective re ection by the trial authors, was that implementation of the 'CF My Way' protocol into the clinics was a key issue. Barriers such as competing clinical demands, clinic space constraints, limited time to conduct the intervention and low clinician attendance at supervision and training sessions were identi ed. These unexpected barriers to implementation led the authors to question the feasibility of implementing a complex adherence intervention into the CF clinic environment [20].
The challenge of implementing and operationalising combined adherence interventions in daily practice within existing clinic structures and processes is an ongoing challenge for health care teams [6,19]. Factors such as the uptake of clinician training, support from other team members, unclear delineation of team roles and team communication have also been identi ed as barriers to implementation [9,11,20]. In the area of adherence, research is needed to determine not only which adherence interventions are effective but also the strategies by which these interventions can be translated into clinical practice outside the academic arena.
The emergence and growth of the eld of implementation science provides researchers and clinicians with a theoretically or conceptually-derived systematic approach to identify factors likely to inhibit or enable successful translation (or implementation) of evidence-based interventions. Implementation science has been de ned as "the scienti c study of methods to promote the systematic uptake of research ndings and other evidence-based practices into routine practice, and, hence, to improve the quality and effectiveness of health services" [21, pg 1]. Understanding and addressing the factors impacting on the effective implementation of 'CF My Way' in the real world has the potential to not only improve adherence to treatment in patients with CF but to change the way practitioners assess and perform adherence work. The aim of this study was to explore the feasibility, acceptability and appropriateness of implementing the 'CF My Way' protocol into an outpatient CF clinic in a paediatric healthcare setting.

Study Design
We employed a pre-post observational study design using mixed methods to evaluate implementation of the 'CF My Way' adherence protocol into existing CF outpatient clinic processes. Using a mixed methods approach allowed for the collection of quantitative data in terms of service use, delivery and adherence to the CF adherence protocol as well as the collection of rich qualitative data to understand the factors likely to inhibit or enable program feasibility, acceptability and appropriateness that ultimately impact on program sustainability. The Replicating Effective Programs (REP) framework [23] was selected to inform the implementation of 'CF My Way'. The study received ethical approval from Children's Health Queensland Hospital and Health Service (CHQ-HHS) Human Research Ethics Committee (HREC) (HREC/18/QCHQ/44458) and administrative ethical approval from University of Queensland HREC (2018002220/HREC/18/QCHQ/44458), and Queensland University of Technology HREC (1800001158/ HREC/18/QCHQ/44458). The study will be reported following the Standards for Reporting Implementation Studies (STaRI) [22] and Standard Protocol Items Recommendations for Interventional Trials (SPIRIT) guidelines.

Study Setting
The study was conducted in the outpatient CF Clinic of a publicly funded, tertiary-level, teaching hospital located in South East Queensland, Australia. This clinic is the primary care provider for all children and young people diagnosed with CF (0-18 years) across the state of Queensland and northern New South Wales in Australia, with approximately 350 patients. The CF clinic team is multidisciplinary and is comprised of respiratory physicians, CF specialist nurses, physiotherapists, dietitians, social workers and occupational therapists.

Inclusion Criteria
A young person with a con rmed diagnosis of CF and receiving care at the study clinic Parents of child or young person with a con rmed diagnosis of CF who receives care at the study clinic Exclusion Criteria Young person receiving care at the study site as an inpatient Young person receiving care by the CF clinical team off site (i.e., not at the outpatient clinic).
For clinicians:

Inclusion Criteria
Clinicians providing care to consumers at the study site clinic during the study period Exclusion Criteria Student trainees of any health discipline

Recruitment and Consent
Recruitment occurred between Sept 2018 to July 2019.
Pre-Implementation: All clinicians in the CF team were approached and invited to participate in focus groups via the existing weekly CF business/team meeting. A purposive sample that included senior leaders, clinical staff, members of each allied health/ medical profession and clinicians of variable levels of experience was sought. The anticipated recruitment numbers for focus groups was 10-12.
Consumers (de ned as young people aged 8-18 years with CF, and their parent/s) were invited to participate in the pre-conditions phase interviews. All eligible consumers were approached while attending their CF clinic and invited to participate in an interview to explore their perspective on adherence interventions in the clinic. Young people and their caregivers were initially approached by a clinical team member not involved in the research project. If interested, the research assistant then provided information on the project and obtained written informed consent. Consumers were informed that the information provided would be utilized for the purpose of improving service delivery as well as for research purposes. The anticipated number of consumer participants was ve to 10 caregivers and young people with CF.
All participants were provided with an information sheet outlining the purpose of the research and their rights and responsibilities when participating in the study. A consent form was provided to all participants by the research assistant at time of recruitment. Clinicians, young people and parents were informed that participation in the research was voluntary, any information provided would be con dential and that they could choose to withdraw at any time.
Post -Implementation: Clinicians and consumers were recruited to complete a questionnaire to assess acceptability and appropriateness of the nal adherence protocol. The same recruitment methods were utilized as the initial interviews / focus groups; however, participation in the pre-implementation phase was not required to participate in the post-implementation phase.
The 'CF My Way' Protocol -The Intervention The 'CF My Way' protocol [19] is made up of ve core components. According to the published protocol, 'CF My Way' can be completed by any member of the clinical team who is trained to deliver the components, face to face, within existing outpatient clinic encounters. The protocol involves provision of three standardised assessment tools assessing the young person's (1) adherence, (2) treatment skills and (3) CF knowledge. The two nal components are (4) the completion of a 15-minute structured problemsolving session and (5) the provision of a written treatment plan outlining all prescribed treatments.

Implementation Plan
This study design was guided by existing implementation science frameworks and current CF literature. These frameworks were selected to support translation of the 'CF My Way' protocol into an Australian CF clinic environment and systematically identify factors that impact implementation. The REP framework was designed to support translation of effective health service interventions into health care and focuses on key stakeholder engagement through all phases of implementation [23]. Embedding the CF adherence protocol into routine service delivery requires a change in attitude and behaviour on the part of the clinicians as well as the young people and their families. Thus, the REP framework is an ideal implementation framework because it focuses on key stakeholder engagement through all phases of introducing of the new model of service delivery. The implementation plan based on the REP phases is outlined in Table 1. Implementation intervention activities have been named in accordance with the taxonomy of implementation strategies described by Powell et al [24] which re ect de nitions compiled by a panel of 71 implementation and clinical experts through a systematic consensus development process. In accordance with the Replicating Effective Programs Framework [23], a local needs assessment was conducted in the pre-conditions phase and the 'CF My Way' protocol was modi ed and adapted (Table 2) in the pre-implementation phase to improve its t for the local context (termed 'Modi ed CF My Way' herein). Annually.

Clinic Communication Tool
Included A document created for this trial. The onepage is used by the clinic nurse/s to facilitate communication between the clinic and families as to their concerns to plan for clinic appointment priorities and action accordingly.
Every visit.
During the implementation phase, the 'Modi ed CF My Way' was piloted across the outpatient CF clinic for three months. All families who accessed the clinic during the trial period received the 'Modi ed CF My Way' protocol as part of standard care. All clinicians working within the clinic were informed of the changes to the modi ed adherence protocol and trained to utilise the protocol components prior to the implementation phase. A two-page educational handout was circulated to the team via email, outlining the newly introduced protocol components and clinic processes, one week prior to implementation. Three presentations to the team occurred within existing meetings in the rst two weeks of implementation. A researcher was present for up to 60 minutes to provide face to face training and problem solving to the clinical team for four clinics across the rst two weeks. Reminders were provided monthly when the audit results were presented.

Outcomes
The outcomes of interest in this study are the conceptually distinct implementation outcomes of feasibility, appropriateness and acceptability, as de ned by Proctor et al [25]. For conceptual clarity we contextualised the outcomes for the purposes of this study as follows: Feasibility is de ned as the extent to which an innovation can be successfully used or carried out within a given setting [25]. As an aspect of feasibility, we also measured delity. Fidelity is the degree to which an innovation was implemented as it was intended by the program developers [25]. Fidelity was evaluated across three areas: adherence to the 'Modi ed CF My Way' protocol, dose of program delivered, and quality of program delivery [25].
Appropriateness is the perceived t, relevance, or compatibility of the innovation for a given place, provider, or patient [25]. Appropriateness evaluates the observed t between the 'Modi ed CF My Way' intervention and the clinic setting. Proctor et al suggest that measuring appropriateness is valuable for picking up resistance or "push-back" to implementation efforts, particularly if the innovation is seen to be inconsistent with the organisation's mission or the individual clinicians' skills, role, or expectation for their job [25].
Acceptability is the perception among key stakeholders that an innovation is agreeable and satisfactory [25] and acceptability outcomes allow the measure of perceived successfulness of the new innovation. Acceptability was based on individual clinicians' and consumers' knowledge of, or experience with, ongoing use of the 'Modi ed CF My Way' protocol and evaluated by assessing their level of satisfaction with various aspects of the protocol such as the content, complexity or comfort [25].

Data Collection and Analysis
Quantitative and qualitative data was collected to support evaluation of the implementation process at the organisation, team, individual provider and consumer levels.

Clinician focus groups
Clinician focus groups conducted during the pre-conditions phase were facilitated by the principal investigator using a structured question guide (Additional File 1). The question guide contained ve questions designed to facilitate clinician communication around current adherence practices, perceived barriers and enablers in the clinic environment to adherence work and clinicians' readiness for change.
Each group was composed of three to six members of the clinical team and ran for an average of 47 minutes (SD = 4.98). All focus groups were audio-recorded and transcribed verbatim; eld notes were also included in this data set. All participants provided written consent to participate.

Consumer interviews
Interview data was collected to determine the consumer perspective on adherence promotion interventions currently used by clinicians and the perceived need for change. Interviews were audiorecorded and transcribed verbatim. The consumer interviews were conducted on site using a six-question semi-structured interview guide (Additional le 2). The average interview length was 38 minutes (SD = 6.64). When recruiting young people, the parent was involved in co-consenting and consulted as to whether they felt the child was capable to participate either alone or with their parent present. By parental consent, all young people who participated were interviewed independently.

Analysis
Thematic analysis [26] was completed on the clinician focus groups and consumer interview transcripts to identify key contextual factors to consider prior to implementation and examine the appropriateness and acceptability of the 'CF My Way' protocol. All transcripts were independently coded by two coders. A third coder was consulted as required to achieve consensus.
Two rounds of coding were completed. The rst round of coding was an 'open coding' round, where inductive, natural codes were identi ed in the transcripts. The second coding round involved the application of the Consolidated Framework for Implementation Research (CFIR) [27] to guide evaluation of the factors in uencing implementation. The CFIR is a widely cited and rigorously developed determinants framework for implementation that was developed through a process of consolidating earlier published implementation literature, including nineteen previously developed frameworks, such as the Greenhalgh [28] dissemination and sustainability of innovations model. The CFIR consists of ve domains, containing 36 key constructs considered most prominent in in uencing program implementation in terms of valence (positive or negative in uence on implementation) and strength (strong or weak in uence on implementation) [27]. Two researchers (BR, MS) rated the predicted valence of each factor as identi ed in the qualitative analysis [29]. Consensus was reached via discussion.
As the interviews explored the local context prior to the proposal of the 'CF My Way' protocol the CFIR domains of 'Intervention Characteristics' and 'Process' were not applicable as these explore aspects of the innovation and implementation plan which were not discussed.

Medical record auditing
Throughout the implementation phase, audit and feedback was conducted as described by Ivers et al [30]. Medical records of patients who attended the outpatient CF clinic for their annual review appointment were audited to assess clinician delity to the outlined protocol [25]. Tailored checklists were used to compare the care provided with the core components outlined in the 'Modi ed CF My Way' protocol. Results were compiled monthly [30] to facilitate regular feedback to the clinical team. Feedback was presented to the clinical team via existing meetings as a month-to-month comparison. From this, target behaviours to improve implementation were identi ed with the clinical team and target goals created for the following audit cycle [30]. Fidelity use was quanti ed as a percentage (i.e., number of protocol components completed and documented compared to the total number of applicable protocol components).

Technical assistance log
The technical assistance log was a record of all key implementation events and/ or external events that may have impacted the study, as well as all formal and informal feedback received from stakeholders throughout the process. This log was recorded by the principal investigator throughout the preimplementation and implementation phase. Any modi cations to the protocol or implementation plan including further consultation activities, top-up training and alterations to processes were recorded in the technical assistance logs. The descriptive data from the technical assistance logs was reviewed at project completion alongside the quantitative and qualitative data to provide a narrative description of the implementation process and observe key activities that may have impacted implementation success. This data was compared with audit data to determine the overall feasibility of the 'Modi ed CF My Way' protocol in the outpatient CF clinic.

Clinician and parent surveys
Structured questionnaires were provided to clinicians and consumers face-to-face at the CF clinic.
Questionnaires were distributed at week six and in the nal week (week 12) of the implementation period to assess acceptability and appropriateness of the adherence protocol. The questionnaires outlined each component of the protocol and assessed familiarity and perceived value and t using Likert scales and open comments.
Clinician and consumer comments on the written surveys were collated and analysed using content analysis to determine key themes relating to acceptability and appropriateness of using the 'Modi ed CF My Way' protocol as standard care. The CFIR framework was applied to collate qualitative data into implementation factors. Quantitative data was analysed using descriptive statistics.

Digital assessment component
Digital screening results were reviewed at 6 months post implementation and data collected on the number of screens completed, the percentage of young people identi ed as "at risk" and the percentage who received follow up care were evaluated using descriptive statistics.
All ndings collected, including data from quantitative and qualitative sources were triangulated by the research team (BR, SO and MS) using the 'following the thread' methodology [31] to identify key factors and implications for future practice and research.

Pre-Implementation results
Thirteen clinicians (92% female) participated in the pre-implementation focus groups. All disciplines within the CF clinical team were represented (three CF clinical nurse consultants, one nurse practitioner, three physiotherapists, one occupational therapist, two respiratory physicians, one social worker and two dietitians). Consumer interviews were completed by eight parents (87.5% mothers) and two adolescents (50% female).

Adherence Assessment
Both clinicians and consumers reported that adherence was discussed by a member of the clinical team at each clinic appointment. Conversations about adherence were reported to be consistently initiated by a member of the clinical team and assessment of adherence was based on child, adolescent or parent selfreport. Parents reported the experience of repetitiveness around adherence assessment with multiple team members discussing adherence in the same clinic appointment as a challenge of adherence assessment in the clinic. One parent reported, "Okay, so I don't mind everyone coming in and asking the same question. But I do for him [son], he gets frustrated because its different people asking exactly the same question constantly to him, so by the time you get to the last person, he's over it." (Parent interview 6)

Adherence Interventions
Adherence interventions were reported by clinicians and parents to take place in a variety of settings (CF outpatient clinic, home visits, additional outpatient appointmentstelephone, email and telehealth).
Clinicians discussed using a range of adherence interventions in the CF clinic including; educational interventions, goal setting, encouragement, coaching, creation of reminder systems, supporting families to create daily treatment routines and providing ideas or options to modify how treatments are completed. Some clinicians discussed that if they had concerns about a child or young person's adherence, they may increase the frequency of follow up to allow greater monitoring.
In contrast, parents and young people interviewed were able to identify a smaller number of perceived adherence interventions. The main adherence intervention that they could identify was that clinicians generated new, novel or alternate ideas and tips for how treatments could be completed. Family participants did not recall other coaching, goal setting or habit interventions as recognisable adherence interventions or did not separate this from the main intervention of "doing things differently". Parents discussed that clinic clinicians' conversations about treatment expectations in clinic could support adherence at home as parents could refer to members of the clinical team an authority to their child, thereby supporting the need for treatment to be completed. Overall, parents reported a positive perception of their interactions with the clinical team. All reported that the CF team typically presented a consistent message regarding adherence to treatments. Both clinicians and parents discussed that the education about CF and treatments was provided in depth at diagnosis, however, age-appropriate education became less frequent, available and well-resourced as the child aged.
Parents expressed a desire for easily accessible, up-to-date CF education resources to be available throughout their child's lifespan. Parents reported that they sought support and education through external channels beyond the CF clinic, most commonly from other CF parents and members of CF communities. External support and education were typically sought via online social networks or from alternate health care providers (e.g. Facebook pages, chiropractor, GP, local hospital).

Factors impacting adherence assessment and intervention
Factors within the clinic structures and processes that were likely to impact the implementation of a new adherence protocol were identi ed and mapped against the CFIR domains [27]. Three additional implementation factors, not included in the CIFR framework, were perceived to further impact the provision of adherence care and parental adherence decision making: clinician-family relationships, paternalism, and parental decision making. All identi ed factors are outlined in Table 3 and were used to inform the modi cations made to the 'Modi ed CF My Way' protocol. The predicted valence of each construct is represented with a (+) positive, (-) negative or (U) unclear impact on the adherence protocol implementation into the local clinic [29]. ** 'Paternalism' is de ned a set of attitudes and practices where the clinician/s is guided by their perception of what is best for the family when making decisions about a child/ young person's treatment regime, without involving the family in the decision-making process. [32] As outlined in Table 2, four components of the original protocol were deemed suitable by the research team based on t with the setting and end user priorities and consideration of research support for the component. The Treatment Adherence Questionnaire-CF component of 'CF My Way' was not included in the implementation plan as data from the focus groups and interviews identi ed that asking about adherence was a consistent facet of standard care. The four components included were assessment of treatment skills, CF knowledge assessment, the written treatment plan and problem-solving intervention.
It was identi ed that the component 'assessment of treatment skills' was met by the existing practice of physiotherapy and nursing annual review processes in the CF outpatient clinic, therefore it did not need to be included in the implementation planning. The CF knowledge assessment and written treatment plan components required minor modi cations (i.e., integrating with medical record systems, updating to Australian medication names) to support translation into the local context.
Use of the problem-solving intervention was considered a central tenant of the adherence protocol by the research team, however as it was not identi ed as a priority during pre-implementation interviews. A phased approach to implementation rollout was utilised, whereby intervention components were introduced progressively [33]. The elements with a high level of interest and perceived need were introduced rst to support clinician uptake and allow the clinical team to adjust to these changes rst before the introduction of the problem-solving intervention requiring a more signi cant shift in adherence practices. The research team developed an electronic learning program to support train staff members in the problem-solving intervention, anticipated to be rolled out post the initial three-month implementation phase. A digital platform was selected to support sustainability as clinician rotations were identi ed as a key implementation factor in pre-implementation. In summary, the 'Modi ed CF My Way' protocol included two standardised assessments (knowledge assessment and mental health screening), provision of a written treatment plan to the family and the clinic communication tool (for use in outpatient CF clinic).

Implementation Effectiveness
During the three-month implementation phase, 359 outpatient appointments were completed within the cystic brosis clinic. Thirty-ve families attended their annual review appointment and were therefore eligible to receive the full 'Modi ed CF My Way' protocol.
For a clinic of this size, every month 8% of the clinic population should undertake an annual review to ensure that all patients are reviewed in a yearly cycle. As a result of the auditing process it was uncovered that the number of eligible consumers was signi cantly lower than expected. In the rst month of the study, only 2% of consumers were booked for annual review appointments. This resulted in a lower number of eligible recipients for the modi ed adherence protocol and highlighted a barrier to universal adherence care.
As a result of this auditing process, tailored implementation strategies were launched to improve monthly annual review numbers. An implementation group (consisting of CF nurses and later allied health representatives) was initiated to explore barriers to annual review processes, support booking processes and revise the multi-disciplinary team processes which resulted in low booking numbers. By the third month of the implementation period, 5.7% of the clinic appointments were identi ed for annual review.
Medical chart audits demonstrated that the adherence protocol was being used with 43.8% (knowledge assessment) to 57.1% (young person mental health screening) of eligible consumers by the third month of implementation. Rates of use over the three-month period are outlined in Fig. 1. As multiple team members were required to complete the written treatment plan, differences were noted between the commencement and completion rates. Commencement rates averaged 65% in the nal month of the study, however completion rates were lower at 45%.

Post Implementation Survey results
Data was collected from eight members of the clinical team at the midway questionnaire. Data was collected from ten clinicians and ve parents via the post-implementation surveys during the nal week of the implementation phase (week 12). Qualitative data from the surveys were compiled alongside researcher notes in the technical assistance log to identify the key factors that impacted implementation during the three-month period. The key implementation factors identi ed are outlined in Table 4 were aware of the knowledge assessment and 90% aware of the mental health screening.

Screening results
The impact of screening was retrospectively evaluated after the rst six months of use. Digital screening assessments were completed with 135 young people, adolescents or parents. The most utilized assessment tool was the parent mental health screening (N = 80, representing 26.2% of clinic population).
These tools were used to identify young people and parents presenting with 'red ags' (such as depression and/or anxiety) that may negatively impact adherence as well as other health outcomes.
Screening identi ed 41.67% (n = 10) of consumers that were assessed (12 years+) required further input to support their mental health and support ongoing engagement in treatments. Of the eighty parents assessed, 18.75% (n = 15) were identi ed as requiring action due to their scores on the depression or anxiety screening. Only young people over 11 years were identi ed as having sub-optimal understanding through the knowledge assessment, which detected 15% (n = 20) of young people who scored less than 50% on this assessment. Of the consumers identi ed as 'at risk', 85.71% (n = 24) received follow up care (such as social work review or referral to further services) to address the concerns identi ed by the use of screening tools.

Discussion
This is the rst study using a theoretically and conceptually-derived implementation strategy, underpinned by implementation science, to introduce an adherence protocol into a CF outpatient clinic. Through a rich exploration of the structures and processes of the outpatient CF clinic, factors that impact adherence care provision and in uence whether clinics are able to change towards more uniform approach to adherence management were uncovered. The 'CF My Way' protocol was designed to embed adherence interventions into outpatient CF clinic encounters however, this protocol required extensive modi cation and planning to support compatibility to the local environment in this study. Audit results showed that the modi ed protocol components were used with up to 65% of eligible families, with the use of a screening tool and commencement of a written treatment plan the most successfully implemented components. Whilst both clinicians and parents indicated via survey that they considered the Modi ed CF My Way protocol both appropriate and acceptable, the rates of clinician use suggest inconsistent clinical utilization over the three-month observation period. These ndings, like the original 'CF My Way' study, bring into question the feasibility of integrating an adherence protocol into standard care in an outpatient setting [17]. Factors that likely in uenced uptake were identi ed and, if addressed, may allow greater implementation success in future studies.
Clinical teams in the CF clinic have become more multi-disciplinary to deliver high quality, family-centred care. As the clinical team has expanded, there has become a greater need for strong coordination and communication networks to integrate care as each team member is likely to address their discipline speci c piece of the adherence puzzle [36,37]. Adequate team coordination is required to avoid the pitfalls of the "too many cooks in the kitchen" effect where clinics become repetitive, unhelpful or ine cient [36,37]. Results from this study support that clinicians experience an overlap of professional roles regarding adherence and poor visibility of what other members of the team are doing with the family regarding adherence. Without clear allocation of roles, many clinicians expressed concerns about the risks of providing contrary recommendations to families or overloading the young person or family with too many recommendations or expectations. This was con rmed by parent and adolescent participants who described care provision that was often repetitive, especially in the area of adherence assessment, and acknowledged the negative impact of receiving different advice on sustaining their child's care at home. This phenomenon has been described in multidisciplinary clinic care more widely, where overlap of professional roles can both be a strength when the team reinforces the health message but also lead to confusion for families when recommendations vary or change according to practitioner, leaving parents to decide how to interpret the recommendations themselves [36,37,38]. The level of compatibility between the clinic work ows and systems and the new protocol was identi ed by clinicians as impacting the protocol's acceptability and appropriateness. The highest number of modi cations were required to the written treatment plan, due to its digitization and need for multiple team members to contribute their recommendations. Digitization offered bene ts in terms of data storage, accessibility and team collaboration. However, it also introduced a higher training requirement as the digital systems were not familiar to the large healthcare team (as existing systems were not suitable to serve the requirements of the protocol components) [33]. Clinicians identi ed that the components added additional paperwork and time burden. The technical logs displayed ongoing modi cations were underway until the last week of the implementation period, as a result of organizational efforts to address clinical governance. As a result, changes to the originally launched protocol were necessary throughout the 3-month implementation period which likely impacted upon the quality of training and clinician's understanding of the digital components of the protocol. It is recommended that a higher training budget be considered when introducing digital adherence resources as well as use of adequate pilot testing to ensure compatibility with local systems [33]. The bene ts of digital transformations often take longer timeframes to be realised due to both training requirements and their increased value when the data collected can be harnessed for other purposes [33].
An unexpected nding of medical chart audits was that the number of monthly annual reviews relative to the total clinic population was lower than anticipated by clinicians. In effect, there were low numbers of eligible consumers for the 'Modi ed CF My Way' protocol as four of the ve core components were linked with this timepoint. Despite annual review being considered standard care [39], it was identi ed that consistent processes were not well established or understood by the team. Subsequently, the annual review processes were reviewed and formalized to support compatibility between a consistent structured adherence protocol and a similarly consistent annual review structure. As annual review numbers increased signi cantly between month one and three, this may have impacted the adherence protocol completion rates as clinicians adjusted to greater number of eligible patients per clinic than previously experienced. Systems analysis practices could be utilised to observe how clinic processes work prior to implementing an adherence protocol to ensure underlying process barriers are addressed before introducing new processes [40]. In this study, use of systems analysis in the pre-condition phase of planning may have uncovered the low annual review numbers. When adherence protocol processes are designed to be linked with clinic ow processes, it is imperative that these underlying systems are optimised.
Future teams implementing the 'CF My Way' protocol should consider modi cation, adaptation, and implementation strategies to increase compatibility with CF outpatient clinics. Using a small clinic pilot group is recommended to allow implementation teams to re ne the protocol and troubleshoot issues prior to scaling up. As evidenced by this study and the original randomized control trial [17], signi cant modi cation and localization is required to align within clinic adherence protocols with existing clinic work ows. Initial modi cation alone did not su ciently support compatibility. Using a pilot clinic with a small group of patients/ clinicians would facilitate 'on the ground' learnings to inform changes to the protocol before scaling up. This would also reduce the need for clinician retraining.
An unexpected nding of this study was the signi cant impact of culture on adherence care provision. The pre-implementation results brought organisational culture and underlying clinician beliefs into view.
In implementation literature, culture has been identi ed to play a signi cant role in organisational change [41,42]. At the time of this study, the approach to adherence management by the CF team was typi ed by high levels of decentralization and exibility for clinicians. Adherence care provision was highly oriented towards the users (i.e., parents and young people) of the service, rather than explicit rules or formal processes. This is characteristic of a "humanistic" organizational culture [41]. Humanistic organisations rely on shared values and allow employees a high level of individual decision making to govern behaviour and place limited structures in place to control these processes. It is considered that this culture impacted upon adherence protocol implementation in this study in two distinct ways. Firstly, as the intervention was considered family-centred (according to post implementation clinician survey results), this positively impacted clinician acceptability of the protocol. Secondly, as adherence care falls to a large group of clinicians within a humanistic, decentralized workplace culture, it was not surprising that underlying process gaps and variability in practices were identi ed in the local setting when the modi ed adherence protocol was introduced. Maintaining a balanced culture is perceived optimal, which supports both exibility and structure as well as greater consistency of care [42].
Clinician's beliefs and the underlying workplace ethos around adherence were uncovered to be an important factor in understanding CF adherence practices. Clinicians discussed in the pre-implementation focus groups that they held an underlying belief that "adherence" is an unattainable target. Team members reported that they believed prescribed treatment plans are unrealistic and place a high level of burden on families. The way that clinicians deliver family-centred care was in uenced by this belief.
Clinicians reported they act in a protective way to reduce overwhelm and change their adherence targets and recommendations according to their perception of the family's capacity to manage. In contrast, parents reported that they would prefer their team to equally discuss more treatments options with them rather than assuming the family's burden. Some parents spoke to a feeling of having to ask for additional treatments that they had read about online or feeling their care team was holding back in their recommendations.
When teams rely on a humanistic culture, equity of care may be compromised by the paternalistic instincts and best intentions of clinicians. Where clear organisational policy or structure is not available to guide clinician's delivery of care, a high level of exibility and variation in the provision of care is likely to be observed. Although this places families at the fore, it raises a question of whether clinicians and families differ in their interpretation of family centred care and whether the introduction of systems that outline basic roles and responsibilities around adherence would increase the equity and predictability of multidisciplinary care.
An opportunity is presented here for families to be engaged actively and openly in co-designing their optimal treatment plan. Consumers who are well engaged are more likely to seek help when needed, ask for and follow advice, are more satis ed with their care and have improved adherence and health outcomes [43]. Parent participants discussed a desire to be an active participant in care planning as well as in planning how they use their clinic appointment (e.g., requesting multidisciplinary team members as their needs indicated). From this study, it would appear that parents, particularly parents of teenagers, want more opportunities to be a key member in treatment decisions and have insight into which clinicians they wish to see in their clinic appointment to support their care.
Limitations of the study should be noted. The purposive recruitment of clinicians and families to the study may have introduced respondent bias, as interested parties were more likely to allocate time to discussing the subject matter of the study. Therefore, views of stakeholders with low interest or investment in adherence service redesign may be under-represented. Also, the implementation of the problem-solving component of the original protocol was not observed in this study due to considerations of the number of sustainable changes that could be managed and availability of training materials. However, following service evaluation using the key factors outlined in this paper, the implementation of a collaborative problem-solving intervention is recommended.
In summary, adherence protocols show promise in their ability to de ne professional's roles and care pathways in an area of practice that has previously been relatively unde ned. Currently the 'CF My Way' protocol and the modi ed version observed in this study are not "off the shelf" solutions to this problem. However, lessons can be learnt from observing their implementation using an implementation science approach. When improving adherence care, attention should be directed at strategies that support effective interprofessional collaboration to address the underlying fragmentation and role overlaps in adherence care. Further exploration is indicated regarding how multi-disciplinary teams can collaborate or utilize care coordinators to increase integration of adherence care and clarity of treatment recommendations for families.

Page 41/47
The use of objective assessment data to direct adherence care also showed potential in this study. The screening of adherence related factors and inclusion of family preferences in appointment planning were observed to assist clinicians to allocate clinical time and support families demonstrating risk of adherence deterioration, rather than clinicians relying on intuition. Objective screening data may support more effective allocation of resources and prioritization of patients requiring timely adherence intervention. For example, the use of preclinic family communication or screening (via digital automation) could be used to enhance pre-appointment planning. This may be especially valuable for teenagers with CF and their families where clinic reviews are seen to add less value [44] and where families are more con dent and comfortable to manage how they use their appointment. Bringing the child, adolescent or their family into care planning decisions would represent a directional shift away from traditional paternalistic approaches to adherence management towards family demand driven care [45,46]. Further research should explore pathways for family data (including communication of their preferences) to have a more active role in directing the allocation of health professional time in the CF clinic. Predictive analytics using objective adherence data could also be used to streamline timing of review appointments or target group programs or resources towards families with similar needs to increase e ciency of health care.

Conclusion
Translating newly designed clinical tools into existing, complex health care settings can be challenging.
Research is needed to determine not only which adherence interventions are effective but also to evaluate whether these interventions can be translated into clinical practice within real clinical contexts [47] -and how best to do this.
An understanding of the factors that characterize the CF outpatient clinic enables clinical teams to better understand the context in which adherence changes are often recommended. This includes the less obvious cultural and systemic factors that may in uence practice and inform the design of protocols that can improve care delivered by the service. Results of this study support the use of communication strategies such as using implementation teams with multidisciplinary representation to optimize team awareness and engagement. Enhancing local compatibility via smaller pilot trials and systems review prior to implementation and exploring automation are suggested to reduce the time impact of adding additional tools into the clinic. Lastly, it must be understood that adherence work is directed by systems and processes as well as clinician beliefs and organisational culture. Notably, it was identi ed that clinician beliefs and organisation culture may be at odds with client values and desires/wishes, as identi ed in this study. This nding warrants further exploration to understand the impact that clinician beliefs about adherence have on provision of care and communication styles. Implementation factors should be considered alongside adherence process changes to detect the in uence of these factors on delivery of adherence care. Abbreviations