Effective maNagement of depression among patients witH cANCEr (ENHANCE): A protocol for a qualtiative study on stakeholder perspectives of Network Meta-Analysis findings on the most effective treatments for depression [version 1; peer review: 1 approved]

Depression is very common among patients with cancer, yet many patients with depression do not receive treatments for depression that may help them. The best available evidence for the management of depression among patients with cancer comes from systematic reviews of randomised controlled trials (RCTs) that use network-meta analysis (NMA) to compare different treatment approaches and rank these in terms of effectiveness. However, patient and healthcare provider stakeholders should be involved in interpreting this evidence and determining how it can best be put into practice to meet patients’ needs. The aim of the current study is to explore stakeholders’ views of the best available evidence on the management of depression among patients with cancer, and their views on the implementation of this evidence in Ireland.


Introduction
One in four patients with cancer experiences significant symptoms of depression, a five-fold increased odds compared to the general population (Hartung et al., 2017).Depression is associated with lower treatment participation, less satisfaction with care, poorer quality of life, greater symptom burden, higher healthcare costs and mortality (Smith, 2015;Walker et al., 2014).Nevertheless, most outpatients with major depression do not receive interventions for depression (Holland, 2013;Walker et al., 2014).
The National Cancer Strategy (NCS) 2017-2026 highlights the deficiency of psycho-oncology services in Ireland and proposes a stepped care model where interventions (ranging from information provision to psychiatric care) are allocated based on patients' level of distress and morbidity (Department of Health, 2017).Incorporating information on intervention effectiveness and acceptability would enhance this model.However, data from direct comparisons between depression interventions for people with cancer are currently unavailable.Furthermore, there is no explicit consideration in the strategy of patients' perceptions of these treatment options or their efficacy.Indeed, the model focuses on the type of care provider appropriate for a given level of distress rather than the type of intervention that would benefit patients most.
Our research group is currently carrying out a systematic review and Network Meta-Analysis (NMA) to determine which treatments (including pharmacological, psychotherapy, combined care, collaborative care, exercise, and complementary and alternative medicine interventions) for depression are the most effective and acceptable to patients with cancer (Pertl et al., Submitted for publication).By using NMA we can compare the effectiveness of interventions that have not been compared head-to-head in previous research and can even rank all available interventions from most effective to least effective.NMA provides the highest level of evidence for developing treatment guidelines (Leucht et al., 2016), supplying patients, clinicians and policy-makers with vital information for informed decision-making.The aim of the current study is to present the evidence from this NMA to key stakeholders during focus groups to obtain their views of the findings and to discuss on how they should be applied in clinical practice in Ireland to best meet patients' and service needs.
This focus group study addresses growing evidence of gaps between evidence-based medicine and clinical practice, and recognition that global data do not always transfer to local circumstances (Djulbegovic, 2014;Hay et al., 2008).Local stakeholder input identifies factors affecting intervention effectiveness not addressed in randomized controlled trials (RCTs), and can help to determine whether interventions can actually work in the "real world".Involving stakeholders in interpreting and applying best evidence is critical for developing guidelines.Key stakeholders for our project include patients with cancer who have experienced depression and hospital-and community-based healthcare professionals, who may have different perspectives on managing depression among patients with cancer.

Study aims and research questions
The aim of this study is to explore stakeholders' (i.e., patients' and hospital-and community-based healthcare professionals') perceptions regarding the evidence for the management of depression obtained through an NMA of RCTs on interventions for depression among patients with cancer, and their views on the implementation of this evidence in Ireland.

Design
This qualitative study constitutes the second phase of a larger mixed-methods project, which aims to determine the most effective and acceptable treatments for depression among patients with cancer.In the first phase, we are conducting an NMA of existing RCTs to establish the best evidence for the management of depression among patients with cancer, while in this second phase, we will use a qualitative descriptive design to explore stakeholders perceptions of this evidence using focus group discussions (FGDs).Focus groups are group discussions with several participants and at least one interviewer who acts as a moderator or facilitator.In focus groups there is an emphasis on the interactions and collective meanings constructed by the participants when discussing the topic in question (Bryman, 2012).Given their interactive and flexible nature focus groups are particularly appropriate to explore the reasons why people hold the views they do, as well as producing a variety of opinions on the issue being discussed (Bryman, 2012).This methodology has been used to explore patient's experiences of illness and their use of health services, and staff's attitudes and needs (Tausch & Menold, 2016).
The value of qualitative methods using focus groups has already been asserted in psycho-oncological research; for example, in studies looking at the experiences of patients with cancer receiving pharmacotherapy and/or psychotherapy treatments (Rodríguez Vega et al., 2012), the acceptability of internet cognitive behavioural therapy (CBT) programs (Karageorge et al., 2017), and music therapy (Thompson et al., 2017).Focus groups have also been conducted successfully among healthcare professionals to explore their perspectives on the care of patients with cancer and their needs (Rohrmoser et al., 2017).

Participants
We will include two types of participants: (1) people with experience of cancer and depression (referred to here as patients) and (2) healthcare professionals delivering interventions for depression or referring patients with cancer for psycho-social care.We will run 8 FGDs, with 4 to 8 participants each, leading to a maximum of 64 participants in total.

Inclusion criteria
The inclusion criteria for the two types of participants are as follows: To be eligible to participate, (1) Patients must be: • Over the age of 18; • Currently or previously treated for cancer (regardless of cancer type).
• Have had experience of seeking, being referred for and/or receiving treatment for depression after being diagnosed with cancer, during and/or after treatment for cancer.
• Able to attend FGD online.
(2) Health-care professionals must be: • Working in Ireland in a community and/or hospital-based settings (e.g., general practitioners [GPs], physiotherapists, occupational therapists, voluntary sector cancer care providers, public health nurses, psychologists, medical and radiation oncologists, psychologists and psychiatrists, and clinical nurse specialists).
• Providing psycho-social care for patients with cancer or referring patients with cancer for psycho-social care in a professional capacity.
• Able to attend FGD online.

Exclusion criteria
The following exclusion criteria will apply to: (1) Patients: • <18 years old (NMA findings will include data on adult cancer populations only) • Patients in an emergency medical setting • Current psychiatric in-patients • Patients who do not have the capacity to provide informed consent (2) Healthcare professionals.
• Healthcare professionals who have not been involved/ do not have experience in the treatment or service provision for patients with cancer.• Healthcare professionals delivering care primarily outside of Ireland (as the study seeks to inform the provision of treatments for depression in Ireland and the healthcare context is important).

Recruitment
Given the qualitative nature of the study, our approach is to use purposive sampling to recruit participants with as broad a range of different experiences as possible.For the recruitment of patients, we will consider representation across the following characteristics: age; gender; type of cancer, time of diagnosis and stage of cancer treatment; timing of initial depression onset (before, during or after cancer treatment) and diagnosis of depression; types of treatments (if any) obtained for depression and whether these were sought in a hospital and/or a community setting.For the recruitment of healthcare professionals, we will consider representation across the following characteristics: age, gender, profession (e.g., GPs, physiotherapists, oncologists, clinical nurse specialists, psychologists, psychiatrists and voluntary sector care providers), whether work is hospital or community based, length of time working in cancer care, and career stage.
No specific breakdown of these characteristics has been established as part of the inclusion criteria; however, we will prioritise the participation of individuals with different characteristics to maximise our ability to capture different experiences, while acknowledging that it will not be possible to represent all perspectives in our findings.
We will recruit participants through social media using both institutional and personal Twitter accounts, snowball sampling and word of mouth among the research teams' diverse networks of GPs, clinical psychologists, psychiatrists and physiotherapists, and through newsletters and e-mails sent by professional and voluntary organisations (e.g., we will ask for information about the study to be disseminated by the Irish Cancer Society, Patient Voice in Cancer Research, The Irish Society for Lifestyle Medicine, The Irish Psycho-Social Oncology Network, The Irish Association of Nurses in Oncology, the Psychologists Working in Irish Hospitals group, The College of Psychiatrists of Ireland, The Irish Society of Chartered Physiotherapists, The Irish Association for Cancer Research, The Irish Society of Medical Oncology, Irish Practice Nurses Association, the Primary Care Trials Network Ireland among others).
Advertisements for the study will contain contact information for the research team and potential participants will be advised to contact them if they are interested in learning more about the study and potentially participating.When a potential participant contacts the team, they will be sent a participant information leaflet and consent form by post or e-mail.The information leaflet will comprehensively detail the study, the rationale for the project, what participation involves, potential risks and the inclusion and exclusion criteria.Information leaflets will contain all details necessary for the prospective participants to make an informed decision on whether or not to give consent and take part of the study.A researcher will contact interested participants by phone to discuss the study and their participation and answer any questions they may have.The researcher will also assess eligibility and obtain informed consent from all participants prior to participation either electronically using Microsoft Forms secure platform or in writing using paper copies of the consent form, which will be returned by mail to the project team.Informed consent will be obtained from all participants to (a) participate in the research study and (b) have their data processed for research purposes.Participants will be given a copy of their signed consent form for their records.

Telephone survey
Participants will be asked to complete a brief telephone survey to collect demographic information.This information is needed to provide some context to the FGD results, as experiences of and views on the management of depression may vary depending on these factors.
For patients, this will consist of questions on their: • age [in 5-year bands], • gender, • type of cancer they were diagnosed with (e.g.breast cancer), • cancer stage (I -IV), • the number of years since their cancer diagnosis (in bands: within the last year, 1 -2 years ago, 2 -5 years, 5 -10 years, 10+ years), • The onset of their (first episode of) depression (in bands: before period around cancer diagnosis/treatment, during cancer treatment, or after cancer treatment was completed), • whether they received a formal diagnosis of depression (yes/no), • treatment for depression, if any (e.g., antidepressant medication, psychotherapy, exercise, combination therapy, etc.), • context of depression treatment (i.e., within the hospital system or community-based care).
For healthcare professions this will consist of questions on their: • gender, • profession (e.g.GP, physiotherapist etc.) • grade (e.g., senior, clinical specialist etc.), • work context (i.e., in a hospital [cancer-centre or regional hospital] or a community based setting), and • length of time working in cancer care (in bands).

Focus group interview
We have developed a semi-structured topic guide for the FGDs with input from the project Patient and Public Involvement (PPI) advisory panel and the clinical members of the research team with experience delivering the three main types of depression interventions.The semi-structured format of the guide will give the facilitator flexibility to adapt the questions and expand on topics when new points are raised.FGDs will address participants' experiences of the treatments for depression included in the NMA, their views on how acceptable and appropriate they think these treatments are in practice, and their views on how the findings should inform and be implemented in practice including any perceived barriers and facilitators to accessing these treatments.

Procedure
A member of the research team will contact participants by telephone, at a time convenient to them, to complete the brief telephone survey and answer any questions ahead of the FGDs.
At least one week prior to the FGDs, the participants will be provided with a briefing sheet on the findings of the NMA so that they can review these before the FGD if they so wish.The briefing sheet will be developed in consultation with the project PPI panel to ensure it is written in lay language.Reviewing the briefing sheet will not be a requirement for taking part in the study.
We will hold eight FGDs (4-8 participants each); some with a mix of patients and healthcare professionals and some with only either patients or healthcare professionals (Finch & Lewis, 2007).Participants will have a choice whether to participate in a mixed FGD or a FGD with only other patients or healthcare professionals and participant preference will determine the number of mixed FGDs that we ultimately include.The rationale for having mixed focus groups of patients and healthcare professionals is to maximize the sharing of perspectives and the discussion of different views on treatments for depression and implementation in practice.However, we are cognizant that not all patients or healthcare professionals may be comfortable expressing their views in mixed FGDs.Furthermore, power dynamics between patients and healthcare professionals and among different types of healthcare professionals may deter some participants from speaking openly.Therefore, we will give participants a choice of which type of FGD they would feel most comfortable in.If it proves impossible to organise FGDs with some participant groups, and further exploration of particular issues is needed to reach data saturation (Mason, 2002), we will consider running individual interviews.
We will carry out the FGDs online through MS Teams or Zoom and will be video-recorded.Online data collection is comparable in quality to face-to-face and we opted for this format to support COVID-19 preventative measures, to minimise the costs of participation both in terms of money and time, and to facilitate participation for patients and healthcare professionals across Ireland FGDs (Willemsen et al., 2022).In line with recommendations, we will aim to limit the number of participants in each FGD to five if possible (Willemsen et al., 2022).
The FGD will each be facilitated by two members of the research team, one of whom will run the FGD while the other will act as a note-taker/rapporteur and assist in managing the FGDs.
In line with guidelines on online FGDs, participants will be given the opportunity to join a practice session using the online platform ahead of their scheduled FGDs and they will be informed that the online room will be opened 15mins before the start time so that they can join early and resolve any technical problems should they arise (Willemsen et al., 2022).
In addition, participants will be asked to follow "house rules" including using headphones if available, going on mute when not speaking to minimise background noise, enabling the video function, and using the "raise hand" function when they want to speak (Willemsen et al., 2022).
At the beginning of each FGD, one of the researchers will give a brief (~10 -15 minute) PowerPoint presentation of the main findings from the NMA to outline the evidence, effectiveness and acceptability rankings for each treatment type included, as well as the limitations of the NMA.This presentation will be used to initiate discussions into participants' views and experiences of these treatments, how acceptable and appropriate participants think these treatments are in practice, and any perceived barriers and facilitators to accessing these treatments.
Participants will be free to ask questions during or after the presentation.In line with guidelines for online FGD (Willemsen et al., 2022), the FGD will be a maximum of 90 minutes including the presentation, with a brief 5 min break approximately halfway through.

Ethical approval
A Data Protection Impact Assessment (DPIA) has been completed and approved for the project.Ethical approval for the study has been given by the RCSI research ethics committee.

Data management Transcription
The built-in recording and transcription functions of MS Teams will be used to video record and transcribe the FGDs.MS Teams saves video recordings and transcriptions directly onto the RCSI OneDrive, which is secure and private; from here they can be deleted by the administrator (the person who started recording).Immediately after each focus group, we will save the video recordings and transcriptions to the study folder on RCSI's OneDrive and the original video recordings and transcriptions will be deleted and the files will also be erased in the recycle bin to ensure they are permanently deleted.
The Principal Investigator and project research assistants will review all transcriptions and manually correct any errors and remove all instances of identifying information (e.g., names), replacing these with pseudonyms.In the event that MS Teams transcriptions are not of sufficient quality, Trint Transcription Services (www.trint.com)will be used to transcribe the video recordings.Trint is a secure online platform that automatically transcribes uploaded audio and video and is fully compliant with the EU General Data Protection Regulation (GDPR).
All participants will also have the opportunity to review their transcripts and change or remove their data before analysis takes place.Digital versions of the data will be erased once transcriptions are completed and verified by participants.

Data storage
Hard copies of consent forms will be securely stored in a locked filing cabinet in RCSI until they are scanned, after which they will be destroyed.All electronic data will be stored on an online secure server, RCSI's OneDrive, using encryption software and it will be password protected.Video files will be deleted after transcription is completed and transcripts are verified.
All data will be pseudonymised and the Principal Investigator will hold the key to re-identify participants.Data from the telephone survey will be entered into an excel spreadsheet under participants' pseudonyms.Access to the transcripts before pseudonymisation will be restricted to the Principal investigator and the research assistants (EK and RB), who will carry out the transcribing.After all transcripts have been pseudonymised and all identifying information has been removed, access to the transcripts will only be granted to other members of the research team for input on analysis; this will be done using a password protected shared drive.
The transcribed data and the data collected from the telephone survey will be retained pseudonymised for 5 years; thereafter, it will be deleted.

Analysis
Descriptive statistics (largely counts and percentages etc.) will be calculated in Excel and used to present participant characteristics (e.g., the number/percentage within each age band).
Transcribed focus group data will be analysed using the Framework Method (Gale et al., 2013).Framework analysis involves a clear systematic approach and generates very structured outputs of summarized data, making it especially useful when multiple multidisciplinary researchers are working on a project with a large data set.It also lends itself well to studies where both deductive and inductive approaches to analysis are called for.In the current study, a deductive approach -that is driven by the results of the NMA -will be used to address some research questions (e.g., "Do stakeholders agree or disagree with the NMA results in terms of what treatments are the most effective and/or acceptable?" and "How accessible do stakeholders consider the most effective/acceptable treatments to be in Ireland?", while a more inductive data-driven approach will be used to address others (e.g., "What are stakeholders' experiences of these treatments?", "How could the most effective/acceptable treatments best be implemented in Ireland?", and "What are the biggest facilitators and barriers to implementing the results in clinical practice?").
Analysis will be carried out in Nvivo 12 software, which enables data storage, organisation, and comparisons within and between transcripts and data from individual participants.Analysis will begin directly after each focus group is completed using the procedure described in Table 1, based on the methods for Framework analysis described by Gale et al. (2013).
Individual participants will be used as the unit of analysis of the data, whereby the framework matrix will enable each participants' views to be connected to their responses across other codes so that their individual context and account is considered throughout (Gale et al., 2013).However, similarities and differences across stakeholder groups (e.g., patients vs. healthcare professionals, community-vs.hospital-based healthcare professionals, and professionals referring for or delivering

Stage Description of method
Stage 1: Transcription Transcription will be carried out as described in the data management section, and will be used as an opportunity for data immersion.As analysis will focus on content rather than the structure of participants' responses, only very long pauses, interruptions and nonverbal communication that convey information about group interactions and dynamics will be noted in the transcriptions.
Transcriptions will also be supplemented with notes made during or after the FGDs (e.g., documenting background information of note).
Stage 2: Familiarisation Members of the research team will read and re-read the transcripts and contextual notes, and listen back to audio-recorded FGDs to become familiar with the entire data set.Researchers will record initial impressions, interpretations and analytic notes in a notebook or word document during this stage and throughout the analytic process.This will include any thoughts around the nature of group interactions within the FGDs.
Stage 3: Coding Transcripts will be read line by line and coded by attaching labels that describe the data, so that the data relating to these can be compared systematically across other parts of the data set.Both pre-defined and open coding will be used to reflect the combination of a deductive and inductive analytic approach.For example, initial pre-defined codes will include stakeholder agreement and disagreement with the NMA results; however, open coding will be used to develop codes that capture these views in more depth (e.g., to describe reasons for agreement/disagreement).At least two researchers will independently code the first two transcripts.
Stage 4: Development of an analytical framework After two transcripts have been independently coded, the research team will meet to discuss and compare codes, group codes into meaningful categories, agree on a set of draft codes, and develop brief descriptions of these codes that explain what they relate to and examples of what might be summarised under that code.Following this, the draft coding framework will be independently applied to a third transcript, discussed amongst the research team and refined further.It is envisioned that several iterations of this process of refinement may be needed to develop an analytic framework that adequately captures the data.
Stage 5: Application of the analytic framework The analytic framework will be applied to each of the transcripts by systematically going through each one, and selecting and attaching an appropriate code to sections of text.If codes can be meaningfully grouped under clear categories (for example, at a minimum it is envisioned that there will be separate categories for stakeholders' perceptions of NMA results and for their views around implementation of the findings), a separate framework matrix will be used for each category.
Stage 6: Charting of data into the framework matrix All data will be charted into a matrix that is comprised of one row per participant and one column per code.This will involve reducing the data by summarising it in a way that retains it's original meaning or key idea, and documenting any illustrative quotations that relate to a given code.
Researchers will agree on abbreviations and compare and discuss their individual summarising styles after one transcript to maximize consistency among team members.At least 25% of the transcripts will be double coded.
Stage 7: Interpretation of the data Themes will be developed by reviewing and discussing researchers' notes and the framework matrix and drawing connections within and between participants' data for and across the codes and categories.Discussions among the research team and drafting of analytic memos will be used to generate, explore, and develop themes around the original research questions and any new concepts that emerged from the data.
different types of treatments for depression) will also be explicitly considered.In addition, the analysis will also include group interactions within the FGDs by considering a) data on group discussions (e.g., does the group have consensus/dissent, and do opinions change over the course of the FGD) and b) data on group interactions (i.e., how the group interacts with one another) (Onwuegbuzie et al., 2009).Such data will also be considered in the context of whether particular FGDs involved only patients or healthcare professionals or both.If individual interviews are carried out, these will be analysed and reported separately.
To draw on the strengths of quantitative NMA and qualitative FGD methods, we will triangulate findings from the two project phases at interpretation level (Fetters et al., 2013) (see Figure 1).This will involve combining results into meta-themes and using a convergence coding matrix to report whether the findings are in agreement, silence, or dissonance.

Results and dissemination
This will be the first study to obtain stakeholders' views of an NMA of depression treatments in cancer.The study described in this protocol has not yet commenced.It is anticipated that data collection will begin in June 2023.
While the results of our NMA will provide the best available evidence for managing depression among patients with cancer, the findings from this qualitative study will facilitate the direct national impact of NMA results across key aspects of health services, including policy, practice and service delivery.We will co-produce a report with patient and healthcare representatives summarising the evidence for the different depression management options available to patients with cancer, outlining how effective and acceptable they are for patients.
In the report, our recommendations will focus on the findings that are most important to patients and healthcare professions delivering cancer care and their ideas about how to best apply these in Ireland, as evidenced by the results of the FGDs.In addition, we will present our findings at international academic conferences, patient conferences and public talks and submit our findings for publication to a peer-reviewed journal.

Conclusion
The findings from this study will be valuable for healthcare professionals and patients, helping them to make choices about how to best manage depression among patients with cancer.
Insights from this study will help to inform and enhance the delivery of psycho-oncology services in Ireland and bridge the gap between evidence-based medicine and the experiences of patients with cancer in the "real world".
First, I wonder if the authors could provide more detail regarding the network meta-analysis so that a naive reader could understand what the outputs might look like (i.e.get a sense of the participant-facing information for this study).NMA results can be complex, so how will these be given to participants in an accessible way?I appreciate that there will be a presentation (which I commend the team for), but more information about what might be included within the briefing sheet might be helpful here.I would also encourage the team to reflect more broadly on the limitations of NMA (e.g.considering aggregate rather than individual patient data; what cannot easily be manualised from a therapy point of view, and the implications of this; language and publication bias, and so on) when presenting their findings.
Second, I wonder if the team had considered recruiting patient who had not sought treatment or support for their depression?At present, there is the risk of recruiting a motivated, help-seeking sample which may not adequately reflect the reality of living with cancer and depression.
Third, will any adaptations be made for people with cognitive or communication difficulties, or physical health difficulties, to aid accessibility?(e.g. head and neck cancer survivors who may have difficulty with speech).
Fourth, are there any exclusion criteria?We know that prior experience of depression is a risk factor for subsequent depression; do patients need to have experienced depression within X months before/after a diagnosis?What about patients in the palliative phase, or patients with comorbid mental health difficulties that may influence their experiences?
All in all, the authors are to be commended on a really excellent proposal.The above suggestions are designed to strengthen the work but are just suggestions.I wish the authors all the best with their work.
Is the rationale for, and objectives of, the study clearly described?Yes Is the study design appropriate for the research question?Yes Are sufficient details of the methods provided to allow replication by others?Yes Are the datasets clearly presented in a useable and accessible format?

Not applicable
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Psycho-oncology; systematic review; evidence synthesis; depression; anxiety; fear of cancer recurrence; psychological processes; clinical psychology; health psychology.
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Figure 1 .
Figure 1.Sequential explanatory mixed-methods design with emphasis on the qualitative phase.