Organizational challenges to equity in the delivery of services within a new personalized risk-based approach to breast cancer screening

Abstract

Emerging evidence opens new possibilities to improve current breast cancer mammography screening programs. One promising avenue is to tailor mammography screening according to individual risk. However, some factors could challenge the implementation of such approach, specifically its potential impact on the equitable delivery of services. This study aims to identify the barriers and facilitators to the equitable delivery of services within a future integration of a personalized approach in the Québec screening program. We then propose different means to address them. We conducted 16 semi-structured interviews with stakeholders with a role in the management, implementation or assessment of the Québec screening program. The barriers and facilitators identified by respondents were regrouped in two themes: 1) Reproduction of social inequities, and 2) Amplification of regional disparities in access to services. We consider that fostering inclusion through communication strategies and relying on electronic communication technologies could help in addressing these issues.

Keywords:

• breast cancer
• equity
• screening
• personalized medicine

1. Introduction

Currently, Canadian breast cancer screening programs provide biennial mammography to asymptomatic women between the ages of 50 and 69, regardless of individual risk factors (Canadian Task Force on Preventive Health Care 2011). Similar programs to detect cancer at an early stage exist around the world. However, there remains an ongoing debate about whether the benefits of these programs outweigh their possible disadvantages (such as overdiagnosis, false-positive results, unnecessary follow-up biopsies and anxiety) (Myers et al. 2015; Houssami 2017). Emerging evidence in personalized medicine opens new possibilities to improve these screening programs. One possibility that might be soon implemented is the assessment of personalized risk in order to target women most likely to benefit from mammography screening (Shieh et al. 2017; UK Government, Department of Health 2017). In this way, women who are at higher risk of developing breast cancer could be advised to start mammography screening at a younger age and/or have more frequent mammograms. This could improve patient outcomes, detection of earlier-stage cancer, and allocation of health care resources (Hall and Easton 2013; Onega et al. 2014).

Recent discoveries have demonstrated that it is now possible to identify hundreds of genomic markers that slightly or moderately increase the risk of developing breast cancer (Mavaddat et al. 2015; Chatterjee, Shi, and Garcia-Closas 2016; Michailidou et al. 2017; Milne et al. 2017). These genomic factors may be combined with non-genomic risk factors (such as age, mammographic density, family history of cancer) in a computational algorithm, providing a useful assessment of individual risk (Figure 1). Calculated individual risk would then allow recommended screening measures to be tailored to the appropriate level of risk (Gagnon et al. 2016; Rainey et al. 2018). This personalized risk-based approach would improve current mammography screening programs that primarily use age as the inclusion criterion (Hall and Easton 2013). Although this emerging approach has not yet been implemented by health authorities anywhere in the world, many ongoing research projects involving thousands of women could well favor its future adoption (e.g. BRIDGES [Breast Cancer Risk after Diagnostic Gene Sequencing]; FORECEE [Female Cancer Prediction Using Cervical Omics to Individualize Screening and Prevention]; MyPeBS European Commission 2017; PERSPECTIVE I&I [Personalized risk assessment for prevention and early detection of breast cancer: Integration and Implementation]; WISDOM).

Figure 1. Breast cancer individual risk assessment process: An example. The percentages used here are approximate; they were used to facilitate discussions between experts during the development of the PERSPECTIVE recommendations as explained in Gagnon, J.; Lévesque, E.; the Clinical Advisory Committee on Breast Cancer Screening and Prevention; et al. Recommendations on breast cancer screening and prevention in the context of implementing risk stratification: impending changes to current policies. Current Oncology 2016, 23, e615-e625, doi:10.3747/co.23.2961.

Because risk-based screening using genomic factors is a radical new approach with regard to screening programs, various organizational issues need to be considered prior to implementation (Dent et al. 2013). Equity in the delivery of services is one of the organizational issues that could challenge the implementation of this approach in current screening programs (Anderson and Hoskins 2012; Hall et al. 2014; McGowan et al. 2014; UK Government, Department of Health 2017). Equity is a fundamental value of the Canadian health system, and means that “citizens get the care they need, without consideration of their social status or other personal characteristics such as age, gender, ethnicity or place of residence” (Romanow 2002). In the literature, equity in the delivery of services is also articulated in terms of the concept of “distributive justice,” referring to situations in which individuals or groups unfairly benefit from health services due to their socio-economic status, educational background or ethnicity (Hall et al. 2014).

Although some potential equity issues have already been identified in the literature (Anderson and Hoskins 2012; McClellan et al. 2013; Hall et al. 2014; McGowan et al. 2014; UK Government, Department of Health 2017), they may not be automatically transposable to the approach under study. The specificities that might prevent transposition of prior results include: the type of genomic markers used (e.g. when mutations on the high risk genes like BRCA1/2 are not targeted), the Canadian insurance context (e.g. where health services are public and tax-funded), and the public health-oriented approach (when services are population-based). Nevertheless, it is essential to recognize the factors likely to challenge the equitable delivery of services with the integration of this personalized risk-based approach in the Canadian context (Hall et al. 2014), and then identify the solutions that could be envisioned to manage them. In addition, it is recognized that exploring health professionals’ perceptions related to personalized screening might facilitate implementation (Rainey et al. 2018). To our knowledge, our study is the first to examine equity in the delivery of services for the risk-based approach through stakeholder perceptions.

2. Context and methods

As part of the Personalized Risk Stratification for Prevention and Early Detection of Breast Cancer (PERSPECTIVE) research project (Genome Québec), which is developing tools to implement a risk-based approach (e.g. calculation algorithm, genomic test, economic simulation model, screening policies), we had to consider the various parameters of tools developed through this project. This presupposes an individual risk assessment step to complement Canadian breast cancer screening programs. It is conducted with an algorithm (next version of BOADICEA (Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm)) which computes both genomic test results (only variants increasing lowly or moderately the risk) and extended personal risk factors (e.g. mammography density, alcohol intake, body mass index).

Each Canadian province has its own public screening program, but provincial programs share common bases and a similar health services delivery context. Our study focuses on one of them: the program in the province of Québec. Since 1998, this screening program has provided one mammogram every 2 years to all women between 50 and 69 years of age. It is managed by the Ministry of Health, implemented by hospitals and clinics and assessed by various governmental bodies.

Our global aim was to study the barriers and facilitators of the implementation of the personalized risk-based approach, by focusing on: 1) the influence of organizational factors, 2) the use of information by insurers, and 3) the equitable delivery of services. As analysis of parts of our interviews pertaining to the first and second aspects were presented in previous publications (Hagan, Lévesque, and Knoppers 2016; Dalpé et al. 2017), this paper focused on the results of the third remaining aspect.

Since our aim focuses on organizational aspects of implementation, we designed our data collection and analysis within an institutional framework, which offers a relevant sociological perspective to study processes of organizational change in healthcare (Pedersen and Dobbin 2006; Currie et al. 2012; Zilber 2012). We planned semi-structured interviews (open-ended questions) with stakeholders having a good level of knowledge of the delivery of services in the Québec breast cancer screening program and assuming some responsibility in this program at the organizational level (i.e. management, implementation or assessment).

The general themes selected for the interview guide (see Table 1) reflect the main dimensions of the institutional framework, while the sub-themes were identified by analyzing the documents published on the current Québec breast cancer screening program (framework, reports, action plan and evaluations) and via analysis of the international literature on barriers and facilitators to the implementation of genetic technology and personalized medicine (using PubMed). Approval was obtained from the ethics committees of McGill University and the Québec University Health Center (CHU de Québec-Université Laval).

Table 1. Development of our interview guide on the barriers and facilitators of the implementation of the risk stratification approach (focusing on the influence of the organizational factors, the use of information by insurers, and the equity in the delivery of services).

Theme	Sub-themes of the questions	Example of question
Resources (human, material, etc.)	Specific challenges for remote regions Follow-up/management of genetic information Sufficiently of the resources Potential form of injustice in targeting women at higher risk	To what extent could the specific challenges of remote regions influence the implementation of a risk stratification approach?
Cognitive dimension	Needs of health professionals to understand and explain results Impacts of current relationships between health professionals Training needs Translation of research discoveries in the clinic Perception of genomics among health professionals	To what extent could the complexity of the test (in terms of interpretation, for example) affect the implementation of a risk stratification approach?
Formal norms	Influence of current management schemes and methods Management methods Impact of current guidelines, policies and programs	In which manner could the current guidelines, policies and programs encourage the implementation of a risk stratification approach?
Relationships between policy scales, sectors and professions	Relations between different policy levels (local, regional, provincial) Adaptation to regional realities	To what extent should the implementation of a risk stratification approach be adapted to the regional realities?
Practices	Changes in practices (habits, attitudes, etc.) of the work environment Access to medical or genetic information Use by insurers	Can you identify, in your work environment, the practices that could challenge the implementation of a risk stratification approach?

In total, we completed 16 individual interviews (12 in person and 4 by phone), lasting each on average 50 min, between March and June of 2014. Interviewees were identified through our network of collaborators as well as via public documents available on the Internet, in addition to snowball sampling. Potential participants were approached directly or via email. Almost all the interviewees had a clinical background (14/16) and they were selected to ensure variation in the following criteria: region, gender, type of involvement in the program, affiliation and prior knowledge of the risk-based approach (Mason 2010) (see Table 2). Recruitment continued until data saturation (i.e. until all viewpoints were represented without repetition) (Lai-Goldman and Faruki 2013). Interviews were conducted in French (excerpts presented in this paper were translated). However, 3 interviewees did not answer questions related to the equitable delivery of services because this topic was outside the scope of their expertise or due to lack of time. These 3 interviews are not included in the analysis presented in this paper.

Table 2. Profile of the 13 stakeholders who answered questions on equity in the delivery of services. Stakeholders who had previously taken part in an expert consensus meeting on the personalized risk-based approach (to achieve another aim of our PERSPECTIVE project) were considered to have prior knowledge of the approach.

Background	Doctor (9), Nurse (3), Lawyer (1)
Principal involvement in the program	Management (5), Implementation (5), Assessment (3)
Principal affiliation	Public health authority (4), Breast screening program (3), Health practices assessment body (3), Ministry of health (3)
Gender	Female (5), Male (8)
Region	Metropolitan (3), Urban (6), Rural (4)
Prior knowledge of the approach	Yes (7), No (6)

At the beginning of each interview, participants were provided with standard information on the risk-based approach (also called “risk stratification,” see Table 3) in order to ensure a common basic understanding of this new approach, and avoid confusion with the traditional approach used with women having a family history of cancer (i.e. through identification of high-risk mutations in the BRCA1/2 genes after genetic counseling). Stakeholders were informed that, with the risk-based approach, each woman would be provided with a statistical estimate of the risk of developing the disease until the end of life expectancy (80 years), thereby situating each woman within a risk level.

Table 3. Factual background provided to the stakeholders at the beginning of each interview.

What is a risk stratification approach for breast cancer?

When one does a risk stratification, one establishes who are the individuals with the higher risk of developing a health issue in order to offer personalized care, thus preventing or detecting the health issue sooner. This project aims at offering risk stratification for breast cancer to women. A computer program combines the risk and protection factors of a woman in order to establish her risk level. This includes personal factors (e.g. age, age at first birth, breast density), family history as well as genetic factors. The genetic factors tested consist of numerous variants having an impact on breast cancer risk. These are not genes significantly increasing the risk for individuals (e.g. BRCA1/2). These are genomic variants that increase or decrease the risk less markedly. However, when put together, all the genomic variants tested can change the risk level significantly.

The interviews were recorded (except for one who preferred handwritten notes) with the participants’ written consent and then transcribed. Interview transcriptions and handwritten notes were coded thematically (by J.H.) to address the issues identified in the literature (Currie et al. 2012) and new codes were assigned to the interview sections addressing empirical issues that we were not able to address using the initial literature-driven themes (Srivastava and Hopwood 2009). Transcriptions and thematic categories were discussed with a second researcher (E.L.) throughout the analysis process. Data relevant to equity were placed in the two broader thematic categories that emerged from our analysis.

3. Findings

Almost all answers provided by the interviewees were in relation or in comparison to the Québec screening program currently in place. As such, to support understanding, the presentation of our findings will be preceded in each sub-section below by background information on the Québec healthcare context. Due to the empirical experience of the interviewees, the facilitators identified were mostly formulated as solutions or proposals to address an existing or anticipated issue. The barriers and facilitators identified by respondents were grouped in two themes: (3.1) Reproduction of social inequities, and (3.2) Amplification of regional disparities in access to services.

3.1. Reproduction of social inequities

The stakeholders interviewed raised concerns about reproduction of social inequities across 3 stages: a) Solicitation of women b) Informed consent, and c) Return of results.

3.1.1. Solicitation of women

Under the Québec screening program, an invitation letter and a one-page information pamphlet (Québec Government, Ministry of Health and Social Services) is sent every two years to the home address of each woman between the ages of 50 and 69. The pamphlet encourages women to obtain a mammogram in one of the designated screening centers. Having an assigned primary care provider does not affect eligibility, though some women could be encouraged and/or reminded to participate by their assigned primary care provider. The methods that would be used to invite women to have a risk assessment in a personalized risk-based approach have yet to be determined. For instance, who will invite the women, when will women be invited and how they will be contacted, are still unknown.

Almost all respondents (12/13) raised concerns about the impact that the chosen method of invitation could have on equity of participation. They underlined current challenges to the inclusion of certain disadvantaged sub-groups of women in regard to language, ethnicity, literacy level and/or education. Some interviewees mentioned that this difficulty could be exacerbated by the complexity of the risk-based approach (e.g. several steps before mammography, probabilistic risk estimation process) and the contribution of a genomic component.

According to some respondents, in-person solicitation by the assigned primary care provider could be an adequate alternative to the traditional postal invitation to address these potential inequities. However, most of the interviewees (8/13) pointed out the fact that many women in Québec do not have an assigned primary care provider who oversees their preventive care. In this context, some stakeholders considered that relying only on the assigned primary care providers to solicit women could also replicate/reinforce difficulties in access to health services that these women currently experience. One respondent underscored the fact that, no matter how efficient the solicitation tools are, the recommendations of primary care providers will likely remain decisive:

Because health professionals have credibility among the population, more than any tools we can build. We can use the best tools to convince someone to participate, but when the doctor says “That’s not worth doing,” we lost everything we just obtained.

To address the issue of equity during the solicitation of women, some interviewees suggested using competing solicitation approaches (e.g. postal letter associated with solicitation from primary care providers and community health leaders), adapting communication tools to the particular concerns of sub-groups (e.g. cultural groups less comfortable with image of breasts) and involving women in the development of these tools (e.g. focus groups including women of different ethnicities). One interviewee pointed out a successful strategy developed to inform women in Indigenous communities:

We made a kind of collaboration with them to find the best strategy […] In the end, it was very simple. This was the portrait of the female chief, on a poster, who encouraged women to learn about the program.

3.1.2. Informed consent

Currently, written consent (Québec Government, Ministry of Health and Social Services) to participate in the Québec screening program is obtained without doctor involvement at the mammography clinic (Québec Government, Ministry of Health and Social Services). The informed consent process that would be adopted in a risk-based approach might take several forms, specifically with regard to the involvement of healthcare professionals and to the informational tools that would be used (e.g. website, poster, letter, pamphlet, video). It is still undetermined which professionals will be involved, as well as when and how they will obtain informed consent.

Several interviewees mentioned that the risk-based approach could add complexity to the informed consent process. But, only few of them expressed the opinion that this complexity could foster social inequality and challenge the informed nature of the consent. A stakeholder pointed out that this additional complexity could:

[…] significantly increase health inequalities by favoring university graduates (…) and by clearly disadvantaging those who have not finished high school.

For about half of the stakeholders (7/13), specific concerns emerge when health professionals need to discuss the concepts of genomics and risk with patients. Some of them underscored the barriers for participating women (e.g. “women are very uncomfortable with risk”) or from the health professionals (e.g. “physicians have difficulty understanding risk”), while others shed light on health professionals’ role in supporting a more complex informed consent process (e.g. “if your patients don’t understand, it’s because you didn’t explain it properly”). One stakeholder explained the burden on health professionals of an extended duty to inform and to advise:

Even when you try to obtain the most informed consent possible from women, it’s the doctor’s responsibility to explain things to the patient. Not to decide for them, but to recommend the right course of action […].

In addition, some interviewees mentioned the importance of considering the cultural specificities of different sub-groups (e.g. those not understanding the language of the majority; having a cultural taboo about breasts; showing reluctance to undergo screening without involvement of their husband; lacking trust in health authorities) and hence the need to adapt communication tools. For instance, one interviewee raised concerns about immigrant women:

With an immigrant population, I do not know how the message can really reach them when French is their third or fourth language. We must be very careful not to increase social inequalities in health.

Yet, many of them also emphasized the fact that the current Québec screening program information document was already adapted to local particularities in its many versions in order to decrease the impact of cultural particularities on participation in the program.

3.1.3. Return of results

Within the current Québec screening program, mammography results requiring follow-up are discussed between the women and her assigned primary care provider (or a primary care provider associated with the program if she does not have her own). They both receive the mammography results. Unlike the current program, a personalized risk-based approach would generate three types of results before starting the mammography screening: the estimated risk, the risk level and the risk-based screening recommendations. Communication of these results will need to be organized in order to support decision-making about the risk-adapted screening recommendations. Some of the unknowns are who will return the results, when and how this will be done, and also to whom the results will be communicated.

Many stakeholders (8/13) expressed concerns about the possibility of making results understandable for women due to their inherent complexity or an insufficient level of health literacy. One stakeholder explained why there should be additional resources to facilitate efficient communication:

There is a high proportion of the population who don’t have a sufficient level of literacy to understand these health concepts. Numeracy level, understanding numbers, is even worse because you talk about risk.

Some respondents (5/13) also underlined that understanding could be compromised by the lack of knowledge and/or expertise of the primary care provider himself/herself.

To mitigate inequity at the step of the return of results, stakeholders proposed to: use adequate communication tools (e.g. simple, efficient, with a visual design, easy to use by healthcare professionals who have little time to communicate information); rely on the experience of risk communication acquired from other health conditions in preventive medicine (e.g. cardiovascular diseases, colorectal cancer); and offer access to women and/or health professionals to a specialized resource (e.g. genetic counselors).

3.2. Amplification of regional disparities in access to services

In Québec, some remote areas are poorly equipped with health facilities and have scarce access to primary care providers or specialized doctors. The current screening program, with its decentralized implementation strategy, aims to provide equivalent health service delivery irrespective of geographical location. For example, mobile mammography services (i.e. portable mammography machines that are transported by plane, boat or train) currently provide screening to some remote communities that are hard to reach by land, for instance to Indigenous communities in Northern Québec (Cree Board Council of Health and Social Services of the James Bay). In addition, women who do not have an assigned primary care provider to follow-up with the mammography results are provided with an ad hoc primary care provider. How health services would be geographically available in a risk-based approach remains to be determined. For instance, it has yet to be decided which services will be available, and where and how women will have access.

Regional disparity in access to services in a personalized risk-based approach was identified as a key issue by most of the interviewees (9/13). They were concerned that “some regions would be advantaged compared to others.” Their concerns included the services required to calculate risk (e.g. sampling, genomic testing), but also the follow-up measures recommended for each risk level (e.g. imaging exams, consultation with a specialized doctor, genetic counseling). Many expressed the view that women living in rural or remote areas should have access to the same services offered to women in urban centers. For instance, equitable access to specialized imaging technology was mentioned:

What speaks to me is really to ensure that equitable access is offered. If a program is implemented, is offered to everyone, some regions don’t have magnetic resonance imaging.

The geographical accessibility issue was illustrated by a stakeholder, who mentioned how it could be challenging to convince women from remote areas to go to urban medical centers hundreds of kilometers from their place of residence for genomic testing. Many interviewees were specifically concerned with the accessibility of genetic services (genetic laboratories and genetic counseling), which are concentrated in urban centers located far from many rural areas.

In addition, access to services raised concerns among interviewees in respect of the services that would be required in a personalized risk-based approach for women experiencing anxiety, stress or other concerns. Almost all respondents (10/13) underlined the management of anxiety as an expected issue in any personalized risk-based approach. The potential lack of financial resources and of adequately trained health professionals and genetic counselors were the main challenges mentioned by the respondents in regard to psycho-social service delivery.

Many stakeholders (6/13) identified the measures and inventiveness of the current Québec’s screening program as a way to mitigate inter-regional inequities. One interviewee suggested setting up a specialized national telephone resource as a solution to the scarcity or uneven geographic distribution of genetic counselors and geneticists.

4. Discussion

Our findings revealed that interviewees involved in the management, implementation or assessment of the Québec breast cancer screening program expect that equitable delivery of services would be challenged by the addition of a risk-based component. Their concerns were chiefly concentrated on two issues: the reproduction of social inequities, and the amplification of regional disparities in access to services.

These findings are consistent with the concerns expressed by many authors in the literature on the implementation of personalized medicine in breast cancer and of genomic testing. In 2013, Burton et al. clearly summarized the issue of equitable access in the context of the future implementation of a risk-based screening approach:

With respect to distributive justice, it will be important to ensure that different societal groups have equal access to the program and that it is delivered in a fair and objective fashion so that, as far as possible, different societal groups have an equal chance of benefit (Burton et al. 2013).

Chowdhury et al. also raised concerns that the “introduction of personalized screening programs could also be seen as undermining the principles of solidarity and fairness on which current screening programs that offer universal coverage are based” (Chowdhury et al. 2013). These authors highlighted the fact that complexities inherent to personalized risk assessment could disadvantage people from certain ethnic and socioeconomic groups due to a lower uptake (Chowdhury et al. 2013). Anderson & Hoskins maintained that more research “on the use of risk assessment models in underserved minority populations is critical to informing national public health efforts to eliminate breast cancer disparities” (Anderson and Hoskins 2012). They supported their claim by the knowledge gap of how the relationships between breast cancer and genetics can affect underserved minority women.

More specifically, the literature has reported that: risk perception is affected by health literacy, health numeracy, and cultural beliefs (Anderson and Hoskins 2012); limited health literacy poses a particular challenge to communicating with underserved communities about genomics (Lea et al. 2011); people with lower health literacy may not be able to take advantage of precision medicine in the same ways as those with higher health literacy (Ferryman and Mikaela 2018); risk is a difficult concept that is not well understood by the public (Lea et al. 2011); and even a well-educated population might not fully understand the terminology associated with genetic testing (Hooker et al. 2014).

On the issue of regional disparities in access to services, the literature also reveals concerns in regard to the provision of genetics services in rural areas. Guttmacher et al. expressed this concern 17 years ago—long before the implementation of genomics in population-based programs was envisioned—in regard to the widespread integration of genomics in medicine generally:

Undoubtedly, many factors contribute to lack of equity in access to genetic services. Geographic barriers to access have been one such factor. With only a few thousand genetic specialists nationally, many cities have no genetic specialists and some states have only a few (Guttmacher, Jenkins, and Uhlmann 2001).

Jonhson et al. also reported that individuals living in rural areas face significant barriers that prevent access to genetics services (Lea et al. 2015). The negative impact of regional disparity in the delivery of services is an issue that was already observed in breast cancer screening programs. For instance in the United Kingdom, where travel is often needed to get to a mammography facility, the availability of a car for the household was considered a “significant predictor” of accessing mammograms (but not a predictor for cervical screening that is usually provided locally) (Moser, Patnick, and Beral 2009).

In order to manage the issue of equity for the future implementation of a risk-based approach, we looked into solutions proposed by the interviewees and in the literature. We identified, through discussions among authors, who are closely involved in the PERSPECTIVE project (J.S. is lead investigator and B.M.K. is co-lead investigator), what might be the most promising solutions. The two solutions relevant for the Québec context were considered to be: 4.1) fostering inclusion through communication strategies and 4.2) relying on electronic communication technologies.

4.1. Fostering inclusion through communication strategies

Interviewees underlined the complexity of communicating genomic information, pointing out that difficulties in making key concepts of this approach understandable could negatively impact the participation of women with low health literacy or who come from less privileged socio-economic groups. In the province of Québec, significant socio-demographic variances occur from one region to another, and so do literacy and numeracy rates. In general, Québec rural areas present a lower level of health literacy than metropolitan and/or urban areas. (Institut de la statistique du Québec) Cultural barriers may also affect communication about health issues, for instance with immigrants (Paternotte et al. 2015) or Indigenous communities (National Collaborating Centre for Aboriginal Health). Having a public healthcare system irrespective of the ability to pay doesn’t guarantee equity in its access to marginalized populations, such as homeless people (Hwang et al. 2010) or people from Indigenous communities (Tang and Browne 2008). In addition, interviewees expressed concerns about the impact of solicitation that would rely essentially on assigned primary care providers (in 2014, 28% of women in Québec lacked a primary care provider) (Institut de la statistique du Québec).

Due to its complexity, the personalized risk-based approach will likely require “more interactions between the service and the recipient” (Dent et al. 2013) than current age-based programs. While more interactions will occur, the efficiency of communication with women could have greater impact on participation. The adaptation of solicitation tools that address the particular needs of sub-groups, as well as the participation of women in developing these tools, have been offered as a solution by some respondents. They also mentioned that the current Québec screening program already uses a similar process in some regions. As reported in the literature, information “should be culturally sensitive, and culturally appropriate support should be provided for different groups of service users to ensure maximum coverage and inclusivity” (Hall et al. 2014). To address lower uptake among some ethnic or socioeconomic sub-groups, “active communication strategies […] will be needed to mitigate any exacerbation of existing inequalities” (Chowdhury et al. 2013).

We consider that fostering the inclusion of disadvantaged groups through communication strategies might support equitable delivery of services. This could consist of developing appropriate communication measures to support participation of those affected by inequities, including those with a minority language or culture, limited knowledge of the health system, low level of health literacy and numeracy, and reduced mobility. These measures could also ensure equitable solicitation of women that don’t have an assigned primary care provider, for instance by using a variety of methods of invitation. Finally, as the impact of ineffective communication can occur at different stages (solicitation, consent, and return of results), equitable communication strategies would need to cover all these steps.

However, some organizational barriers may impede the implementation of inclusion strategies. First, an increase in costs could be reasonably expected if methods to communicate were multiplied (e.g. in terms of language, format, place, and staff involved) or if they would involve more personalized interactions with healthcare workers. In addition, if adapted communication strategies relied on non-conventional healthcare channels of communication to reach the disadvantaged groups (e.g. community organizations, churches (Karcher et al. 2014)), it could be challenging to build collaboration and align procedures with organizations outside of the healthcare system (Maxwell et al. 2014).

4.2. Relying on electronic communication technologies

During the interviews, concerns were expressed about whether regional equitable access to health services (i.e. specialized expertise and psychosocial support) would be provided. In Québec, regional access might be challenged by the size of the territory and its highly variable population density, the unequal geographical distribution of healthcare services and resources, and difficulties in accessing certain types of healthcare professionals in some regions (e.g. primary care providers, genetic counselors). For example, many Indigenous communities don’t have year-round road access or a local hospital, and patients must leave their community to access more specialized care (National Collaborating Centre for Aboriginal Health). In addition, there are a small number of genetic counselors and geneticists, and their practice is concentrated in urban and/or metropolitan areas (Québec association of genetic counselors).

The literature suggests that providing genetic consultations via videoconferencing offers a solution to provide services in a context were specialized resources are scarce and unevenly distributed (Elliott et al. 2012; Buchanan et al. 2015; Bradbury et al. 2016; Mette et al. 2016). Research results demonstrate that communication technologies contribute to avoiding unnecessary transportation of people from First Nations communities in remote areas, allowing them to stay with their family (Khan et al. 2017). This alternative method of genetic counseling delivery by a videoconference system that incorporates non-verbal cues is also called “telegenetics.” According to the literature, telegenetics increases patients’ knowledge and decreases depression and anxiety (Hilgart et al. 2012). Some results suggest that telegenetics is acceptable and convenient in the context of genetic counseling for hereditary breast and/or ovarian cancer (Zilliacus et al. 2010; Zilliacus et al. 2011; Pruthi et al. 2013). In contexts where genetic counseling resources are lacking, the potential of telegenetics should be considered. This technology could also be used to provide specialized expertise to health professionals themselves in order to address the lack of expertise of primary care providers and difficulties accessing this expertise.

Information technology can also be used to provide both clinical and psychosocial support to women. One interviewee proposed setting up a national telephone resource offering specialized expertise regardless of location. Specifically, electronic technologies enabling communication over distance between women and persons offering clinical or psychosocial support should be considered, by phone (Davis et al. 2015; Casey et al. 2017), text messaging (SMS) (Hall, Cole-Lewis, and Bernhardt 2015; Rico et al. 2017; Uy et al. 2017) and videoconference (Mette et al. 2016). Studies have demonstrated that communication technologies could be used to provide remote support, while maintaining the quality of the service (Melton et al. 2017).

We consider that the use of electronic communication technologies could offer a solution for regional inequitable access to clinical and psychosocial support. Not only could this solution be used to support women and health professionals at the stage of the return of results, but should also be considered at two earlier stages in the process: solicitation of women and informed consent. Several interviewees expressed apprehension in regard to the complexity of the information used to communicate with women at these stages. Electronic information technologies (e.g. video, chat, text messaging) might help address this concern at these stages, regardless of the region of residence of the women.

Nonetheless, implementation of electronic communication technologies in the healthcare system might encounter some organizational barriers. One of the most notable in the province of Québec is the strict, complex framework currently in place to ensure the confidentiality of communication between patients and healthcare providers. This framework includes laws, by-laws, professional codes of conduct, governmental policies, agreements on confidential data exchange, and institutional computer security rules. In addition, barriers related to the technology itself should not be overlooked. Mandatory public tenders in the health system could force the use of some software and/or interface brands that are not preferred by the users (Regulation respecting contracting by public bodies in the field of information technologies). Many remote regions, specifically where Indigenous communities are located, could also face limitations to access video communication technologies due to limited Internet capacity (The Canadian Radio-Television and Telecommunications Commission).

5. Limitations and conclusion

There are a few limitations that should be considered when interpreting our findings. First, the size of the Québec screening program made our recruitment challenging. Although our sample appears small, it is adequate for a qualitative study considering the size of the health system (the Québec population is 2,5% of the U.S. population, and 12% of the U.-K. population). Second, half of the stakeholders interviewed had previously taken part in an expert consensus on the risk-based approach to achieve another aim of the PERSPECTIVE project. These interviewees could have been influenced by some general information gathered during the expert consensus, although only medical aspects were discussed (Gagnon et al. 2016). On the other hand, the interviewees who participated in the expert consensus could also be better informed in identifying the issues relevant to implementation. Finally, while many respondents relied on their clinical experience to anticipate the potential needs and behaviors of women, our study did not include direct assessment of patient perspectives. Our focus was on the organizational barriers and facilitators, while women’s views are under study by other PERSPECTIVE researchers.

We should mention one emerging aspect of equity in the delivery of service that was not covered in our study: the reduction of mammography screening for women classified in the lowest risk level. Although some authors had raised this sensitive issue in terms of distributive justice (Hall et al. 2014) or social acceptability (Meisel et al. 2015), we did not include this topic in our interview guide because reducing the services currently offered was not planned in our PERSPECTIVE project. As the evidence available in 2013 did not support recommending decreasing mammography frequency for the lowest risk level, new evidence may well suggest exploring this issue in the future.

Many benefits are expected from risk-based screening, including higher detection rates at an earlier stage and a better allocation of financial resources in the health care system. Indeed, this new approach brings a “novel screening and prevention paradigm [that will] require a more complex framework” (Rainey et al. 2018). Health authorities will have to address these issues at the organizational level, including that of equity in the delivery of services. Our analysis allows us to conclude that fostering inclusion through communication strategies and relying on electronic communication technologies could help in addressing equity issues. Although our results came from analysis of the Québec health system context, they can be useful for the implementation of the risk-based approach in other similar health systems. Our findings will inform and advise our new research project, a continuation of the PERSPECTIVE project (Personalized Risk Assessment for Prevention and Early Detection of Breast Cancer: Integration and Implementation (PERSPECTIVE I&I)). Since this new project aims to offer a breast cancer risk assessment to 10,000 Canadian participants and to recommend risk-adapted screening measures, we will propose to foster participant inclusion through various communication technologies and to rely on electronic communication technologies if possible.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Emmanuelle Lévesque http://orcid.org/0000-0002-4618-6635

Bartha M. Knoppers http://orcid.org/0000-0001-7004-2722

Additional information

Funding

This study was conducted with the financial support of the Quebec Breast Cancer Foundation, the Government of Canada through Genome Canada and the Canadian Institutes of Health Research, and the Ministère de l’Économie, de l’Innovation et des Exportations du Québec through Genome Québec.