Background
The clinical problem
Cystic fibrosis (CF) is the commonest inherited, life-limiting disease in white populations. In 2018, 222 new cases of CF were diagnosed, totalling 10,509 people with cystic fibrosis (PWCF) in the UK, with a median age at diagnosis of 2 months.1 Although CF is a multisystem disorder, the upper and lower airways and digestive system are more commonly affected.2,3 Gene mutations lead to clinical manifestation of the disease whereby secretions in the lungs become congealed and sticky.4 In > 80% of cases, PWCF experience respiratory failure due to lung damage.5 In the UK, the median predicted survival for a child born with CF is currently 47.3 years.1
Cystic fibrosis management
Preventative medications both preserve lung function and reduce pulmonary exacerbations.6–12 Typical treatments for CF target airway clearance, reduction in bacterial infection and inflammation. Patients self-manage their disease through a complex, time-consuming daily regimen of treatments that include inhaled therapies delivered via a nebuliser.13,14 As with other long-term conditions, adherence levels to medication are low. Estimates by the World Health Organization suggest that 30–50% of prescribed medicines are not taken as intended.15 In CF, observed objective adherence rates to nebulised therapies are 36% in adults16 and 67% in children,17 despite self-reported rates of adherence of 80%.16 Treatments work only if they are taken; poor adherence predicts periods of exacerbation requiring intravenous antibiotics (IVAB).18,19 Exacerbations carry both the burden of systemic side effects and significant mortality;20,21 frequent exacerbations are linked with accelerated decline in lung function and increased 3-year risk of lung transplant or death.22
The cost of non-adherence
The cost of pulmonary exacerbations is high. The total UK spend in 2012 for CF was estimated to be £100 million, £30 million of which was spent on inhaled antibiotics and mucolytics (Paul McManus, Lead Pharmacist, South Yorkshire and Bassetlaw Area Team, NHS England, 2015, personal communication). In the UK, CF patients received 171,907 days of IVAB, with 93,455 days occurring in hospital, at an estimated cost of £27 million (Stephanie McNeil, Chief Statistician, UK Cystic Fibrosis Registry, 2016, personal communication). Medicines possession ratio (MPR) data suggests that PWCF with good adherence (MPR > 80%) have lower health-care costs than PWCF with poor adherence (MPR < 50%). Hospital admissions for IVAB are responsible for most of the excess costs.23,24 At the heart of enabling preventative therapy is the task of making adherence visible and allowing this key metric to be available to both PWCF and their clinical teams to move from rescue to prevention.
Potential for cost savings
In the early stages of the programme grant reported here, Tappenden et al.24 conducted an early, model-based cost–utility analysis in adults with CF with chronic Pseudomonas aeruginosa infection in a project funded separately from the programme grant. Therapies included nebulised or dry powder inhaled antibiotics prescribed by the NHS and Personal Social Services (PSS) over a lifetime horizon. The analysis suggested that an intervention such as ‘CFHealthHub’ had the potential to generate cost savings.
Aims
We developed a complex intervention with several components25 to support adherence to medication for PWCF, in line with identified research priorities.26,27
Objectives
Our specific objectives map to three overarching work packages (WPs):
- to develop a mechanism for objective measurement of adherence through (1) development of a data-capture and transfer infrastructure that could collect time- and date-stamped data and (2) display this data both on a CFHealthHub website, for use by both PWCF and their clinicians, and on a CFHealthHub mobile application (app), for PWCF, and (3) develop the CFHealthHub web interface with patients and clinicians
- to develop an evidence-based behaviour change intervention (BCI) to increase adherence to nebulised CF medications that works synonymously with the CFHealthHub digital platform
- to evaluate the CFHealthHub intervention (digital platform plus BCI) in terms of (1) clinical effectiveness, (2) acceptability and (3) cost-effectiveness and to examine the processes that drive these outcomes.
These objectives were met through three WPs from intervention development to evaluation, details of which are outlined in Figure 1.
FIGURE 1 Work packages in the research programme. RCT, randomised controlled trial. (Figure not reproduced here; see the original article.)
Key changes to the original programme
Interventionists
We adapted our initial plans to employ physiotherapists as interventionists because, early in the programme, CF clinicians identified that it was challenging to recruit physiotherapists because of workforce shortages. Intervention training was adapted to ensure that a range of health-care professionals could deliver the intervention competently through the use of tests and certification. Consequently, any member of the multidisciplinary team (MDT) could be trained to deliver the CFHealthHub intervention. Additional funding from NHS England allowed us to increase the amount of interventionist time from 0.8 to 1.0 whole time equivalent at randomised controlled trial (RCT) sites in England.
Primary outcome
The definition of the primary outcome – number of exacerbations – was changed from the standard Fuchs criteria (IVAB plus 4/10 symptoms)28 to the modified Fuchs criteria (IVAB plus 1/10 symptoms) to increase the sensitivity of the primary outcome, in line with other, recent CF research.29
Number of centres and treatment costs
As a way of paying treatment costs, the NHS England patient activation quality improvement scheme authorised the CF Commissioning for Quality and Innovation (CQUIN). The CQUIN was offered to all 26 adult CF centres in the UK, adoption of which was a means of engaging with quality improvement by delivering the full-scale RCT in our programme. A total of 19 sites participated in the RCT, nine of which were funded by the CQUIN and 10 by a combination of local and central Department of Health and Social Care commissioning.
Fidelity
Fidelity checklists together with audio-recording sessions were used to assess treatment fidelity. CFHealthHub also records exactly what intervention components have been prescribed at each session, and the extent of participants’ independent engagement with the intervention. Interventionists were trained to audio-record all intervention delivery sessions of which a purposive sample were assessed based on the type and time point (phase) of the session administered. Data were analysed and rated for fidelity by two independent assessors for inter-rater reliability.
Changes to work packages
The WPs completed are shown in Figure 1. Changes to WPs included:
- Combining distinct WPs together – WP 2.1B was completed as part of integrating the findings of WP 2.1A with WP 1 to develop the CFHealthHub portal.
- Creating a smaller, distinct WP [WP 2.2B(1)] within WP 2.2B to allow us to recruit an initial set of participants to start refining the intervention.
Patient and public involvement
Active involvement of the patient and public involvement (PPI) panel was integral to the design and implementation of key aspects of the ACtiF (Development and evaluation of an intervention to support Adherence to treatment in adults with Cystic Fibrosis) programme. PPI was embedded in the early intervention development phase, the feasibility and RCT phases and the dissemination of the programme findings. As outlined by Shippee et al.,30 patient and service user involvement was considered in each of the three phases of our research: preparatory, execution and translational.
Specific objectives were to:
- seek iterative cycles of advice and feedback from PWCF and the wider clinical community during the development of the CFHealthHub intervention, including from our PPI panel
- receive guidance on key research procedures and materials used in the feasibility study and full-scale RCT
- contribute to discussions on a dissemination strategy, including the most effective methods of communicating programme findings to the CF community.
Service user initiation
Dan Beever, a co-applicant on the ACtiF programme, is a PWCF and was the PPI panel chairperson throughout the programme. Dan Beever was involved in providing patient input to research design at the bid stage. The remaining panel consisted of five PWCF and two parents of PWCF. The initial PPI panel consisted of PWCF members from the Sheffield CF centre who were involved during the intervention development phase. To establish a more diverse panel, membership was extended to PWCF from across the country. A leaflet campaign was used to recruit patients, carers and other people with an interest in CF through sites involved in WP 2.1C. In addition, the opportunity was advertised on the People in Research website (URL: www.peopleinresearch.org/; accessed 12 May 2021), and contact made with some individuals known to Dan Beever. Roles and responsibilities of the group were discussed at the first meeting. During the course of the programme, the study manager provided regular updates on programme progress to allow co-learning30 and maintain engagement.
Development phase (work packages 1.1 and 2.2)
A number of frameworks and approaches that were used during the intervention development phase necessitated input from the PPI panel. Use of the agile software development approach required iterative cycles of development and feedback from both the research team and the PPI panel. This included giving feedback on the proposed input of the educational content, the CFHealthHub landing page and demo versions of the CFHealthHub website and giving advice on practical issues of sharing data on the platform. The PPI panel ensured that the CFHealthHub user guide was accessible and coherent. Frameworks to guide decision-making and prioritisation of work implementation in this process required PPI input to understand which tasks were the most important and relevant to PWCF.
Our PPI panel also contributed to discussions on the themes identified in the qualitative research conducted in WP 2.1 to provide context and additional insights. Feedback on the content and nature of PPI meetings was continuously sought from the panel. Alterations were made to agenda items to include agenda items that the panel wished to discuss, for example ongoing progress.
Feasibility and randomised controlled trial phases (work packages 3.1 and 3.2)
The PPI panel advised on research procedures, in preparation for both the feasibility study and the full-scale RCT. This included input on the best way to contact participants to arrange consent visits and collect prescription data as well as discussions on the most appropriate terminology and communicative approaches to use during intervention sessions and on the CFHealthHub digital platform. The panel advised and provided input on the essential trial documents and their administration with participants. One important element was testing the participant-administered questionnaires, with feedback on completion times and appropriateness of both questionnaire and text message wordings to participants.
Dissemination phase
The PPI panel advised the research team on dissemination strategies across the clinical community and participants. Co-learning facilitated sharing knowledge in a multidirectional way between the study management team and the PPI panel. The RCT findings were presented to the PPI panel members on the Trial Steering Committee (TSC), who provided input on the way in which these findings should be presented to PWCF. One of the PPI panel members with expertise in managing communications assisted with the development of a formal dissemination strategy for feedback of results to multiple stakeholders in multiple ways. The PPI panel agreed the dissemination plan.
Challenges and successes
We had regular contact with the PPI panel but input was limited by the inability of the group to attend face-to-face meetings because of infection control requirements. Visual conferencing facilities were a challenge to use in practice and this was not pursued owing to technical issues. Recruitment of the PPI panel was a challenge but we were able to expand recruitment from the initial Sheffield CF centre to several other CF centres through our leaflet campaign. Involvement was much more frequent in the earlier intervention development phase than during the evaluation phase. Two members of the PPI panel also became members of the TSC and, as such, were able to provide oversight to the programme as a whole.
One of the key successes of our PPI panel was their continued input and engagement, which was sustained throughout the duration of the programme. This is in part a result of having a patient co-applicant on the ACtiF grant who acted as the PPI panel chairperson. Dan Beever’s contributions were of particular value in the recruitment of the PPI panel, as well as in maintaining the interest and engagement of the panel throughout the duration of the programme. A short survey was distributed to members of the panel at the end of the RCT phase to seek their views on their involvement during the course of programme. Three completed responses were received, all reflecting positively on involvement, particularly in terms of contributions being valued by the research team.
Specific impacts
- When PPI panel members used the prototype website they found it difficult to locate appropriate patient videos, so suggested that these be tagged with labels for ease of use (change recommended 14 September 2016). This change was made to the intervention.
- We asked for advice from the panel around contacting potential RCT participants to arrange consent visits. The group recommended sending a text message to the potential participant the day before the planned telephone contact to prepare them, followed by a telephone call the next day, leaving a voicemail message if there was no answer (discussions held 23 June 2017 and 2 October 2017). This approach was used.
- A guide was produced for people using the intervention. The panel recommended a number of changes, for example referring to it as a ‘guide’ rather than as a ‘how to guide’ and being clear that the ‘problem-solving’ related to solving problems with adherence and not with using the intervention (changes recommended 13 November 2017). These changes were made to documentation.