Though ≥80% of urban, suburban, and rural survivors agreed that a tool would be useful, agreement that a tool would increase confidence to discuss exercise was higher among suburban survivors (83.7%) compared to urban (69.0%) and rural survivors (63.6%). By contrast, rural survivors reported greater agreement (90.0%) that they would use a tool to obtain a referral compared to suburban (81.6%), and urban (66.7%) survivors. Survivors in a relationship had greater agreement for access to home exercise resources (87.7% vs. 63.3%) and lower agreement for childcare (60.0% vs. 80.0%) and finances (60.8% vs. 70.0%). Women with a baccalaureate degree or higher tended to report greater agreement for most inputs compared to women with some college education who, in turn, tended to report greater agreement than women with a high school education or lower.
Contemporary cancer exercise guidelines recommend providing individualized exercise prescriptions [10,11,12]. A variety of approaches (e.g., the Moving Through Cancer initiative [32], decision algorithms [i.e., models] such as Exercise in Cancer Evaluation and Decision Support [EXCEEDS] [33,34,35], and Cancer Exercise application [36]) have been developed to support offering exercise prescriptions to or triage to services for cancer survivors. However, to our knowledge, there are no evidence-based clinical support systems capable of generating individualized outcomes (e.g., risk of breast cancer death) associated with physical activity for breast cancer survivors considering their individual characteristics [30].
Overall, breast cancer survivors’ perspectives on a tool that could offer individualized exercise recommendations to support exercise-related decisions in clinical settings were positive. Education and encouragement were the highest rated tool uses, indicating that a tool may help to address gaps in knowledge and support relating to exercise participation among breast cancer survivors [27, 52,53,54]. Women also indicated that they would use a tool to receive exercise referrals. Therefore, a tool may help increase the frequency and uptake of exercise referrals, which are persistent issues in cancer survivor exercise promotion [55, 56].
Most women indicated that they would use a tool to support shared decision-making with their healthcare provider regarding exercise. Shared decision-making is an important part of cancer survivors developing survivorship care plans in collaboration with their healthcare providers [57], and is associated with breast cancer survivor exercise participation [58]. Thus, findings align with those of previous studies that found breast cancer survivors want exercise to be included and better addressed a part of survivorship care plans [59,60,61]. Further refinements may be required to facilitate shared decision-making about exercise in clinical settings. For example, providing survivors with individualized exercise information generated by a tool prior to clinical consultations may help survivors to deliberate and be aware of their options under less time pressure [62,63,64]. Different modes of communication could also be considered. For instance, a pre-visit evidence-based educational video has been shown to increase knowledge and decrease distress among breast cancer survivors while stimulating thoughts and questions about their care [65].
Breast cancer survivors rated age, readiness to exercise, and access to exercise resources at home as the most important tool inputs. Studies have shown that behavioral change theories are useful when promoting exercise among breast cancer survivors [66,67,68]. Our finding that survivors viewed exercise readiness as an essential input reinforces the importance of considering behavioral change theories when developing a tool and accompanying messages to promote exercise, especially since readiness to exercise is closely aligned with constructs from behavioral change theories such as self-efficacy, motivation, intention, and attitude [69,70,71,72,73,74].
The importance of access to home resources highlights the need for healthcare professionals to consider the personal preferences and resources at the disposal of survivors when providing exercise recommendations. For example, exercise preferences with respect to the physical location (e.g., home vs. healthcare setting), supervision (i.e., supervised vs. unsupervised), and the social environment (i.e., alone vs. with company) can vary among survivors [75,76,77,78,79,80,81,82]. However, some survivors may not have home exercise resources and/or the financial means to participate in supervised exercise. A tool that considers the resources available to survivors in addition to their clinical needs and preferences could provide healthcare providers an opportunity to individualize care based on survivor’s needs and preferences. For instance, women diagnosed with more advanced cancer could be referred to a lower intensity supervised exercise program, whereas women with earlier stage cancer may receive information and resources to support home based exercise or a referral to community exercise programs. Better consideration of women’s exercise preferences may improve women’s perspectives on the tool, as well as their uptake of provider advice and referrals.
It is worth noting that ‘race and ethnicity’ was among the lowest rated inputs. Some have advocated that race be included in clinical decision tools when it is directly linked with racism and contextual factors [83]. However, race-based medicine has a problematic history in the U.S [84]. Further work is needed to establish whether, and if so how, race and ethnicity should be included in a tool that offers individualized exercise information for breast cancer survivors, particularly in light of the ongoing debate, combined with recent evidence that healthcare algorithms could mitigate, perpetuate, or exacerbate racial/ethnic disparities irrespective of the explicit use of race or ethnicity [85].
Shorter-term outcomes such as the ability to do everyday tasks, quality-of-life, energy, and sleep rated among the top benefits. Though many of these ‘benefits’ are the inverse of the ‘conditions’ associated with breast cancer and related treatment that were examined, the outcomes were rated higher as benefits compared to conditions. The opposite was true for cognition, which was lowest rated benefit (55.2%) but the highest rated condition (75.5%). Further work is warranted determine survivors’ preferences regarding of gain or loss framing for each outcome. For example, many breast cancer survivors experience sleep distrirbances [86, 87]. Whether a sleep-related outcome should be framed in terms of ‘improved sleep’ and/or ‘less likely to experience sleep disturbances or insomnia’ is worthy of investigation in future research.
We assessed the potential differences in tool usefulness, uses, and inputs based on breast cancer survivors’ demographic, clinical, and contextual characteristics. The differences revealed highlight the need for individualized exercise recommendations for breast cancer survivors [10,11,12], and may offer insight to inform the development of a tool to address existing disparities in exercise communication and participation. For example, women diagnosed with later-stage or more aggressive forms of breast cancer perceived the tool less favorably. These findings suggest that the tool needs to be updated to consider the unique exercise preferences of these women [76, 77, 88] and support them in overcoming barriers they may encounter such as, financial toxicity (i.e., financial problems related to the cost of medical care) [89] and impairments specifically related to cancer and treatments [82, 90]. The perspectives of women who participate in less physical activity, are older, and are less affluent are also needed as the sample recruited was not representative of U.S. breast cancer survivors.
The lower ratings of shared decision-making compared with other tool uses may be indicative of a level of medical mistrust or uncertainty regarding the motives clinicians or healthcare institutions [91,92,93]. Studies have shown lack of trust in the healthcare system may vary based on race, ethnicity, and education among breast cancer survivors [92, 93]. Health literacy may also influence patient-healthcare provider communication and shared decision-making [94, 95]. Further research is warranted to understand whether, and if so how, a tool could address medical mistrust and health literacy to support patient-provider communication, especially among historically underserved and underrepresented populations. For instance, translation of information into women’s preferred language may mitigate health literacy barriers related to language [96]. Culturally-tailored exercise education and information may also help mitigate the effects of medical mistrust and health literacy [97, 98].
The benefits with the greatest agreement for inclusion in a tool were the ability to do everyday tasks, quality-of-life, and energy. Breast cancer survivors who initiate exercise are likely to experience these exercise-related benefits [10, 99], which in turn may improve their motivation, efficacy, and intention to habitually adhere to exercise long-term [100]. Thus, a tool providing individualized information about the exercise benefits for which breast cancer survivors have an interest and are likely to experience has the potential to meaningfully increase exercise participation, and ultimately, health.
Our study has several limitations. The women in our study were younger, more affluent, and more physically active compared to the wider population of breast cancer survivors in the U.S. This is indicative of self-selection bias and may limit generalizability and explain some of findings. For example, inputs concerning women’s physical and functional impairments were rated low, despite such impairments being frequently reported as barriers to exercise by older breast cancer survivors and/or women diagnosed with advanced breast cancer [27, 28, 82]. As such, items concerning physical and functional may still need to be included in a tool for it to meet the needs of the wider population of breast cancer survivors. The modest sample size constrained our ability to rigorously examine differences in perspectives based on demographic, clinical, and contextual characteristics. Reliance on self-reported measures also means responses were susceptible to social desirability bias [101]. However, our sampling strategy ensured the inclusion of the perspectives of women who have historically been excluded from cancer research [102, 103]. We did not examine specific exercise participation preferences (e.g., intensity, setting), insurance coverage for exercise, or willingness to pay out-of-pocket of and/or travel to participate in exercise. Future research is warranted to examine how these factors could be incorporated into a tool. Finally, given the time constraints experienced in clinical settings, establishing how many inputs can feasibly be included in a tool is a crucial next step.
Future directions
Future work should seek to acquire insight into the perspectives of a more representative sample of breast cancer survivors. In particular, survivors who are less physically active, older, have lower socioeconomic status, and reside in rural communities. Further research is also needed to understand how to best communicate exercise information generated by the tool to survivors. The performance of the tool in comparison to usual care and other methods of offering exercise prescriptions to breast cancer survivors will also need to be tested.
Our intention is to continue to iteratively develop this clinical decision tool with input from survivors and healthcare providers. The perspectives of healthcare providers (i.e., breast oncologists, primary care physicians, exercise specialists, occupational therapists, physical therapists, advanced care providers, nurses, patient navigators, and social workers) on the same clinical decision tool are reported in a separate paper [37]. The wireframe will be revised based on the findings reported in this formative research before being revised again following future cognitive interviews with healthcare providers and more representative breast cancer survivors. The revised wireframe will then be translated into a functional prototype web-based clinical decision tool, which will be tested for feasibility and acceptability. The version of the tool that is tested will incorporate items that elucidate patient preferences concerning exercise participation (e.g., exercise frequency, intensity, type; willingness to pay and/or travel for exercise) and communication (i.e., how survivors would prefer to receive exercise information and from whom) using a question prompt list currently being developed. Ultimately, a Beta-version tool will be used to conduct a clinical trial that will evaluate the effects of individualized exercise prescriptions on exercise self-efficacy, exercise intentions, and exercise participation of breast cancer survivors.
