Models of support to family members during the trajectory of cancer: A scoping review

Aims and objectives: To map the existing literature on support models provided to family members during the cancer trajectory. Background: Cancer diagnosis, treatment and survivorship have a profound influence on the surrounding family members. This scoping review is part of the development of a support model for family members of persons diagnosed with colorectal cancer. Design: The method was guided by the Arksey and O’Malley framework, described in the Joanna Briggs Institute guidelines, and the reporting is compliant with PRISMA-ScR Checklist. Searches were conducted in PubMed, CINAHL and PsycINFO from November 2019– February 2020 with no limitation in publication year or study design. Complementing searches were conducted in reference lists and for grey literature, followed by an additional search in September 2020. Inclusion criteria were primary research about support provided by health care, to family members, during cancer, of an adult person, in Swedish or English, of moderate or high methodological quality. Quality


| INTRODUC TI ON
Cancer diagnosis, treatment and survivorship, described as the cancer trajectory, have a profound influence on the surrounding family members.The impact of the diagnosis leads family members not only to needing support of their own (Lambert et al., 2012;Lavallée et al., 2019;Sklenarova et al., 2015) but also to having needs that shift throughout the cancer trajectory (Given et al., 2012).Consequently, numerous interventions combining information with emotional support have been designed to support family members (Treanor et al., 2019).However, their effect on the family members' quality of life (QoL) is small and short lived.This, along with persistent reports of unmet needs for emotional, informational and relational support (Lambert et al., 2012;Lavallée et al., 2019;Mollica et al., 2020;Sklenarova et al., 2015), calls for increased focus on how to provide adequate support.Consequently, this scoping review was designed as an initial step in the development of a support model for family members of persons diagnosed with colorectal cancer in Sweden.Before designing a new intervention, examining the existing support models appears beneficial.Thus, this scoping review was conducted to map and review existing support models for family members of a person diagnosed with cancer.To examine factors affecting the likelihood of meeting the needs of family members, this study contrasts the existing support models against unmet needs described by Lambert et al. (2012) and cancer trajectory phases described by Given et al. (2012), since these are widely cited and clinically applicable.
In this scoping review, 'family member' is used when referring to a person connected to the person stricken by cancer and to Wright (2013) definition: 'the family is who they say they are'.Thus, family members also include, for example, next of kin and friends.
In addition, it encompasses 'family caregiver' and 'informal caregiver'.The rational to also including these roles are that the definition of family caregiving by Weitzner et al. (2000) describes not only the management of disease symptoms and treatment of side effects, coordinating or administering treatments in the home, but also the providing of emotional support and assisting the person with activities of daily living.This implies that the role of a family member and/or a family caregiver appears to be close connected, or even intertwined, and assumingly not static, meaning needs might change over time related to their level of engagement.Thus, as a result of this position, this scoping review does not distinguish between being a family member and a family caregiver when searching for models of support.
A cancer diagnosis not only affects the family members themselves but also challenges the relationships within the family, established communication patterns, and roles and responsibilities (Northouse et al., 2005).In addition, the role of the family members during the cancer trajectory is multifaceted-comprising practical, psychological and empathetic aspects (Blindheim et al., 2013).They experience a physical, mental and emotional struggle.A cancer diagnosis even has a negative effect on the family members' health; for instance, they have higher rates of depression and weakened immune response (Aizer et al., 2013), as well as increased risk of ischaemic heart disease (Mollerberg et al., 2016) and of coronary heart disease and stroke (Ji et al., 2012).These risks persist over time.
Moreover, self-reported poorer QoL also persists through years of survivorship, especially if the family members' needs for support are not met during the cancer trajectory (Kim & Carver, 2019).
Consequently, researchers have emphasised the necessity of focusing on family members as well as on the person diagnosed (Ji et al., 2012;Kim et al., 2019;Norlyk & Martinsen, 2013).Lambert et al. (2012) have categorised family members' need for support in six domains of unmet needs: • Comprehensive cancer care, for example, being told about the help healthcare professionals can offer, having possibilities to participate or help in the person's care and receiving appropriate information from healthcare professionals.
• Emotional and psychological, for example, receiving help in dealing with own emotional distress and getting emotional support for self/having someone to talk to.
• Partner or caregiver impact and daily activities, for example, receiving financial support, help dealing with impact on work and help looking after own health.
• Relationship, for example, receiving help communicating with the person about illness and concerns and communicating with others, such as family and friends.
• Information, for example, knowing what to expect from the illness, treatment and prognosis.
• Spirituality, for example, feeling there is hope for the future and receiving spiritual support.
The categories of unmet needs echo the patterns generally reported in the cancer family caregiver literature (Chen et al., 2016;Girgis et al., 2011;Osse et al., 2006); therefore, it is used in this scoping review to contrast the existing models of support.In addition, researchers have proposed that the cancer trajectory significantly influences what the family members need and when they need it (Given et al., 2012;Northouse, 1984).Thus, meeting the needs of family members requires recognising phase-specific problems and needs.The cancer trajectory follows two winding paths that sometimes overlap, for What does this paper contribute to the wider global community?
• In general, the reviewed support models consider neither the impact of the specific diagnosis nor the cancer trajectory, which may negatively affect the likelihood to meet the family members' needs.
• A majority of the support models reviewed contains multiple components, and as a result, may be difficult to apply in clinical practice.
• Future research needs to include the prerequisite of clinical practice when designing support models, to enable for the support to be implemented and thus reach the targeted family members instance, in case of recurrence (Fletcher et al., 2012).One path leads to palliative care and the end of life, and the other to rehabilitation and survivorship.This scoping review focuses on the latter.Consequently, to elucidate the support models' recognition of the trajectory, we have sorted all the reviewed models into a phase of relevance as described by Given et al. (2012)-namely, diagnosis, treatment and survivorship.
Many interventions have been developed over the past decades to support family members of persons with cancer.For instance, systematic reviews have described interventions aiming to support family caregivers in pain management (Chi et al., 2020) and interventions aiming to improve cancer caregivers' QoL (Waldron et al., 2013).In addition, they have reported on interventions supporting family caregivers who care for persons with advanced cancer at home (Ahn et al., 2020) and on interventions aiming to improve the care for persons diagnosed with lung cancer (Kedia et al., 2020).The effectiveness of cancer helplines has also been systematically reviewed (Heckel et al., 2019).Fu et al. (2017), Ferrell and Wittenberg (2017), and Applebaum and Breitbart (2013) all reviewed randomised control trials of psychosocial interventions.Further, Treanor et al. (2019) have discussed quasiexperimental trials.To the best of our knowledge, no review has been conducted aiming to include support models regardless of study design.By doing so, this scoping review seeks to broaden the picture of existing models and examine them in relation to family members' unmet needs-thereby elucidating the relevant factors for designing support models for family members of persons with cancer.In preparation for this scoping review, we made searches to locate comparable, published or ongoing, scoping reviews in PubMed, Cumulative Index to Nursing and Allied Health Literature (CIHNAL), Cochrane Library and Joanna Briggs Institute Systematic Review Register.However, none were identified.

| AIM
This scoping review aimed to map the existing literature on support models provided to family members during the cancer trajectory.
Therefore, we asked the following research questions: (i) What are the characteristics of the models described?(ii) During which phase of the cancer trajectory is the described support provided?(iii) What are the aims of the support?(iv) To whom is the support directed?

| Design
The study was designed as a scoping review.According to Arksey and O'Malley (2005), this design allows for further exploration by including unlimited study designs, settings and outcomes.
Furthermore, the design allows for additional questions to be asked on the research topic of interest.The design and implementation were guided by a methodological framework developed by Arksey and O'Malley (2005), refined by Levac et al. (2010) and Colquhoun et al. (2014), and described by the Joanna Briggs Institute (2015).Prior to the review, a protocol was constructed to describe the planned methodology and search strategy (https:// bmjop en.bmj.com/content/10/9/e0376 33.info).To achieve rigour, the reporting was compliant with the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews Checklist (File S1).To enable replication and enhance reliability, we documented the search and selection process using a PRISMA flow chart (Figure 1).

| Search strategy
The Population Concept and Context (PCC) mnemonic recommended for scoping reviews was used to establish effective search criteria (Table 1).Search tools such as Medical Subject Headings, Headings, Thesaurus and Boolean operators (AND/OR) were used to expand and narrow the search.Keywords (e.g. 'support' and 'neoplasm') and synonyms (e.g.'family', 'next of kin' and 'partner') were applied to the different databases.
Grey literature identified in the databases was scoped to identify unpublished studies and ensure that no relevant reference was missed.For the same reason, additional searches for grey literature were conducted in Google Scholar and SwePub.Grey literature identified included books, conference abstracts and research posters.

| Inclusion criteria
The following inclusion criteria were applied: a primary research in Swedish or English about support provided by healthcare professionals to family members during the cancer trajectory of an adult person.The reason for the limitation in language was that the available resources of the research team were not enough to hire a professional translator.In addition, the following exclusion criteria were applied: first, to exclude studies describing support targeting specific illness-related issues as they are not applicable to other diagnoses.Second, to exclude studies focusing on the terminal/palliative phase and support postmortem as this study focuses on the trajectory towards survivorship.Third and last, to exclude studies focusing on children (< 18 years old) due to their specific needs of support related to maturity.
A quality cut-off was set at studies not presenting one of the following: aim, criteria for inclusion and exclusion, sample, data collection, the process of analysis (all study designs), and a description and analysis of dropouts (quantitative studies).The rationale behind this decision was that their presentation of the support models provided to family members would be incomplete without this information.Consequently, these studies were categorised as having 'low quality' and excluded (n=5; Table 2) whereas the remaining studies containing the required methodological information were included.Assessment was conducted by two reviewers independently and, if any disagreements, discussed in the research team.To enable systematic exclusion of studies with incomplete methodological description, the Joanna Briggs Institution Critical Appraisal tools (Tufanaru et al., 2017)-'Checklist for Qualitative Research', 'Checklist for Quasi-Experimental Studies' and 'Checklist for Randomised Controlled Trials', were used.

| Study selection
The study selection process is presented in a flow diagram (Figure 1).First, all titles (6,140) were screened for relevance by the first author (MS) in collaboration with the librarian, consulting the research team at any hesitation.Searches in grey literature and reference lists identified another 33 studies.All relevant titles were imported into a reference programme, Endnote X9, and checked for duplicates.After removal of duplicates and screening of titles, 83 studies were excluded due to being reviews, studies about instrumental development or studies describing family caregiver needs.Second, 431 abstracts were screened and assessed with the inclusion criteria using the Rayyan system for systematic reviews by Ouzzani et al. (2016).
The rational for using Rayyan is that its system enables a systematic collaboration and sorting.To enhance the cohesion during the study selection process, all abstracts were reviewed independently by two reviewers.Eligible studies (75) were retrieved and read in full text by the research team members, as described above regarding the abstracts.Studies were assessed with the inclusion and exclusion criteria using a charting form, and 38 studies were consequently excluded.

| Analysis
In this scoping review, a semantic deductive thematic analysis was conducted.This approach was chosen as it allows a descriptive analysis, during which data were sorted using four research questions as well as the six domains of unmet needs described by Lambert et al. (2012) and mentioned in the introduction.Thus, the deductive approach described by Braun and Clarke (2006), in which the coding and theme development are directed by existing concepts or ideas, seemed most appropriate to answer the research questions, resulting in the study protocol (which previously stated that the analysis would follow an inductive approach) had to be modified.Consequently, the analysis was directed by the following: (i) What are the characteristics of the models described?
(ii) During which phase of the cancer trajectory is the described support provided?
(iii) What are the aims of the support?
(iv) To whom is the support directed?
First, the extracted data were read through and the parts of text answering the research questions were highlighted and sorted into new Microsoft Word documents named 'model', 'trajectory', 'aim' and 'participating family member'.Data in the document 'trajectory' were sorted into the trajectory phases described by Given et al. (2012).Data in the document 'models' were further divided into 'description of the intervention', 'sessions', 'mode' and 'by whom'.Lastly, based on the description of the support, the models grouped were divided into three themes.The models identified were then compared with the unmet needs described by Lambert et al. (2012).

| Consultation
In line with Arksey and O'Malley (2005) purpose of consulting with stakeholders, contact nurses were approached with the preliminary results and asked to offer additional information, perspective, meaning and thoughts on applicability.A contact nurse is a registered nurse with specialist knowledge in oncological and psychosocial care who is assigned to support both the person diagnosed and their family members throughout the cancer trajectory.A total of five contact nurses were asked to take part in this process and reflect over the schematic results' relevance and applicability to clinical practice.The contact nurses were from two regions in Sweden and cared for persons with gastrointestinal, urological and breast cancer.A telephone or digital meeting was arranged with each nurse individually at their convenience.The contact nurses recognised the content in the support models as part of the support they provided and acknowledged its importance.However, they did not provide such support in a structured manner.Instead, they offered support when a need was identified, for example, if a family member contacted them and expressed a wish for information.Furthermore, they found the support models difficult to apply, due to both lack of time and lack of knowledge.To explore MBSR participation for couples affected by cancer.In particular, it examines the programme's impact on symptoms of stress, mood disturbance, and mindfulness for both cancer patients and their partners; it also begins to explore relationships between couples' outcomes.9. Budin et al. (2008), USA To conduct a randomised controlled clinical trial of phase-specific evidence-based psychoeducation and TC interventions to enhance emotional, physical, and social adjustments in patients with breast cancer and their partners 10.

Carlson et al. (2017), Canada
To report the results of a negative randomised controlled trial that piloted brief supportiveexpressive therapy (SET) for partners of men with prostate cancer and to discuss lessons learned for future clinical trials 11.

Carlsson and Strang (1998), Sweden
To evaluate whether educational groups have an effect on the perceived level of knowledge and whether this has a positive effect on mood 12.

Chambers et al. (2014), Australia
To compare the effectiveness of two low-intensity approaches for distressed patients with cancer and caregivers who had called cancer helplines seeking support 13. Chien et al. (2020), Taiwan To understand the effectiveness of a couple-based psychosocial information package (PIP) and multimedia psychosocial intervention (MPI) on patients with prostate cancer and their partners 14.

Chiquelho et al. (2011), Portugal
To describe proFamilies: a psychoeducational multi-family discussion group intervention for cancer patients and their families 15.

Gabriel et al. (2019), South Africa
To implement and evaluate the effectiveness of a psychosocial intervention programme on the QoL and caregiver burden of the primary caregivers of women with breast cancer 16.Gjerset et al. (2019), Norway To investigate differences between female and male caregivers regarding (a) health status at the start of the 1-week educational programme and at 3 months after termination, (b) self-reported needs for support at the start of the programme, and (c) changes in health status from the start to 3 months after termination of the programme 17.

Heckel et al. (2019), Australia
To evaluate the utility of a telephone outcall programme for cancer caregivers and to examine longitudinal changes in their distress levels and supportive care needs 18.
The included studies describe 39 support models provided to family members during the cancer trajectory, presented in three themes: psychoeducation (n = 26), caregiving training (n = 6) and psychological support (n = 7).Each theme describes how, by whom and to whom the support is provided.Table 4 offers detailed information on the aim and design of the support models.Table 5 relates the support models to the six domains of unmet needs described by Lambert et al. (2012) and shows during which trajectory phase the support was provided as described by Given et al. (2012).

| Psychoeducation
Out of the 32 studies reviewed, 19 described 26 different psychoeducational support models provided to family members during the cancer trajectory.The models predominately addressed the treatment phase of the cancer trajectory and were focused on addressing emotional and psychological unmet needs, unmet needs related to the partner or caregiver impact and daily activities, and relational and informational unmet needs.
A nurse-led, telephone, psychoeducation programme that comprised structured telephone calls to understand carers' situation and identify problems in order to provide information, education, and psychosocial support.Specific caring problems or psychological issues were identified, and related information or psychological support was given to carers.
The nurse also provided education to carers according to the patient's needs at different recovery stages.Before the end of the call, the nurse asked about any additional problems and ensured that carers' needs had been met Three sessions, three and 5 weeks after discharge.Each call lasted no longer than 45 min information relevant to living with cancer.Hand-outs were provided focusing on orientation to the cancer centre and the treatment team; suggestions for communicating with healthcare providers; resources for health information, psychosocial support, and financial concerns; evaluating health information on the internet; the impact of cancer on different domains of quality of life; and suggestions for maintaining quality of life.
Four sessions.The first session lasted approximately 75 min, and sessions 2 through four lasted 45 min each.Four sessions Schellekens et al. (2017) Mindfulness-based stress reduction includes mindfulness practice and teachings on stress.Mindfulness exercises (such as body scan, gentle yoga, sitting, and walking meditation) were practiced.Participants received teachings on stress and grief and were invited to share their experiences Weekly 2.5 h sessions for 8 weeks, including daily 45 min home practice and one 6 h silent day TA B L E 4 (Continued)

| Caregiving training
Out of the 32 studies reviewed, 6 described 6 different caregiving training support models provided to family members during the cancer trajectory.The models predominately addressed the treatment phase of the cancer trajectory and were focused on addressing informational and relational unmet needs.
Caregiving training focused on supporting family members by teaching them strategies and practical skills in how to support and care for the person diagnosed with cancer.The interventions in the reviewed studies taught practical caregiving training (Belgacem et al., 2013), symptom management (Hendrix et al., 2013;Kozachik et al., 2001), caregiver-assisted self-care strategies (Sherwood et al., 2012), coping-skills training (Porter et al., 2011), emotional support, coordination of services and caregiver preparation to care (Kozachik et al., 2001).Of the studied interventions, two were provided in hospital setting (Belgacem et al., 2013;Hendrix et al., 2013), one in person plus telephone (Kozachik et al., 2001) and two by telephone only (Porter et al., 2011;Sherwood et al., 2012).The number of sessions ranged from 1-14, and the sessions lasted for 20-180 min (the median was 52 min).The interventions were all provided by nurses.As with the psychoeducation, the invited family members were identified by the person diagnosed with cancer, focusing on those providing informal care.There were some variations in terminology in the studies describing this intervention as well: 'caregivers' (Hendrix et al., 2013), 'informal caregivers' (Belgacem et al., 2013), 'family caregivers' (Kurtz et al., 2005;Sherwood et al., 2012) and 'primary caregivers' (Porter et al., 2011).

| Psychological support
Out of the 32 studies reviewed, 7 described 7 different psychological support models provided to family members during the cancer trajectory.The models predominately addressed the treatment phase of the cancer trajectory and were focused on addressing emotional and psychological unmet needs and relational unmet needs.
The psychological support discussed in the reviewed studies involved counselling and therapy rather than education, and it was provided through a series of sessions conducted by psychologists or therapists.Its focus was on mindfulness (Birnie et al., 2010;Kohle et al., 2017;Schellekens et al., 2017), coping strategies (Carlson et al., 2017;Heinrichs et al., 2012) and emotional disclosure (Porter et al., 2009).Furthermore, the interventions were dyadic, focusing on the relationship between the person diagnosed with cancer and their family member (Birnie et al., 2010;Carlson et al., 2017;Heinrichs et al., 2012;Kohle et al., 2017;Kuijer et al., 2004;Porter et al., 2009), and only two were conducted in group settings (Carlson et al., 2017;Schellekens et al., 2017).The number of sessions ranged from 4-8, and the sessions lasted for 75-150 min (the median was 90 min).The psychological support differed from psychoeducation and caregiver training in terms of participating family member; the studies describing psychological support did not focus on the person who provided support to the cancer person but on the relationship between the two.Hence, apart from one study that invited a life partner or another close relative/friend (Schellekens et al., 2017), all persons invited their spouses.No effect was found in partners, possibly because they were more focused on the patients' well-being rather than their own TA B L E 5 Support models related to unmet needs and phase of trajectory five of the six domains of unmet needs (Table 5).Out of those, the most commonly addressed domains are Informational and Emotional needs, which correspond with the most frequently reported needs.

| DISCUSS ION
This could be an indication that the psychoeducation support model is tailored to meet the needs reported in the literature, which increases their likelihood to be met.
In a majority of the support models reviewed, it is not articulated to which specific part of the cancer trajectory the model of support is applied.Out of the 39 described models of support, only 14 were provided during a specific phase (see Table 5), of which 11 were provided during treatment.However, eight of those 14 models did not describe the support as intentionally targeting phase-specific needs.Only two of these models of support were designed explicitly to meet phase-specific needs (Budin et al., 2008;Northouse et al., 2007;Porter et al., 2011), of which the support model described by Budin et al. (Budin et al., 2008) was the only one intentionally targeting phase-specific needs in each phase of the breast cancer trajectory.The lack of focus on phase-specific problems and needs was previously described in a systematic review by Ussher, Perz, Hawkins and Brack (Ussher et al., 2009).Unfortunately, due to the sparse amount of interventions considering the cancer trajectory, no conclusion on its importance can be made based on the existing support models.However, although the effect of the phase in a cancer trajectory on support models has not been established, research on family members' experiences and needs (Fletcher et al., 2012;Given et al., 2012;Northouse, 1984) emphasises its significant impact; accordingly, it should not be ignored.To illustrate, the diagnostic, or initial phase-on which none of the studied psychoeducational interventions focused-has been described by Given et al. (2012) and Northouse (1984) as one of the most stressful times for family members.In addition, during this phase, the prevalence of reported unmet needs is among the highest (Lambert et al., 2012).The other end of the trajectory, the survivorship phase, also has its specific problems and needs.Even though the overall needs for support decrease over time, family members continue to report unmet needs that require attention (Kim et al., 2019).Nevertheless, only one support model in the reviewed studies focused on this phase (Badr et al., 2016).Thus, based on the literature, it appears important to consider the trajectory when designing a support model, but its significance requires establishment by future research.
Another relevant finding is the surprisingly high number of support models provided to family members of persons with different cancer diagnoses.In this scoping review, only half of the support models were illness specific.Interestingly, only one of those (Nga Fan et al., 2014) focused on colorectal cancer, compared to six focusing on breast cancer and four on prostate cancer.The remaining models included either multiple diagnoses or did not mention the targeted diagnoses.Persons diagnosed with colorectal cancer participated in four of the reviewed models of support (Birnie et al., 2010;Kohle et al., 2017;Kozachik et al., 2001;Porter et al., 2009), together with persons diagnosed with breast, lung, pancreatic and prostate cancers.Arguably, a universal design is advantageous due to its transferability and its ability to facilitate recruitment of study participants.In particular, studies have emphasised that a crucial share of experiences and needs are shared by family members regardless of cancer diagnosis (Given et al., 2012;Northouse, 1984).For example, the experiences of family members of persons with colorectal cancer (Mosher et al., 2016;Sjövall et al., 2011) are similar to those of family members of persons with breast cancer (Holst-Hansson et al., 2017) or prostate cancer (Piervil et al., 2019)-such as feelings of shock at the diagnosis, a need for information, an adjusted everyday life and feelings of uncertainty (Holst-Hansson et al., 2017;Mosher et al., 2016;Piervil et al., 2019;Sjövall et al., 2011).Consequently, it could be argued that a support focusing on, for example, informational needs may apply to all family members.On the other hand, comparisons of healthcare use and physical and psychiatric morbidity among partners of persons with the mentioned diagnoses reveal significant differences in which colon and rectal cancer stand out in higher rates (Sjövall et al., 2009).This may indicate the severity of the impact of the diagnosis on the family members, speculatively due to symptoms and treatment or lack of adequate support.Irrespectively, it calls for careful consideration of the multi-diagnosis design to try to meet family caregiver needs.To illustrate, during the early postoperative period of colorectal cancer surgery, the person diagnosed describes experiences of total dependency on support from others (Jakobsson et al., 2017).Family members describe the same phase as having a severe impact on them and as a heavy burden, due to the responsibility for the person's well-being and compliance with everyday life (den Bakker et al., 2018;Norlyk & Martinsen, 2013).Presumably, the early treatment phase may be experienced differently by family members of persons undergoing other treatments, which may mean different needs of support.
For instance, family members of persons with prostate cancer describe experiences and needs related to masculinity and erectile dysfunction (Piervil et al., 2019), which are obviously not applicable to family members of the other diagnoses.Further, persons diagnosed with breast, prostate and colorectal cancers seem to use different coping strategies throughout the cancer trajectory (Lashbrook et al., 2018).This diversity may also have an impact on the effect of a universal emotional support model for family members of the three diagnoses mentioned.Instead, it could be argued that even though most cancer trajectories roughly follow the same path-from diagnosis to treatment to survivorship or end of lifethe specific diagnosis generates its own path based on symptoms, treatments and duration of and emphasis on certain phases.Thus, a universal informational and emotional support model provided to family members of, for example, breast, prostate and colorectal cancer may not be sensitive enough to meet all their varying needs.Instead, it may be more useful to draft the trajectory of the diagnosis in focus before designing a support model.However, future reviews may want to discuss the gains of a universal versus a diagnosis-specific support model by using it as a potential moderator of intervention efficacy.
A final reflection concerns the multi-component design of the support models.The models described in this scoping review, apart from one (Chambers et al., 2014), were provided over numerous sessions using varying delivery modes (Table 4).A multi-component design complicates effect evaluation due to difficulties in separating the components (Ferrell & Wittenberg, 2017;Treanor et al., 2019).Hence, if aiming to establish intervention efficacy, future models of support would benefit from refining their interventions-even though this may require the support model to target fewer of the unmet needs described by Lambert et al. (2012).Further, the multi-component design  Ferrell and Wittenberg (2017) conclude that the support models evaluated in research trials need realistic translation to be applicable to clinical setting.To apply a perspective of applicability of the support models included in this review, we consulted contact nurses and provided them with the preliminary results.As previously mentioned, the contact nurses found the models of support difficult to apply in their clinical practice.In addition, they did not feel confident in providing the emotional support described in the models (e.g.coping strategies and mindfulness exercises) due to lack of knowledge and experience.
Consequently, considering the specific clinical preconditions and context of the intervention provider seems essential, especially since the intervention providers have shown to greatly influence whether a support model has an effect or not (Zimmermann et al., 2007).

| Limitations
In line with the methodology of a scoping review, the aim of this

| RELE VAN CE TO CLINI C AL PR AC TI CE
When designing a support model to be used in clinical practice, knowledge from the literature on both the needs of family members and on lessons learned from existing support models (preferably through a scoping review) should be incorporated with the prerequisites of clinical practice.In this scoping review, contact nurses were asked about the applicability of the models of support identified during the analysis.Unfortunately, they were all found to be too far from their clinical practice to be usable.Hence, instead of targeting multiple needs applicable on multiple cancers and trajectory phases, the models of support should be downsized to become more effective.For example, at the time of diagnosis, they could focus on the most frequently reported need of family members of all cancers-information.Additionally, involving the presumptive intervention providers and learning from the existing support models about the variety of ways support can be provided

R E FE R E N C E S
The search strategy was developed in collaboration with a research librarian well versed in using research databases.An initial broad search (e.g. next of kin, support and cancer) was conducted in PubMed to inform the subsequent searches.In accordance with recommendations from the research librarian, systematic searches were conducted in PubMed, Cumulative Index of Nursing and Allied Health and PsycINFO between November 2019-February 2020 (File S2), with no limitation in publication year or study design.An additional search was conducted in September 2020.
To maintain consistency across the different qualitative and quantitative approaches, the research team designed a data extraction form when writing the protocol based on the research questions and the recommendation by the Joanna Briggs Institute (2015).To validate the extraction form for utility and completeness, three members of the research team independently piloted the form on a study of each study design included.This validation was conducted at the beginning of the inclusion process and resulted in no corrections needed.Apart from study characteristics, the following information was extracted: support model (type of support, delivery mode and intervention provider), phase of the trajectory (e.g. at diagnosis or during treatment), aim of the support and participating family member.
included studies contain the designs of randomised controlled trials (n = 17), quasi-experimental trials (n = 11), a combination of quasi-experimental and qualitative focus group interviews (n = 1) Psychoeducation support models related to Lamberts et al.'s (2012) unmet needs Psychoeducation support models related to Given et al.'s (2012training support models related to Lamberts et al.'s (2012) unmet needs Caregiver training support models related to Given et al.'s (2012support models related to Lamberts et al.'s (2012) unmet needs Psychological support models related to Given et al.'s (2012) phases of cancer trajectory Birnie et al. (2010) raises the issue of applicability to clinical practice.A meta-analysis of randomised trials of interventions with family caregivers of persons diagnosed with cancer (Northouse et al. 2010) indicates that psychosocial interventions seldom move from academic institutions to clinical practice-thereby only rarely reaching the family members in need.
scoping review was to map the existing literature on support models provided to family members during the cancer trajectory.The chosen design enabled the identification of studies of various methodological approaches and, consequently, a broader range of support models.The strengths of this scoping review lie in the inclusion of an experienced librarian in the phase of developing the search strategy in addition to the research team being experienced in both familyfocused care and clinical cancer care.Furthermore, the development and use of a standardised charting form as well as a quality checklist adapted form published tools enabled systematic data extraction and coherence across study designs.All studies were reviewed independently by pairs.In addition, the process of reviewing and discussing papers was iterative, in accordance with Arksey and O'Malley's framework.However, one limitation in this scoping review was the exclusion of studies in other languages than Swedish and English, meaning relevant support models may have been missed.Furthermore, the encountered difficulties related to the broad inclusion resulting in the usage of search terms of double meanings (e.g.'relative' and 'supportive care'), leading to imprecise database searches.The combination of 'family' and 'cancer' search blocks led to titles referring to heredity or screening or qualitative studies describing the needs of the family or the development, translation and/or validation of different instruments for needs, quality of life or anxiety assessment.In addition, there were multiple studies focusing on experiences or interventions in the paediatric context or with children as the next of kin.Furthermore, the inconsistency in the definition of 'support' (e.g.supportive care or psychosocial support) may have affected the searchers.For example, a concept analysis byHui et al. (2013) shows how the meaning of 'supportive care' has ranged from the use of hydroxyurea for persons with acute leukaemia to survivorship care.In addition,Hodges et al. (2011) argue against using 'psychosocial' as the only search term for support in reviews due to its lack of consistency.Subsequently, this scoping review chose a broad definition of 'support', focusing on all non-pharmacological interventions and actions taken by healthcare professionals towards a family member of a person diagnosed with cancer with the purpose of improving the situation for the family member.Altogether, this led to a search strategy that resulted in n = 6,140 titles that had to be reviewed manually.However, this was considered necessary as the research team together with the experienced librarian could not come up with a more precise search strategy.6| CON CLUS IONWhen developing a support model aiming to meet the family members' needs, it appears essential to make informed decisions on whether to have a universal or specific focus concerning both diagnosis, and trajectory phase as well as the context.In accordance with the findings of this scoping review, a tailored individualised support in its unique circumstances provided to family members seems likelier to meet their needs during the cancer trajectory than an universal support model.The latter targets multiple unmet needs, includes multiple diagnoses, is provided across multiple trajectory phases and contains multiple components.Although a universal design might be advantageous in a clinical trial, it appears to lack sensitivity towards both what family members may need and when during the trajectory they need it as well as difficult to apply in clinical practice.Hence, a tailored support model developed in collaboration with both family members and the clinical intervention provider would probably increase the likelihood of meeting the needs of families of persons with cancer even further.
may allow the support to more easily reach the targeted family members.Further, the support models should be compliant with the specific diagnosis trajectory, tracing involved family members to define what is needed and when.Finally, structured assessments of family members' needs could be conducted regularly throughout the trajectory, from diagnosis to treatment to survivorship.Such assessment would help to determine the trajectory of the specific diagnosis and simultaneously enable identification of divergent or changing needs.Hence, it would allow for individualisation of the support by tailoring it to the needs of the specific family member.CO N FLI C T S O F I NTE R E S TThe authors report no conflicts of interest.AUTH O R CO NTR I B UTI O NDesign, search strategy and conceptualisation and drafting the manuscript: MS.Conceptualisation of the review design, inclusion and exclusion criteria and providing feedback on the methodology and the manuscript: AW, JJ and MB.All authors gave approval to the publishing of this protocol manuscript.

E 3 Included studies Author, year, country Aim
Kuijer et al. (2004)1), USA To determine the impact of a 16-week supportive nursing intervention on caregivers of patients with newly diagnosed cancer 22.Kuijer et al. (2004), NZ Explored to what extent a decrease in perceived equity after the intervention could predict relationship quality and psychological distress directly after the intervention and 3 months later To explore to what extent a decrease in perceived equity after the intervention could predict relationship quality and psychological distress directly after the intervention and 3 months later Schellekens et al. (2017) To determine the efficacy of a novel partner-assisted emotional disclosure intervention in a sample of patients with gastrointestinal (GI) cancer 27.Porter et al. (2011), USA To determine the efficacy of a caregiver-assisted CST protocol in a sample of patients with lung cancer 28.Roberts and Black (2002), Australia To report the results of an evaluation of an Australian education and support programme for individuals with cancer and their family and friends-the Living with Cancer Education Programme (LWCEP) 29.Schellekens et al. (2017), Netherlands he aim of this study was to examine the effectiveness of MBSR added to CAU compared to CAU alone to reduce psychological distress in patients with lung cancer and/or their partners he aim of this study was to examine the effectiveness of MBSR added to CAU compared to CAU alone to reduce psychological distress in patients with lung led self-management intervention with feedback about patients' levels of distress and stress reduction instructions.The session focused on eliciting concerns, ensuring access to relevant cancer information, offering brief psychoeducation, orienting participants to the psychological self-management strategies provided in the accompanying resource kit, and, where relevant, discussing specific strategies matched to participant need.Psychoeducation about the psychological impact of cancer, coping and stress management skills, problem solving, cognitive therapy, and enhancing support networks.Participants were given assigned behavioural homework for each core component.Additional components for specific treatment effects were included where relevant (e.g.pain, sleep disturbance, fatigue) The interventions were telephone health education (THE) and telephone or videophone interpersonal counselling (TC or VC).Participants received pamphlets about (1) breast cancer terminology, (2) treatments, (3) side effect management, (4) nutrition, (5) physical activity, and (6) resources.The TC and VC addressed (1) mood and affect management, (2) emotional expression, (3) interpersonal communication and relationships with family and providers, (4) social support, and (5) follow-up, resources and referral to any support services that might be required (e.g. Self-help intervention based on Acceptance and Commitment Therapy (ACT) and self-compassion.In each lesson, one particular theme is discussed.Every lesson starts with a short text about the topic of that lesson.The texts are enriched with short psychological exercises.Each lesson also contains a mindfulness exercise that can be read or downloaded as a mp3-file.Partners also receive practical information, tips, and references to relevant websites.Users can freely decide if they want to receive text messages (one per lesson) with short inspiring texts.Moreover, partners have different options to come into contact with peers