Key stakeholders’ perspectives on implementation of risk-based population breast cancer screening in Australia – “We can either get on the bus or get under it”

Abstract

Background

Risk-based breast cancer screening would be a dramatic shift from the current one-size-fits-all model to a tailored approach where screening modality and frequency is directed by individual risk. This project assesses what key stakeholders, defined as those holding managerial and decision-making roles within BreastScreen, consider the issues are with implementing a risk-based approach to screening.

Methods

A qualitative approach was undertaken, recruiting participants through professional networks with interviews guided by the Consolidated Framework of Implementation Research (CFIR). Participants were key stakeholders defined as those managing, overseeing and influencing Breast Screen throughout Australia. Data were deductively coded against a CFIR-informed codebook, followed by content analysis per CFIR domain.

Results

Twenty interviews were conducted with 21 participants. 144 initial codes consolidated into 17 final themes. Key stakeholders are supportive and optimistic about risk-based screening in principle; however several issues exist, including knowledge gaps precluding support of evidence-based implementation. Concerns about worsening inequities within screening, cost and communication with clients are major issues key stakeholders consider important to address in the planning and implementing a change to the program.

Conclusions

Key stakeholders in Australia were overwhelmingly enthusiastic about the benefits of a risk-based approach however there are concerns about risk assessment utility, cost and the potential risk to equity in the program. Systematic assessment of these concerns will be required to facilitate successful change to the well-established breast screening program in Australia should risk-stratification be undertaken.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13690-025-01690-5.

Text box 1. Contributions to the literature
• There is widespread interest and support for a shift towards a risk-based approach to population breast cancer screening among those currently managing, funding, influencing and researching breast screening in Australia.
• These key stakeholders identified fundamental issues to overcome before implementation of a risk-based approach is feasible namely the incompatible funding model, availability of a robust and use of an easy to administer risk prediction tool and practicalities of performing screening other than mammography in remote locations.
• Systematic consideration of these issues is required prior to the implementation of a risk-based approach to breast cancer screening in Australia.

Background

After 30 years of population-based breast cancer screening in Australia, an improved understanding of the contribution of breast cancer risk factors enabling personalised risk assessment has led to consideration of a risk-based screening approach [24, 32]. Risk-based (tailored, personalised or stratified) screening would change the current one-size-fits-all model by allocating differing modalities and/or frequency of screening as directed by individualised risk. Several trials are underway internationally assessing this approach such as the WISDOM study in the United States, which incorporates polygenic risk scores (PRS) into risk assessment [12] and MyPebs recruiting throughout multiple sites in Europe [27]. BC-PREDICT offered participants recruited through the UK National Health Service (NHS) personalised risk assessment using the Tyrer-Cuzick risk calculator and identified a huge range from < 10% to 76.8% uptake in the program depending on modality of recruitment (onsite compared to paper) [13]. PERSPECTIVE, underway in Canada, compares different stratification methods including with and without PRS which will give valuable information about the feasibility and accuracy about these methods [4]. While there is no current trial underway in Australia, significant work has been undertaken via the ROSA (Roadmap to Optimising Screening) project to identify a roadmap to risk-based screening [23, 31].

A summary of the current Australian BreastScreen program is outlined in Table 1. This current model has not significantly changed since inception and is not without issues [3]. These include interval cancers (cancers which are detected between biennial screening mammograms) and over-diagnosis (detected of cancer that would not have become clinically significant without screening) [1]. There is optimism that a risk-based approach would minimise the known harms of screening by reducing interval cancers by offering higher risk participants more frequent and more sensitive screening and lower risk participants could potentially be screened less. Another proposed benefit would be less treatment intensity with a higher proportion of early-stage cancers detected.

Table 1.

Characteristics of the population breast screening program in Australia BreastScreen

BreastScreen Australia
WHO?	Women aged 50-74yrs actively invited
	Women 40–50 and over 74 years welcome to attend
	Current national participation rate for target age range 51.7%. Significant variation in participation rates throughout the country 34.1–57.9% (Health and Welfare 2024) [17]
WHAT?	2 view 2D mammography
WHERE?	Fixed screening sites primarily. Regional/remote areas serviced via mobile vans
WHEN?	Every 2 years for most clients Provisions for annual mammography for high-risk women as deemed by family history or prior proliferative breast disease (e.g. atypical ductal hyperplasia, lobular carcinoma in situ)
HOW?	Each state and territory service functions independently but with requirements to adhere to the National Accreditation Standards (NAS)[21]
	Funding through both Federal and State Government
	Invitations sent using data from Australian Electoral Roll

Risk-based breast cancer screening requires risk assessment followed by tailoring of screening tests as per individualised risk. Breast cancer risk assessment may include personal risk factors such as family history or reproductive history, radiological findings such as mammographic density, plus or minus genomic information. Several risk assessment tools exist combining a variety of these factors with no single test performing consistently well across different cohorts, but most can reliably identify people at high risk [34]. Tailored screening may involve additional screening (such as contrast mammogram, breast ultrasound or MRI) potentially at higher frequency for people at higher risk or reduced frequency of screening for lower risk. There are numerous possible permutations of this which could also embrace artificial intelligence (AI) as exampled by this Swedish trial using a mammographic AI tool to allocate people with high mammographic density to MRI screening [29].

Current BreastScreen Victoria clients are highly supportive of a move towards risk adjusted screening [20], aligning with overseas experience from both hypothetical and real-life scenarios [4, 26]. However, the practicalities of implementing modifications requires an array of components including support and engagement from stakeholders responsible for managing, funding, and coordinating the program.

Perspectives of stakeholders from other countries has highlighted issues around the need for more evidence, equity and the importance of communication [11, 19, 22, 30]. These studies have recruited mostly healthcare professionals actively involved in the breast screening programs of these countries. Given the paucity of Australian published literature addressing the views of those responsible for breast screening programs and their views on implementation of risk-based screening a qualitative approach guided by the updated Consolidated Framework of Implementation Research (CFIR) [7] was employed. The aim was to elicit and analyse key stakeholder views on the implementation of risk-based breast cancer screening in Australia. Incorporation of key stakeholder views and systematic assessment of the issues when implementing risk-based screening aims to ensure a smoother, more unified acceptance of any change [28].

Methods

Study design

Implementation science provides a structured approach to the operationalisation of a new innovation, with the objective of bridging the gap between basic scientific discovery and established clinical practice [10]. Developed theories, models and frameworks provide a guide to implementation and are grouped into three categories (process, determinant, and evaluation) depending on the specific focus of implementation [5]. The Consolidated Framework of Implementation Research (CFIR), developed utilising psychological and human behaviour principles [6, 7], is one of the most commonly used determinant frameworks, and is appropriate for analysis of large systemic change such as the change of a population screening program as it allows for assessment of multi-faceted systems on multiple levels.

A qualitative approach guided by the 2022 Updated Consolidated Framework of Implementation Research (CIFR) [7] was used to complete a robust, structured and in-depth exploration of the pre-implementation issues of changing to risk-based breast screening in Australia. Consulting this validated framework in conjunction with discussion among the co-authors and literature on the topic, an open-ended interview guide (Supplement A) was generated.

The CFIR consists of 67 constructs organised into five domains: Innovation (the “thing” being implemented), Outer Setting (the external features of the program), Inner Setting (internal features), Individual (personal traits, staff roles and responsibilities within the program) and Implementation Process (change strategies required by an organisation) [7]. Within each of these domains there are several constructs, for example within the innovation domain is the construct of relative advantage referring to the benefit the innovation brings compared to existing practice. In the context of risk-based screening this refers to what the specific benefits (or harms) a risk-based screening approach would be over our current one-size-fits-all program.

Study participants

Key stakeholders were defined as those in high level managerial positions of breast screening services, influential researchers, policy makers and state and federal government officials holding positions responsible for screening and leaders in key breast cancer advocacy groups. A purposeful sampling strategy was undertaken by approaching directors of jurisdictional BreastScreen services of all eight Australian States and Territories. Professional networks and contacts were used to approach key researchers, position bearers in relevant health department roles and peak bodies and advocacy groups. A snowballing technique was also used with each interview concluded with the question “Who else would be important to talk to for this topic?” and those recommendations were subsequently invited to participate.

Invitations were sent via email by JL with an attached Patient Information and Consent Form (PICF). Two reminder emails were sent if participants failed to reply but no further contact was made after the third email. Three people holding positions of responsibility in state, territory, and federal health departments redirected emails to individuals they thought more appropriate to interview. Consent was obtained with signed PICFs prior to the start of the interviews.

Data collection

Interviews were conducted by JL (B.Med, FRACS female researcher and surgeon with eight years’ experience conducting qualitative research). Interviews were conducted between February and September 2023. All interviews were audio-recorded and transcribed using the Rev.com voice to text software program. Quality checking was undertaken and transcripts de-identified against the original audio recording by JL. This quality checking process also allowed for data immersion and preparation for data coding.

Data analysis

Three rounds of coding were undertaken. The first round of coding was performed deductively using the predetermined structured codebooks based on the CFIR constructs and domains. These consisted of all five CFIR domains (Innovation, Outer Setting, Inner Setting, Individuals, and Implementation Process) including their constructs.

Transcripts were managed using NVivo® software. Initially, authors JL, LK and LF independently deductively coded two transcripts against the CFIR-informed codebook and then compared the results collectively. Discrepancies in coding were discussed and clarified so consistency of codes was achieved. After these meetings the descriptions for each construct within the codebook was revised for clarity. Pooled-kappa coefficient was calculated aiming for a score > 0.9 indicating a high level of agreement between coders [8]. A pooled-kappa coefficient of > 0.9 was achieved after two coding meetings.

In the subsequent two rounds of coding, content under each CFIR domain was analysed to identify meaning and context pertinent to each construct and domain. This was a combination of thematic and content analysis dependent on the context of the theme for example the issue of cost to the client was broadly analysed for content. This was performed by a single author, JL, with advice and guidance from LF.

Themes were revised and consolidated between overlapping constructs and domains so that 144 first round themes were consolidated to a final 17. Themes which were duplicated across more than one domain were consolidated into one theme and themes raised from only one or two participants were excluded from final analysis as they related to broader themes and were represented elsewhere.

The consolidated criteria for reporting qualitative studies (COREQ) [33] was consulted in the reporting of this study to ensure all criteria were adhered to.

Results

Participants and interviews

Thirty-five potential participants were invited, three declined to participate (maternity leave and inadequate time), five did not respond to the email invitations and six redirected the invitation to a colleague they deemed more appropriate. Participants were from all Australian States and Territories except for Tasmania. Twenty interviews were conducted with 21 participants (19 one on one interviews and one interview with two stakeholders from the same department who requested a joint interview). All interviews were conducted by JL. Three interviews were conducted in person with local participants and the remaining 17 were conducted online. Interviews lasted between 45 and 75 min. Participant demographics are summarised in Table 2. Specific titles have been withheld to maintain anonymity as different jurisdictions use unique terminology meaning an individual could easily be identified should their title be used.

Table 2.

Participant demographics

	N = 21 (%)
Gender
Female	19 (90)
Male	2 (10)
State/territory
Victoria	7 (33)
New South Wales	7 (33)
Queensland	2 (10)
Aust Capital Territory	1 (5)
South Australia	1 (5)
Northern Territory	2 (10)
Western Australia	1 (5)
Stakeholder category
Health Department Manager	6 (29)
Clinical Lead	5 (24)
Program Manager	4 (19)
Lead Researcher	4 (19)
Director of Advocacy Organisation	2 (10)
Length of time in current role
<2 Years	8 (38)
2–5 Years	6 (29)
>5 Years	7 (33)

Findings overview

The main findings of each CFIR domain (Innovation, Outer Setting, Inner Setting, Individuals and Implementation Process) are listed in Fig. 1. Seventeen themes were identified and discussed below. Themes which arose with the implementation process domain were absorbed by other major domains which is not represented in the Fig. 1 or the text.

Fig. 1.

Summary of main findings for each CFIR domain. Implementation process not included in figure as the themes which arose under this domain were consolidated into other domains

Innovation domain

Cost

The concern that risk-based screening would incur higher costs than the current model was widespread with the corollary that this could be prohibitive to successful implementation.

“the argument that then if you screen the people at higher risk differently, that will cost more money and then those, that money would be offset by screening women who are very low risk less. And that's very contentious.”

(Participant 21)

“any future research in this space is going to be, could change the program quite significantly in terms of costs.”

(Participant 3)

The other concern raised related to the potential shifting of cost onto the individual should the population-based program no longer be able to cater to everyone. This issue speaks to the concern that risk-based screening would increase inequities in the program which is explored under the outer setting domain.

“And then is it what we just push people out into the primary care system, or private system, which is unaffordable for some? So we need to think about that.”

(Participant 3)

Improved screening for all

The perceived advantages of risk-based screening, described as ‘improved screening for all’, encompasses a range of benefits that participants perceived about implementing risk-based screening. These included improved sensitivity for early detection, more appropriate screening for people with high mammographic density, more appropriate resource allocation, fewer interval cancers, and better overall survival. Participants appreciated that if more early breast cancers were detected, there were lower interval cancer rates (cancers presenting as a clinical symptom between screening rounds), and less unnecessary interventions, overall cancer outcomes would be improved which is the primary benefit of risk- based screening.

“..improved sensitivity is the gateway to the ultimate benefit, which is reducing mortality, but also in the context of contemporary treatment, early detection. So that not only do women have a better chance of survival, but they have…the...minimal level of treatment that's needed in order to help them survive.”

(Participant 19)

Evidence base concerns

Two themes were identified that described 1) issues about performing risk assessment and 2) a perceived lack of evidence regarding optimal screening for different risk categories.

Risk assessment issues

The issue of risk assessment was a dominant topic, predictably, as this is a pivotal component to stratify a person’s risk of breast cancer. This theme encompasses the who, what, and when issues of who would be best placed to perform the risk assessment, which risk assessment tool to use, and how frequently a risk assessment should be performed. Without a robust, accurate and easy to perform risk assessment, stratification is not feasible or ethically sound.

“The other thing is who does the risk stratification for breast screening?...Who administers it?…Does a woman do it at home online? Does a person do it at BreastScreen? In which case it's extra work for the BreastScreen staff… Do they go to their GP? In which case you'd need an MBS item, and that's a big deal.”

(Participant 13)

MBS refers to the Medical Benefits Schedule which lists services subsidised by the Australian Government. Healthcare providers are remunerated in Australia based on which MBS item numbers are applicable to the care they provide.

“…to…risk stratify, you need to have good risk assessment and that is the difficulty because you need to...have a tool that we can use that's applicable for the population, and you need that tool to be used systematically because otherwise you're either going to inadequately assess a woman or put them into the wrong category, in which case you could be not screening someone frequently enough.”

(Participant 20)

Optimal screening for different risk categories

There was extensive discussion and concern about the lack of evidence for what is the optimal screening for people of varying risk categories. There are existing guidelines available both by EviQ [14] and the Royal Australasian College of General Practitioners [25] however there was no discussion of these by participants. The use of contrast-based imaging (MRI or contrast enhanced mammograms) was frequently raised together with the practical issues of access to these more complex tests within a screening program.

“also... when we talk about breast density... it's still not clear what the steps are. So yes, someone maybe is more at risk of breast cancer if they've got high breast density, but so what?…I think I'm challenged by how much we can do if we don't have clear direction on what the pathways are.”

(Participant 1)

Outer setting domain

National breastscreen review underway

The current federal health department review of national BreastScreen was interpreted as evidence that government was supportive and interested in change.

“I'm really pleased that the Commonwealth and the States, the Commonwealth has agreed to a review of policy and funding of this program.”

(Participant 3)

Influence from other jurisdictions

Each state and territory screening service operates as an independent body and each jurisdiction has unique elements from the management structure to the needs of their clients. The jurisdictional managers meet regularly as a group and there is a sense that change is coming, and the individual jurisdictions do not want to “miss out” which would result in their jurisdictional populations being screening via an outmoded approach. This theme arose within the construct of “Partnerships and Connections” and highlighted the importance of the interconnectedness of the varying jurisdictions.

“it's very much an acceptance that we can either get on the bus or get under it.”

(Participant 4)

“there probably need to be some persuasion across some of the other jurisdictions because some jurisdictions…might be more conservative, they might have more resources…There's always…comfort and safety in knowing that other people are moving in the same direction.”

(Participant 10)

The consumer drive for improvement of breast screening.

This theme was generated under the construct of “External Pressure” and the subconstruct of “societal pressure” which refers to pressure for change arising from outside of the organisation. Participants held the view that there was a push and interest from their clients for risk-based screening, due to the clients’ perception that risk-based screening would improve cancer outcomes and be in their best interests.

“..there's absolute appetite from our perspective and certainly from the consumers we represent.”

(Participant 17)

However, there was also an appreciation that the perspectives of the stakeholders could be biased in that they are hearing the voices of the vocal rather than the masses.

“I do think part of the consumer perspective on it is coming from consumer advocacy groups who are very much focused on the topic and…that doesn't mean that that generalises to all consumers.”

(Participant 19)

Worsening inequalities

The issue of inequity in the current program was raised, with consideration that people with better access to resources (either because of geography, design or money) would have more opportunities to engage in different screening modalities directed by a risk-based approach than those without. This was especially highlighted for those jurisdictions servicing large regional and remote areas of Australia where a larger proportion of people were serviced by a mobile screening service bus. The importance of equity was identified as a key principle of BreastScreen and concern was expressed that a move to risk-based screening would exacerbate inequities in access, particularly in regional and rural areas of Australia.

“obviously the point about equity falls under this…We really need to make sure that we're bringing everybody with us. And the simplicity of the screening, everybody every two years, I think helps.”

(Participant 9)

“There's a whole lot of risks in that, first of all, that it would be expensive. Second of all, it's about equity and access. So it's no good if..only people living in urban centres close to a MRI machine get access to the…recommended sort of screening. What happens to people who don't live anywhere near a centre? What happens to people living in rural and remote areas?”

(Participant 13)

Dichotomy between individual and population risk

This issue of the difference between the risk for an individual and population risk was raised by several participants as a fundamental barrier to population wide risk-based screening given that population screening programs inherently have the overarching view of the entire cohort rather than the narrow lens on an individual.

“there's a whole bunch of things to start thinking about when you really start peeling back the layers about individualising care, but still making sure that you're looking after everybody.”

(Participant 9)

“while it might be a benefit to for some women to participate less screen less often, it also can be a harm because we might actually miss some breast cancers. So importantly we are talking about balancing benefits and harms in population screening, we're not talking about individuals.”

(Participant 13)

This theme highlights a potential fundamental shift required in thinking and philosophy regarding population wide screening programs to implement risk-based screening to the entire population.

Inner setting domain

Motivation to improve

Participants expressed belief that they, and BreastScreen as an organisation, have culturally shared values about improving screening and so were motivated to change to risk-based screening. All participants agreed that BreastScreen should shift to a risk-based screening model due to the perception that the current model will become outdated.

“..we're not getting it right for whatever reason… it's failing to change. It's not that there's a risk to change it, there's a risk to not changing here.”

(Participant 5)

“So BreastScreen as it currently stands, is not sustainable and is not a program that we will have in its same form in 20 years time. So we need to think about how we use other factors other than age to have a program that is effective and cost effective and makes a difference to cancer outcomes.”

(Participant 16)

Current ‘pay per screen’ model of funding is not compatible with risk-based screening

Given that risk-based screening would include the additional element of a risk assessment as well as supplemental screening, potentially contrast imaging, the current funding model would need to change to include the additional needs of a risk-based approach. There was appreciation this funding model would need to adapt to a change in the program.

“We'd have to acknowledge that there's additional costs with the program now. So, you know, there are changes in technology, medicine, research and changing evidence, so the cost to deliver new and improved practice can be very different. The funding model has remained the same for over 30 years and changes have occurred over this time which also vary. We don't have any room to move with the current model. So, that's a problem. So I think the funding model does have to change to support the changes in the program as it evolve somehow.”

(Participant 3)

BreastScreen is not a research organisation

BreastScreen’s primary aim is service provision and there was concern about introducing risk-based screening whilst the concept was still exploratory or considered to be research. There was a lot of discussion around BreastScreen not having systems and processes in place for exploration of new ideas. This also encompasses the theme of BreastScreen not having a culture of change.

“we don't have a lot of capacity to do research and..push the boundaries”

(Participant 2)

“…it (BreastScreen) a juggernaut. It takes a long time to get going and a long time to change direction.”

(Participant 4)

“One of the strengths of BreastScreen.. is that we have people who have very strong values, who really align with the purpose of what they do. The flip side of that is they're very wedded to the way they do that, that job. So.. it is not an organisation that finds change easy.”

(Participant 4)

Competing priorities—participation rates

Participation rates were extensively discussed as a competing priority, as the current screening rate (ranging from 40–60% throughout Australia) is below the ideal (> 75%). There was dialogue that the ‘one-size-fits-all’ approach may be a barrier to some and that non-attendance may continue to be an issue irrespective of the screening model.

“So we know we are missing 40% of the population…there could be argument made that some of those people aren't coming in because it is a one size fits all and it doesn't fit them. And they see it as a not fitting them.”

(Participant 5)

“well we can risk stratify, because we've talked about this as a team and what we'd like to do and… well we still have to get them through the door first to be able to stratify their risk…and our challenge is still getting the women in the door.”

(Participant 1)

There was also conversation about the potential impact of risk-based screening on participation, with some people believing it would increase and others believing it would dissuade people from screening. Another common finding was the belief that the priority should be improving rates of under-screened populations rather than changing the entire BreastScreen process.

“I guess there is a potential benefit in participation…if you are clear…to the person that you are inviting them on a risk-based model, it might be clearer to them that the invite they're receiving is appropriate and then they're more likely to act on it.”

(Participant 14)

“…if they're considered a low risk or every five years, is there more chance that they'll just not screen at all?”

(Participant 18)

Individuals domain

Opportunity for improved overall health

There was appreciation that risk-based screening creates an opportunity for discussion around improved health strategies more broadly such as the impact of regular exercise decreasing and alcohol intake increasing breast cancer risk.

“..is there an opportunity for us to… most of the things that make you at risk other than your breast density, most of the things that make you at risk of breast cancer also make you at risk of a number of other chronic diseases…heart disease, stroke, diabetes.”

(Participant 1)

“I think…it also provides an opportunity for education…if you enabled those conversations to be...had that also puts in a point about education not only in breast cancer, but all cancer risk.”

(Participant 9)

Health literacy

Concerns were raised about communication of risk, communication of change to the program, communication about recommendations for screening as well as broad communication needs for clients. Under this banner were considerable conversations regarding the issues of health literacy and tailoring communication so that every client is as informed as they would like to be.

“…what I think about often is conversations about risk stratification and things like that involve a level of health literacy...They involve a level of complex conversation that can exclude a vast number of our more vulnerable populations.”

(Participant 9)

This theme highlights the complexity of communication requirements around a change to the program, but especially a large change, as this has broad and in-depth needs.

Perception about the motivation for change

There were concerns current clients could see the motivation for change as cost cutting or not advantageous to them as individuals and this could subsequently impact BreastScreen reputation.

“You have to… get around the possible interpretation that...the government's just trying to save money by…screening some people less. And emphasise that screening does have benefits and harms, and by optimizing the protocols we have for who is screened more often and less we can maximise people's chance of benefit, minimize their chance of harm. But I think that is a difficult thing to get across. And people do sometimes wonder about other motives for, for making changes like that.”

(Participant 18)

Discussion

This work demonstrates the broad support key stakeholders have for a transition to risk-based breast cancer screening in Australia. Nevertheless, we highlight several key areas requiring consideration: the cultural shift required in a large complex State or Territory organisation, as well as issues around the evidence base for risk stratification and supplemental screening, and concern about the impact on equity.

The strength of the current BreastScreen program was acknowledged as the “one-size fits all” approach, with the corollary being that it is relatively simple to administer and can be rolled out effectively in regional and rural areas. In theory, women receive the same service, independent of their socio-economic status or postcode. Some stakeholders feared that the shift to an individualised program would mean losing this key, valuable feature of BreastScreen and that this change could potentially be the driver of increased inequity and loss of a crucial strength of the current program. They also acknowledged the strong sense of value within BreastScreen, both to deliverers and recipients, implying that change will be more difficult.

Our findings echo the recent published systematic review of implementation needs for risk-based cancer screening among all tumour streams [9] that highlights ‘the importance of addressing alternate modes of delivery, education of health professionals, communication needs of consumers, screening options for those at low risk and inequities in cancer outcomes’. These findings also reiterate the recommendations from the Roadmap to Optimising Screening in Australia (ROSA) project (The Daffodil Centre 2023) that reinforces strong support for risk-based screening among key stakeholders in Australia. Acknowledging, however, the importance of evidence and consensus-based recommendations before a change in practice can occur. Our participants frequently described the lack of guidelines as being an evidence gap to implementation however the existing guidelines outlined by both EVIQ [14] and RACGP [25] were not mentioned indicating there may be a lack of trust in these or a need to increase awareness. The DENSE trial [2] demonstrated improved cancer screening (reduced interval cancer rates) for people with high mammographic density using MRI screening every 18 months, however due to funding constraints, this has not been able to be implemented into routine practice in The Netherlands. This highlights the need for implementation approaches to help bring practice changing evidence into a change in clinical care.

The emphasis on needing greater evidence to support the implementation of risk-based screening was also highlighted in a systematic review addressing the perspectives of healthcare professionals in regards to acceptability of risk stratification within population-based cancer screening programs [30]. The issue around resource allocation arose in this review as well as the requirement for clear communication. Most of the participants in these studies were health care professionals although one included the perspectives of policy advisors and decision makers with the United Kingdom National Breast Screening Program and this was concentrating on their views for screening those at low-risk [22] and none were from Australia. Given the unique environment of each country, its culture, resources and workflow of their screening services this study demonstrates unique insights into the Australian context.

Of the ten principles for population screening programs outlined by WHO [36] there are two which require specific consideration in planning a change to risk-based breast cancer screening: 1) Faculties for diagnosis and treatment should be available and 2) The cost of case finding should be economically balanced. In terms of availability, we have demonstrated concern from key stakeholders regarding the access to breast cancer screening beyond mammography alone. There was considerable concern that people living outside of major metropolitan areas could be disadvantaged regarding access to additional screening tests such as contrast based imaging. Acknowledging this fundamental issue needs to be addressed as we consider moving ahead so as not to worsen the current health inequities in our community.

The cost of risk-based screening will require careful modelling and planning. A systematic review reports this approach is cost effective [18] with ‘high-risk women screened annually yielded a higher mortality rate reduction and more quality-adjusted life years at the expense of higher cost and false positives’. However, this will vary depending on screening modalities, time frames and variables to the model, and needs to be specifically addressed in the Australian context especially considering none of the studies included in this systematic review were based in Australia [18]. The issue of cost effective modelling was addressed in the ROSA report (The Daffodil Centre 2023) concluding that most risk-based scenarios would be beneficial however the models which led to the most substantial cancer mortality reduction also led to higher overdiagnosis. The cost savings from reduced treatment intensity, which would occur with finding more early-stage cancers, also needs to be incorporated into models. Although, this is a system-wide benefit rather than a direct saving to the screening program. This potential benefit of finding more early-stage cancers was not raised by our key stakeholders and may represent the bias of interviewing those responsible for managing the program rather than those treating cancer patients.

This work has specific strengths and limitations. The strengths include the use of the CFIR that enabled a structured approach to understanding the factors that might inform the implementation of risk-based breast screening. The qualitative approach ensured an in depth understanding of participants’ perspectives were captured. However, this method does not aim for representativeness in the participant sample. It is possible that this introduced selection bias whereby only those with strongly held views of risk-stratified screening agreed to take part. The other limitation is the over representation of participants from highly populated eastern states. It is possible there would have been unique issues from Tasmania which was not successfully recruited.

The potential to worsen health inequities with risk-based screening is a major issue to be addressed as risk-based screening is implemented [15, 16, 19, 22, 35]. Ensuring that screening clients have equitable access to risk assessment, appropriate modality and frequency of screening, and treatment is of critical importance to any population screening programme and needs to be a key focus for those planning and conducting any change to risk-based breast screening.

Conclusion

Key stakeholders who direct, manage, lead the delivery of, and research the Australian breast screening program were motivated to change the current population-based model to a risk-stratified approach. However, major issues with this change were explored, including concerns about worsening inequities of access, financial cost and the current inadequate participation rates being a competing priority. These findings identify the priority areas that need additional research, for example, health economic analysis of various approaches to risk-based screening, or piloting, for example contrast imaging in remote locations. Developing and testing these types of strategies are necessary to changing Australia’s population-based breast screening to a risk-based program.

Author's contributions

JL wrote the main manuscript text. JL, LF, LK, SB responsible for conceptualisation of the project. JL, LK, RP, LF contributed to coding and data analysis and all authors reviewed and edited the manuscript.

Funding

JL was funded through the Royal Australasian College of Surgeons (RACS) Herbert and Gloria Kees Scholarship for this work. A/Prof Laura Forrest is funded by a mid-career fellowship from the Victorian Cancer Agency (MCRF 20012).

Royal Australian College of Surgeons (RACS),Herbert and Gloria Kees Scholarship,Victorian Cancer Agency,MCRF 20012

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethics approval for this project was obtained through The University of Melbourne Human Ethics and Research Committee. Reference Number 2022–24940-354493–3. Consent was obtained from each participant both verbally with a signed Participant Information Consent Form at the start of the interview.

Consent for publication

As part of the consent process, participants were asked if they would like to review their quotes prior to publication as well as giving verbal consent for publication.

Competing interests

The authors declare no competing interests.