Building a Cancer Coalition: A co-design framework for evaluating regionalism in the South Pacic

Cancer is a signicant problem for the South Pacic region due to a range of complex and unique health challenges caused by shared social, environmental and economic factors. Currently gaps in diagnosis, treatment and palliative care are signicant, and while governmental commitment is strong, economic constrains limit health system strengthening. Collaboration, alliances and partnerships in cancer control have been successful in resource constrained settings. A regional approach has therefore been recommended as an effective solution to addressing many of the challenges for cancer control in the South Pacic. However, comprehensive and appropriate information detailing how to effectively scope and establish a multi-national or regional coalition is scarce. This study therefore aimed to 1) create a Coalition Development Framework, and 2) use the Framework to co-design a South Pacic Cancer Control Coalition through consultation with key cancer control stakeholders working within Fiji, New Caledonia, Papua New Guinea, Samoa and Tonga. Analysis of the consultations with key cancer control stakeholders identied overwhelming appetite and support for a South Pacic Cancer Control Coalition. This paper details the following six coalition development outputs: coalition design and purpose, strategic imperatives, structure, South Pacic foundations, barriers and facilitators to coalition establishment and ongoing operations, priorities for action. Concurrent evaluation of the Coalition Development Framework using theory of change revealed the framework to be an effective mechanism to drive engagement, discovery, unication and action in alliance-building.


Background
Globally, cancer is one of the leading causes of morbidity and mortality, with rates set to continue to rise worldwide 1 . The Paci c is particularly impacted by non-communicable diseases (NCDs) such as cancer, due to a range of complex and unique health challenges caused by shared social, environmental and economic factors 2 . The region comprises 22 Paci c island countries and territories (PICTs) which sit within the geographically and culturally distinct island groupings of Melanesia, Polynesia and Micronesia, with a combined population of approximately 10.5 million [2][3][4][5] . The term 'South Paci c' refers to the region and nations located south of the equator. However, the true scale of cancer burden in the South Paci c is di cult to ascertain due to signi cant challenges with obtaining, measuring and recording accurate diagnostic data [2][3][4][5][6][7][8] . The few national cancer registries in existence are cited as fragmented or dormant; with the notable exception of New Caledonia [3][4][5][6][7] .
Factors which drive high rates of cancer within the region include, but are not limited to: geographic remoteness both within and between PICTs; high rates of cancer risk factors such as obesity, and tobacco and alcohol use; strong in uence from unhealthy commodity industries; food insecurity and lack of access to healthy produce; fragile developing economies; and impacts from climate change [2][3][4][5][7][8][9][10] . In addition, many PICTs are undergoing epidemiological transition, meaning communicable disease risk factors are still highly prevalent 3,4,11 . Together, these factors create a challenging setting for governments in the establishment of effective cancer control policy, infrastructure, workforce and activity 2,5,10 .
Many PICTs in the South Paci c are unable to provide adequate treatment or palliative care services for people diagnosed with cancer 3-5, 12, 13 . Patients often cannot access chemotherapy, or even analgesics that are on the World Health Organization's (WHO) essential medicines list 4,13,14 . Provision of basic cancer care in low resource settings still requires speciality services, highly trained workforce, and costly infrastructure and consumables [15][16][17] . While strengthening cancer treatment is on the policy agenda for many governments within the region, economic constraints are cited as the key barrier 2,18 . Small population sizes in many PICTs further limit the feasibility of implementation of such specialist services at a national level 2,4 . Population health screening, as a key cancer control activity, can increase survival for breast, colon and cervical cancers 15,18 , however, the majority of South Paci c countries do not have adequate histopathology services to diagnose and stage cancer, meaning that many cases are unable to be detected or correctly diagnosed [3][4][5]8 . Further, without adequate treatment, diagnosis can create an ethical dilemma 19 . As noted, cancer risk factors are prevalent in the South Paci c, however population health literacy regarding cancer risks and identi cation of symptoms remains low 13, 20, 21 .
A regional, collaborative approach has been recommended as an effective solution to addressing many of the challenges for cancer control in the South Paci c 4,5 . Collective approaches, such as coalitions or other voluntary multi-stakeholder agreements created to improve mutually agreed healthcare goals, have been demonstrated to be effective and feasible in many settings around the world 13,17,22,23 . Diverse multi-sectoral and multi-national partnerships are likewise considered essential by both the WHO and Union for International Cancer Control when developing and implementing comprehensive cancer control plans and strategies 17,24 . In the Paci c there exists precedence for regional comprehensive cancer control planning. The United States of America (USA) Centre for Disease Control and Prevention (CDC) funded the establishment of the Cancer Council of the Paci c Islands (CCPI) to facilitate a regional multinational coalition of all USA-a liated Paci c island jurisdictions in the early 2000s [25][26][27][28] . Since its inception, the CCPI has successfully developed effective cancer control strategies for the USA-a liated islands, including national and regional registry establishment, population cancer screening, and workforce capacity development [25][26][27][28] . However, despite the existence of successful regional healthfocused alliances, comprehensive and appropriate information detailing how to effectively scope and establish a multi-national or regional coalition is scarce.
This project consisted of two parts: 1) the creation of the Coalition Development Framework, and 2) use of the Framework to co-design a South Paci c Cancer Control Coalition through consultation with key cancer control stakeholders working within Fiji, New Caledonia, Papua New Guinea, Samoa and Tonga. Project aims comprised: i) determine interest for a regional South Paci c Cancer Control Coalition, ii) identify essential requirements for successful Coalition establishment and ongoing sustainability, iii) con rm national and regional cancer control priority areas, barriers and facilitators, and iv) collate learnings to inform continued use of the Framework to develop cancer control coalitions.

Methods
Paci c leadership and oversight Importantly, the drivers for this project were key civil society organisations (CSOs) based in Fiji and Samoa: the Fiji Cancer Society (FCS) and Samoa Cancer Society (SCS) who contacted SA as CEO of Cancer Council Australia (CCA) and through this AH undertook this work as part of her Master of Public Health Studies with the University of Melbourne. FCS and SCS also work in close partnership with national Ministries of Health, tertiary and primary healthcare services, and other health and cancerspeci c non-government organisations to deliver cancer control services in the Paci c. Given the challenges and limitations faced by these nations regarding cancer control, the FCS, SCS and CCA are committed to investigating and supporting a regional approach within the South Paci c.

Paci c Project Advisory Committee
A Paci c project advisory committee was established prior to the commencement of this project to provide stakeholder input into the design and operations of this study. The committee comprised members from the FCS, SCS and additional cancer control specialists with experience in regionalism and coalition-building based in Fiji, Samoa, New Zealand, Hawaii and Australia. This group was responsible for ensuring Paci c leadership, oversight, and input into: project and Framework design, data collection and analysis, and project reporting and dissemination.

Creation of the Coalition Development Framework
The rst step in creation of the Coalition Development Framework comprised determining and assessing relevant materials currently in existence. A scoping review of publications describing the elements, processes, and activities required for the establishment and ongoing success of operational health coalitions and multi-stakeholder partnerships was conducted. This review identi ed a small number of peer reviewed journal articles and one white paper publication. Content analysis of these documents identi ed a range of key elements and activities which informed the Coalition Development Framework (see Table 1.) Identi ed elements and activities were then synthesised into the proposed Coalition Development Framework using program logic. The Coalition Development Framework is proposed as a comprehensive, evidence-informed step-by-step guide to assist leading stakeholders in the design and development of health coaltions (see Table 2). Four key phases across all resources and materials were identi ed as essential: Engagement; Discovery; Uni cation, and Action (see Fig. 1).
The Engagement phase focuses on stakeholder engagement, Discovery focuses collection of information and stakeholder perspectives, Uni cation focuses on synthesis of ideas into coalition design drafting and planning, and Action refers to nalisation of coalition structure and determination of strategic priorities. Each phase includes a dedicated aim and associated deliverables in the form of discrete activities and outputs recommended for completion, from conducting a situational analysis to a proposed coalition organisational design and model. Importantly, once commencing, each phase should continue until the Coalition is established, as activities and processes are non-linear, and involve iterative and continuous connection to facilitate development, information synthesis and consensus-building (described in Table  2). Co-design was employed as the overarching methodology to support coalition development, as the principles and practical application of co-design require active and democratic involvement of a diverse range of participants from concept creation to implementation of solutions developed in response to shared challenges 31 . In particular, co-design has been recognised as a novel methodology increasing useful for engaging stakeholders to nd solutions to complex problems, particularly in the policy context 31 .
Data collection occurred over six weeks from September 10th to October 29th, 2019. During this time, the research team were based in Fiji with the FCS as they commenced stakeholder consultations, and travelled to Tonga and Samoa to meet with participants face-to-face for interviews. Participants located in New Caledonia and Papua New Guinea were contacted via videoconference.

Consultations
Stakeholders working in a cancer control-related role at either a national or regional level in the following organisations were identi ed through the project advisory committee as potential Coalition partners: National Cancer-speci c civil society organisations (CSOs) or non-government organisations (NGOs)

Ministry or Department of Health
Health Services (primary/tertiary/community/public/private)

Regional Development Partners
The project advisory committee invited all identi ed stakeholders to take part in a consultation to share their perspectives regarding the establishment of a cancer control coalition for the region. If the stakeholder was interested in discussing a regional Coalition an interview time was arranged.

Informed consent
As the project concurrently involved evaluation of the coalition framework, potential participants were also emailed with a research participant information and consent form (PICF). At the time of the interview, if the participants agreed for their interview data to form part of the research study, they were asked to sign a copy of the consent form and were given a copy to keep.

Evaluation
A semi-structured interview schedule was developed to facilitate evaluation of the framework, with 12 questions and associated prompts mapping to the four stages within the Coalition Development Framework of engagement, discovery, uni cation and action. Stakeholders who agreed to participate in the research project had their consultations audio-recorded and then transcribed to facilitate analysis.

Field notes
Comprehensive notes were created to supplement interview data and to document information relevant to evaluation of the Coalition Development Framework.

Demographic data
Three demographic questions were also included at the start of the interview comprising participant role, organisation type, and country of focus.

Data Analysis
Quantitative data Demographic data were analysed descriptively, using means and standard deviations, or frequencies and percentages where appropriate.

Qualitative data
Consultation transcripts and other text documents ( eld notes etc.) were uploaded to NVIVO12 for analysis using interpretive description methods 32 . Consultations were analysed in alignment with the Coalition Development Framework phases to facilitate evaluation. Analysis was inductive, with data initially described in broad, generic codes, which were then iteratively and re exively re-labelled and interpreted as categories and themes were developed 18 . Analysis was considered nalised once themes reached a balance of both meaningfully representing stakeholder perspectives, and addressing the research objectives 33 .

Engagement
A total of 43 stakeholders working directly within cancer control, and other key leaders or actors ancillary or adjacent to cancer control within the health or government sector were identi ed and contacted. Of these, 35 were available and interested in taking part in a consultation to discuss a South Paci c Regional Coalition, and all were happy to participate in the research evaluation of the framework. Table 3 summarises the key characteristics of stakeholders interviewed. Role title and organisation name are not reported to preserve participant con dentiality. Papua New Guinea 1 3 The majority of the stakeholder consultations (n = 29, 83%) were conducted face-to-face in country; the remainder were conducted via telephone or videoconference (n = 6, 17%). Consultations ran for an average of forty-two minutes (SD = 20mins), generating a total of twenty-three hours of data. Of the eight stakeholders identi ed but not consulted: three were on leave for the duration of the project, three did not respond to emails, one did not have time-capacity (though was supportive of the project), and one was cancelled due illness and was unable to be rescheduled.

Discovery, uni cation and action
Analysis of the consultations identi ed the following six coalition development outputs: Diagrams were developed to summarise and describe each stakeholder output, with a composite of all outputs diagrams presented in Figure 2. A comprehensive situational analysis was completed and will be reported elsewhere. Development of the nalised coalition model will also be reported elsewhere. .

Output 1: Coalition Design and Purpose
All thirty-ve stakeholders supported the development of a South Paci c Cancer Control Coalition. Stakeholders suggested the rst step in the development of the Coalition should be the establishment of a central hub or secretariat, such as a regional Cancer Agency. This would reduce the burden of administration, management, or potential perceived 'ownership' from any one country.

Output 2: Strategic Imperatives
Stakeholders agreed that the Coalition will need to have a clear mandate, purpose, and objectives. These will be essential to ensure that the Coalition is effective and not passive or tokenistic. The following strategic imperatives articulate both key objectives, and what individuals and organisations would expect from membership (see Figure 2).

Partnerships
Facilitate community and regional networks, and foster partnerships between individuals and organisations.

Data-driven to respond to South Paci c needs
Respond to issues identi ed through data collected by members in region-speci c clinical settings and by cancer registries.

Cancer-speci c, long term and sustainable change
Increase the visibility of cancer in the South Paci c. An established coalition to assist CSOs leverage long-term and comprehensive health system change.

Advocacy and Empowerment
Provide a powerful, uni ed, and respected voice to the political arena to bring cancer to the forefront of discussions; and empower national agencies in advocating in the international setting.
"This Coalition can be a source of empowerment to the national agencies on cancer prevention, cancer treatment diagnosis, cancer medicine" Participant 4

Sharing and Capacity-building
Enable accurate information about cancer care organisations, technical expertise, facilities, and healthcare workforce to be effectively shared and communicated throughout the region.
A regional approach to cancer control would increase regional capacity, knowledge, and resources.

Strengthen, complement, streamline
Enable a strategic approach to all activities from education, awareness, and outreach, through to screening, treatment, and palliative care. All current activities would be strengthened and streamlined through partnerships and coordination. For the Coalition to be viable, it would need to work hand in hand with governments and receive endorsement from governments and Heads of Health at a national and regional level.

Supported by Development Partners
International development partners within the region (such as WHO, SPC or UNICEF) were considered essential for impact and regional cohesion. It was also felt that these organisations would be well placed to provide valuable support for the Coalition in the form of technical and practical assistance. To enable the region to grow and develop, stakeholders strongly felt that Paci c peoples need to be given ownership and leadership positions within the Coalition. Stakeholders emphasised that South Paci c island culture, cancer burden and politics are best understood and navigated by local people.

Maintain Independence
Stakeholders agreed that the Coalition should maintain independence and sit outside the governmental and clinical arenas. Impartiality and distance were seen as essential in order to effectively set policies, evaluate activities, and work in partnership with governments, health professionals and health services.
"My observations and my working with different programs or projects have found that there's always a lot of emphasis on bringing external consultants. But I feel to work as a program, it needs to sit with a country; and discover how the country can develop their own capability." Participant 12

Strong Leadership
A steering committee to drive leadership and governance was suggested, comprising representatives from cancer organisations and health services. A pre-de ned time-period for serving on the committee would ensure that people were willing to take on this role, and to nurture innovation and change.

Dedicated paid staff
Given current capacity constraints faced by all individuals working in cancer in the South Paci c, stakeholders emphasised the need for full-time, paid local staff able to focus solely on the Coalition.
However, voluntary assistance and technical support in the development and running of the coalition were recommended as an opportunity to grow South Paci c capacity without jeopardising ownership.
Output Five: Barriers and facilitators to coalition establishment and ongoing operations Potential barriers and threats to coalition development and activities were outlined by stakeholders to highlight the challenges which need to be thoughtfully considered, discussed, and addressed. Equally, existing and potential future facilitators were also described, so that development of the coalition can leverage or harness these opportunities where possible and appropriate (see Table 4).
"The big challenge with any kind of any network or coalition, whether within country or across countries is keeping it going. Everybody's red up at the beginning, but nobody has time to do because everybody's very busy. And nobody's willing to put the time into doing it. And so gradually, it fades away." Participant 16 Health system design and cancer burden Health systems, cancer burden, risk factors and technical capacity in the region are similar, forming strong facilitators for regional activity and transferrable learnings.

Lack of adequate data
Paucity of health and cancer-related data in all countries is a signi cant barrier to Coalition establishment and development of targeted cancer control activities.
Established multi-national programs and goals Existing international indicators, goals and programs can be leveraged by the Coalition to facilitate governmental support, international funding, partnerships, and national accountability.

Lack of dedicated CSOs or government teams
No governments in the South Paci c have dedicated cancer teams. Many nations do not have dedicated cancer CSOs. This will be challenging for the Coalition to navigate relationships with governments, and maintain cancer-speci c activity.

Established relationships, interest and capacity
Existing partnerships between CSOs, Ministries of Health and public and private healthcare services can be developed and strengthened with the introduction of a Coalition.

Politics
Existing political alliances or issues between nations and regions require the coalition maintain neutrality. Concerns regarding inter-organisational politics and agendas are also present.

Embracing technology
Use of technology can be employed by the Coalition to overcome geographical and travel limitations, support health professional networks, telehealth, capacity development; and communication with international partner organisations.

Sustainability, funding and travel
Funding is scarce and existing CSOs and governments are wary of new competitors. Travel can be prohibitively expensive in the Paci c, and logistically di cult as many countries do not have direct transit routes.
Output 6. Priorities for Action Experts agreed that priority areas for the coalition must hit a balance between what is meaningful for patients and healthcare professionals, but achievable for NGOs/CSOs or health services and systems.
The following priorities were identi ed as being essential for concurrent health system strengthening across the cancer control continuum: 1. Develop Standardised Systems for Quality Data: Stakeholders agreed that a priority is to establish high quality methods to facilitate data collection, monitoring and evaluation, storage, sharing and research at national and regional levels. . Pooled Procurement of Anti-Cancer Drugs: Investigate the potential for a regional approach for the procurement and management of anti-cancer drugs.

Framework Evaluation
Logic model mapping determined that the proposed framework inputs and activities comprising situational analysis, stakeholder consultations, information synthesis and consensus building resulted in the anticipated outputs and outcomes (see gure 3). As stakeholder ideas for Coalition strategy, structure and purpose began to take shape, these were then discussed as part of the ongoing consultation process.
New information and perspectives were iteratively incorporated until concrete ideas were nalised, and consensus reached.
Importantly essential considerations regarding use of the framework were identi ed through this evaluation process. Speci cally, diversity of stakeholders and key actors within civil society, the private sector and the state was needed to optimise contextual understanding, and to determine priorities which ensure a whole-of-system approach. Each stakeholder grouping carried unique perspectives and situational understanding which would have been missed if diversity in stakeholders was not purposively sought. Likewise identi cation of key actors and power-brokers was important, especially to inform targeted engagement where appropriate. In some nations, hierarchy and ceremony are necessary contextual elements to understand and adhere to, particularly as some stakeholders will nd it inappropriate to engage unless formal, culturally-speci c processes are followed. Identi cation and engagement with diverse stakeholders and key actors also ensured that the proposed governance and organisational design is inclusive of all interested stakeholders; and accounts for managing or leveraging commercial interest and/or competing priorities.
Early conversations with key stakeholders regarding leadership and ownership were also important. For many people, understanding that the FCS were the key drivers for this work, and that Paci c leadership and ownership was planned and valued, facilitated their engagement and support. Further, as information regarding proposed coalition structure, purpose and design became clear, mechanisms for action were likewise able to be articulated, as leadership and ownership had been to some degree decided upon.
Other key ndings highlighted that concurrent mapping of key barriers and facilitators ensured that activities are relevant, achievable and measurable, and so issues could be discussed with relevant actors to ensure importation into both organisation design and identi cation of priorities for action. Finally, including discussion regarding establishment and implementation of monitoring and evaluation frameworks also meant that these key elements were included as essential elements of coalition design rather than as an afterthought.

Discussion
Stakeholder consultations and perspectives provided comprehensive information to support the feasibility of a South Paci c Cancer Control Coalition and con rmed the appropriateness and usefulness of the proposed Coalition Establishment Framework. Key barriers and facilitators were outlined, and stakeholders described in detail current cancer control activities, and their perceptions regarding priority actions for the Coalition in strengthening cancer control for the region. Coalitions emphasise that civil society leadership has been crucial in their success, and that identi cation of regional priorities is best commenced where change is aimed to occur, at the community level 13,27 . Ensuring community is at the heart of all coalition activities from identi cation of priorities to capacity building to advocacy, has been further cited by several other island nation health coalitions as a key factor behind their success 23,26,27,34 . Civil society can also be highly effective and in uential at a policy level in cancer control advocacy and universal health system strengthening 35 . Linking back to the Coalition Development Framework, these ndings emphasise the importance of stakeholder determination of leadership, organisational design, and governance in ensuring alignment with local context ( gure 3).
Building local health capacity and allowing organisations and nations to remain autonomous was likewise emphasised by stakeholders; a message consistently reiterated in the literature by those working in health within the Paci c 5, 28, 36 . Underpinning these messages is an issue unfortunately arising in low and middle income countries (LMIC), where global health programs eschew local expertise due to issues such as long-standing colonialist attitudes or belief in the superiority of western knowledge and methods 37 .The CCPI highlights that the essential element behind their success was ensuring that dedicated Paci c staff were hired and appropriately paid to run the Coalition 26 . From these learnings it's clear that assistance provided by development partners in supporting the Coalition must acknowledge past issues of colonisation, respect Paci c local knowledge, leadership and decision-making, and ensure funding is directed to ensuring Paci c islanders are recruited to key coalition roles. Dedicated and highly quali ed Paci c personnel with strong knowledge of local culture and context will be needed to achieve robust scienti c excellence in coalition activities.
Coalition Purpose: Managing politics and power Managing potential government desire to lead or override CSO agenda-setting, as documented in other health Coalitions, are important considerations when designing the Coalition organisational design and structure 17,27 . This was noted by stakeholders as a key potential barrier in Coalition establishment for the South Paci c. Harmonising national priorities with regional Paci c coalition priorities has proven di cult in other contexts such as the Alliance of Small Island States in advocating for climate change, where differences in country versus regional agendas limited action on key issues 38 . Therefore coalition goals must be broad enough to facilitate regional action, yet still allow for connectedness with individual state and national goals and strategies. Regionalism requires support from health leaders (both government and private sector) in all countries involved, as these leaders are skilled at identifying and advocating for what is needed in the Paci c context, and little momentum is likely to occur without their support 2,5,17,36 . Likewise ensuring that proposed agendas are realistic and well-aligned with governments is important.
However, management of power imbalances, or existing political issues, particularly in terms of agendasetting is an important consideration 17,38 . Levels of political in uence and power can impact a nation or organisation's ability to effectively bargain or advocate for their needs, and poor political relationships can impact on successful collaborative decision-making 17,38 . Learnings from the CCPI highlighted that good leadership utilises a collaborative partnership structure, which will be important to consider when creating organisational design to ensure that stakeholder relationships remain cohesive and integrated across all hierarchies, systems and cultures, and to overcome potential power imbalances or political issues 26, 27 .
Shared Context: overcoming barriers by maximising existing strengths Acquiring funding is challenging in a multi-national environment, and as identi ed, the coalition is likely to be seen as a competitor for resources within an already scarce funding environment 27 . However, capitalising on identi ed facilitators, such as existing and diverse stakeholder relationships, can widen the range of funding opportunities outside of usual health avenues 17 . Likewise innovative use of technology can overcome multiple sustainability issues such as: travel, funding, and workforce capacity barriers endemic to the Paci c 2 . Health technologies such as telehealth provide evidence-based solutions to bridge the gap, by facilitating access for patients to healthcare irrespective of geographic location 39 .
Other initiatives, such as project ECHO, which use video-conferencing and other online communication platforms to facilitate clinician capacity-building, mentoring and training have been highly successful in strengthening cancer control in other LMICs 39 .
Limitations Figure 2 Composite of Coalition Development Themes