Introduction

Developing strategies to optimally support older breast cancer survivors is highly relevant to clinical practice, with more than 1.5 million breast cancer survivors aged ≥ 75 currently living in the U.S [1,2,3,4]. This number is anticipated to grow as the U.S. population ages and breast cancer mortality improves [1,2,3,4], yet guidelines for breast imaging in older breast cancer survivors [5] lack individualization or incorporation of life expectancy. Although screening mammography guidelines (for women without a history of breast cancer) recommend discontinuation of routine testing (i.e., in absence of symptoms or exam findings) when life expectancy is limited [6, 7], survivorship guidelines do not address how to tailor surveillance mammography by life expectancy [8, 9]. Thus, it is not surprising that conversations occur infrequently in clinical practice [9], with high utilization of surveillance mammography even when life expectancy is short [10].

We recently published expert consensus guidelines and talking points focusing on the older breast cancer survivor, with suggestions on how to approach surveillance mammography in the context of one’s life expectancy, age, the anticipated benefits and harms of testing, and patient preferences [9]. The guidelines acknowledge the long time-lag required to achieve minimal benefits in breast cancer-specific mortality from mammography[6, 11] and the more immediate harms of testing (e.g., false positives, overdiagnosis, over-treatment). As those guidelines were designed to support clinicians, they did not include patient-facing educational materials.

Engaging patients in patient-centered decision-making[12, 13] can ease exaggerated perceptions of cancer risk, provide reassurance, promote autonomy, and support clinicians by providing talking points for topics that may be uncomfortable, such as discontinuing a test to which patients/clinicians feel attached. Thus, shared decision-making has emerged as an effective approach for de-implementation of testing [14,15,16], particularly when benefit-to-harm ratios are uncertain or when benefits are outweighed by harms, such as the case of surveillance mammography for those with limited life expectancy. The value of shared decision-making in de-implementation has been demonstrated for screening mammography in primary care settings, facilitating informed decision-making, reduction in over-screening, improved patient knowledge, and satisfaction [16,17,18,19,20,21,22,23,24,25,26,27].

Recognizing the lack of decision support for older breast cancer survivors and their clinicians around surveillance mammograms, we created a patient-facing guide to complement the expert consensus guidelines [9]. Herein, we present results from acceptability testing of the “Are Mammograms Still Right for Me?” guide.

Materials and methods

Guide creation

We first obtained broad feedback on the guide’s content from multidisciplinary clinicians and patient advocates serving on our expert consensus panel and from primary care and oncology clinicians participating in five focus groups during guideline development [9]. In collaboration with the Dana-Farber Cancer Institute (DFCI) Health Communications Core, we then created a four-page “Are Mammograms Still Right for Me?” informational pamphlet (also available as a PDF file; shown in its final, revised form in Supplemental File S1), which applied the fundamental theory from the Ottawa Decision Framework[28] and which adapted a decision aid developed for screening mammography [16, 19, 26, 27].

The guide was prepared for an 8th grade reading level (per Flesch-Kincaid literacy criteria) [29], providing information specific for older breast cancer survivors on: (a) why mammography is a decision, (b) estimations of ipsilateral and contralateral breast cancer risk, including risk for recurrences or new primary cancers (i.e., ‘in-breast’ cancer) risks, (c) potential benefits/harms of mammography, (d) how overall health impacts surveillance mammography’s benefits [9], and (e) discussion prompts for ‘what’s important to you?’, encouraging patients to use the guide to deliberate with clinicians. The guide has bulleted text and short sentences to enhance readability, uses pie charts to display the risks for in-breast recurrences and new primary cancer events, and aims to provide a balanced list of the benefits/harms of mammography. The guide’s estimation of in-breast cancer risks is derived from a comprehensive literature review and is based on one’s personal cancer and treatment history [8, 9].

In preparation for in-clinic acceptability testing, we conducted fifteen semi-structured telephone interviews with breast cancer survivors aged ≥ 75. We elicited input on the guide’s content and clarity, while also eliciting intentions (or plans) for mammography and conversations about life expectancy (since one’s life expectancy informs high-quality mammography decision-making). Overall, participants provided affirmative feedback on the guide’s format, length, and content. Because there were no consistent concerns identified, the guide was not modified before in-clinic testing. This research was approved by the DFCI Office of Human Research. Informed consent was obtained from all individual participants included in the study.

In-clinic acceptability testing—approach

We included breast cancer survivors who were aged  ≥ 75 (where discontinuation of mammography may be appropriate given median U.S. life expectancy of ~ 10 years at age 75) and who received care at DFCI (Boston, MA) or two community-based satellite practices (Weymouth and Brighton, both in MA). All patients had to have completed active treatment(s) for their breast cancer (endocrine therapy allowed) and were required to read/speak English. We conducted feasibility testing in oncology practices because in prior focus groups of oncology and primary care clinicians, we learned that primary care clinicians defer to oncologists for decision-making around mammography [9].

Trained clinical research coordinators (CRCs) scanned clinic schedules for the upcoming four to six weeks to identify potentially eligible patients. Because we enrolled during the COVID-19 pandemic, we allowed virtual visits. Visit clinicians (physicians, nurse practitioners, physician assistants) were notified about plans to approach each patient and were allowed to opt out (no clinicians opted out). The CRC then contacted the patient to explain the study’s purpose and obtain verbal consent. Once enrolled, patients were mailed or emailed the guide; visit clinicians were notified.

The day before the visit, the CRC reminded the patient and clinician about the visit, encouraged them to use the guide during the visit, and administered the pre-visit survey to the patient. The survey was adapted from those used in mammography screening settings[16] and included questions about mammography intentions [19, 30], decisional conflict [31,32,33], demographics, numeracy [34], and health literacy [35], using validated scales and definitions (eTable 1). In addition, because the patient guide was designed to complement expert consensus approaches to mammography that emphasize considerations of life expectancy, we surveyed patients’ comfort level and preferences regarding life expectancy discussions. After the clinic visit, the CRC administered the post-visit survey within a week (with up to three reminders) and provided a $40 gift card, concluding study participation. The patient post-visit survey included items similar to pre-visit surveys plus acceptability questions about the guide (discussed below). We surveyed all clinicians via email following each visit, asking questions on guide acceptability, intentions (or plans) for mammography for that patient, and comfort with life expectancy discussions. Participating clinicians received a one-time $25 gift card at end-of-study. All patient and clinician surveys were administered via REDCap.

In-clinic acceptability testing—analyses

Our primary endpoint was guide acceptability, defined as ≥ 75% of unique patients and clinicians reporting each of the following responses in post-visit surveys: (a) would recommend use by others, (b) clear or mostly clear in its explanations, (c) helpful, and (d) with a suitable amount of information (right amount, a little more, or a little less information than needed). We tabulated responses to these questions on post-visit patient and clinician surveys; all analyses for the primary endpoint were descriptive.

For clinician acceptability, we examined responses to each relevant question by clinician (up to 19 submitted surveys, designated as clinicians A-S in Fig. 1). If a clinician stated that they did not use the guide with one patient but then used it with a second patient and rated it favorably, we counted the response provided for the second patient towards acceptability. Clinicians who never used the guide (e.g., clinician D) were not included in the denominator assessing acceptability.

Fig. 1
figure 1

Responses for acceptability of the patient guide by clinician (n = 19)

Full size image

Pre-visit patient surveys collected demographics, health literacy, numeracy, and preferences for decision-making roles. Post-visit surveys included general guide feedback (e.g., length, clarity). Both surveys inquired about intentions for mammography plus past and current (study visit) experiences and comfort level with life expectancy discussions.

From post-visit clinician surveys, in addition to acceptability, we described clinician responses for each participating patient for questions that were relevant to a particular patient (e.g., “What did you recommend in terms of mammograms for this patient?”, “Did you discuss the pros [and cons] of mammography”). In addition to descriptive analyses and extraction of explanatory comments (when provided), we used t-tests to examine changes in patient decisional conflict around mammography from the pre- and post-survey to inform design of a future trial of the guide. All analyses were conducted using SAS.

Results

Among 88 patients approached between August 18, 2019 and May 11, 2020, 18 declined enrollment, 22 could not be reached, three were ineligible, and 45 women enrolled; no clinicians opted out for their patients. The 45 patient participants (ages 75–92, median = 78 years) had clinic visits with 21 unique oncology clinicians (six nurse practitioners, one physician assistant, 14 physicians); participating clinicians had one to seven patients who enrolled. Of the 45 enrolled patients, 40 completed at least some of the post-visit survey (three could not be reached, one rescheduled her visit twice and never completed the survey, one declined). Among 21 clinicians, 19 completed at least one survey. Patient characteristics are summarized in Table 1; 43 (96%) of patients reported White race, five (11%) reported a high school degree as their highest education, and two (4%) reported feeling extremely good about working with fractions, reflecting high numeracy. One woman reported ever having conversations about life expectancy, although 19 (42%) expressed interest in knowing this information, particularly if life expectancy was < one year. Extreme worry regarding new breast cancers was infrequent.

Table 1 Patient participant characteristics from pre-visit survey for in-clinic acceptability testing (n = 45)
Full size table

Acceptability

Acceptability of the guide by patients (Table 2) and clinicians (Fig. 1) was high; overall, clinicians were consistent in post-visit responses across patients. Among patients and clinicians completing post-visit surveys, 33/37 (89%) patients and 15/16 (94%) clinicians stated they would recommend the guide to others; 33/37 (89%) patients and 15/16 (94%) clinicians reported that everything/most things were clear (including one clinician who answered ‘most things were clear’ for one patient and ‘some things were clear’ for another). In addition, 20 patients (54%) and 96% of clinicians reported that the amount of information provided was suitable. Finally, 84% of patients and 80% of clinicians reported the guide was helpful in making decisions about mammography. All of these responses met prespecified criteria for acceptability.

Table 2 Acceptability and feedback on visit and patient guide (patient post-visit survey, n = 40)
Full size table

Additional clinician feedback (Table 3).

Table 3 Post-visit clinician survey responses by patient visit (N = 41 surveys among 19 clinicians)
Full size table

Overall, clinician feedback on the guide was positive, with 73% stating the length was ‘just right’ and only two clinicians preferred a format other than a printed pamphlet. Several clinicians provided written comments: “…the guide is absolutely wonderful”, and “It would be great to have this…readily available for…discussions with our older patients.” Only one clinician reported that the guide made the visit a ‘lot longer’: “I worry about the ability of busy clinicians to use it during their constrained visits.” Even when clinicians recommended that a participating patient continue mammography, open-ended comments acknowledged (a) the general appropriateness of these discussions, (b) the importance of patient preferences, and (c) when it may not be appropriate to have these discussions.

Clinicians did not use the guide in 22% of visits, with some stating that patients required additional testing (n = 3) or were recommended to continue mammography (n = 4) because of higher-than-average risk for future in-breast cancers, excellent life expectancy, or patient preferences: “My bias was against recommending mammograms…the guide helped us think clearly about pros/cons in a very healthy patient with history of bilateral breast cancers who is more reassured to continue mammograms.” Another clinician stated: “[T]here are many patients who you already know will want to continue due to high anxiety.” Overall, patient preferences were strong for mammography (see below); five clinicians reported that they would have recommended discontinuation of mammography before the visit, but only one clinician reported recommending discontinuation after the visit.

Overall, in 29/41 visits (71%), clinicians reported discussing the pros and cons of mammography and less commonly life expectancy (18/41 visits [44%]). Most clinicians found the guide helpful, and none reported uncomfortable conversations (though one perceived the patient as uncomfortable): “…This tool makes it easier to discuss comorbidity and life expectancy; less awkward”, and “…having a printed tool made the life expectancy discussions…comfortable…It made it less like I had some specific concern that I was bringing up; the discussions are more just a part of standard practice.” For those not discussing life expectancy, the most common reasons provided were “it didn’t come up” or the “patient was healthy, so I didn’t think it was important.”

Mammography intentions and additional patient feedback (Table 2 and eTable 2).

Intentions for mammography were very strong in patient pre- (100%) and post- (98%) visit surveys, with only one patient deciding to discontinue mammography after the visit (her clinician also recommended discontinuation). Patients had low decisional conflict on pre- and post-surveys, with mean total scores of 12.8 and 12.0, respectively (p = 0.79; eTable 2). Overall, in post-clinic surveys, 22 patients (56%) recalled discussing mammography’s benefits, eleven (29%) reported discussing mammography’s cons, and nine (23%) recalled life expectancy discussions (Table 2). Although one patient reported feeling ‘very anxious’, no patients reported increased anxiety after using the guide, and 54% of patients felt the guide prepared them ‘a great deal’ or ‘quite a bit’ to make a better decision on mammography.

Discussion

In acceptability testing, 45 women ages  ≥ 75 with history of breast cancer and with strong mammography intentions and their oncology clinicians found an information guide on whether or not to continue mammography acceptable and clear. Overall, 89% of patients and 94% of clinicians stated that they would recommend the guide to others. Patients and clinicians also found the guide reasonable in length and suitable in content. Patients reported no increased anxiety after using the guide, and clinicians were enthusiastic to have the guide available for in-clinic use.

The high mammography utilization observed is consistent with prior observations [10], underscoring the importance of providing support and education to clinicians and patients in this context, likely through longitudinal discussions. Although five clinicians in our study stated they had intended to recommend discontinuing mammography before the visits, only one recommended discontinuation in the post-visit survey, perhaps due to strong patient preferences elicited during visits (though we did not ask why recommendations changed). Another important contributing factor is that 71% of patients in acceptability testing were ages 75–79, an age range where some will have > 10-year life expectancy and will be appropriately recommended to continue mammography [9]. Even when the timing is appropriate for discontinuation, stopping mammograms in survivors may be challenging because of worry related to one’s personal cancer history and the often ingrained, perceived benefits of mammograms, without tools to adequately improve patients’ understanding of the harms of testing.

It is thus not surprising that a one-time intervention with a patient guide may not be sufficient in itself to impact mammography decisions. Instead, this may require repeated discussions over time with engaged clinicians, including primary care clinicians, ideally starting a few years ahead of when discontinuation of mammography will be considered. The patient guide and clinician-facing expert consensus guidelines[9] were developed with this in mind, with the goal to provide reassuring information that promotes informed decision-making, all in the context of one’s life expectancy, preferences, and underlying risk for new cancers. In clinical testing, when the guide was utilized, clinicians reported it was useful in these discussions. However, future testing of a patient guide will include alternative approaches to conveying this information to patients to better accommodate time constraints in clinic.

This high utilization of mammography among older breast cancer survivors in our study also confirms the clinician feedback received during our expert consensus guideline development [9], where focus group clinicians acknowledged the prevailing, habitual continuation of mammograms and difficulties incorporating conversations about life expectancy [9], consistent with the infrequent conversations reported in our study. Interestingly, although nine patients recalled discussing life expectancy, eighteen clinician surveys reported addressing this topic during the visit. Moreover, while 70% of oncology clinicians’ responses indicated they reviewed the downsides of mammography, only 29% of patients perceived this topic was discussed, highlighting differing visit perceptions for clinicians and patients; future study could include observation of discussions to assess for shared decision-making.

To our knowledge, the “Are Mammograms Still Right for Me?” guide is the first resource for older breast cancer survivors to facilitate shared decision-making on mammography. Together with the clinician-facing guidelines and talking points[9], these materials provide much-needed, multifaceted support for patients and their clinicians, including multidisciplinary physicians and advanced practice providers, that informs decisions rather than reflexively and indefinitely promoting mammograms. Based on the constructive feedback we received from acceptability testing, we revised the guide to include a more numerically balanced list of the benefits and harms of mammography, citations for the approximations for in-breast cancer events, and more reassuring text that physical exams and diagnostic evaluations will continue even if routine mammography is discontinued (Supplemental File S1).

We recognize several study limitations. Our study was limited is size and did not include a control population, non-English speaking patients, or a sufficiently diverse sample with regard to race, ethnicity, and socio-demographic factors. In addition, we did not mandate use of the guide or evaluate life expectancy and could not guarantee that the timing for mammography discussions was appropriate. Also, we included only oncology clinicians; we plan to include primary care clinicians in future evaluations and disseminations of the guide. Despite these limitations, clinicians rated the guide favorably and we obtained important initial implementation experience, all providing preliminary results for use of the guide in clinical practice.

Our results for acceptability will inform larger-scale studies that better engage diverse patient populations across various clinical settings and test the implementation of the patient guide through a multi-level, shared decision-making intervention that incorporates longitudinal patient-, clinician-, and practice-facing elements. This guide has the potential to enhance clinical practice by facilitating individualized decision-making for surveillance mammography among older breast cancer survivors but will require further study in larger, diverse practice settings and populations, and in particular, those with more limited life expectancy who are in most need of discussions on discontinuation of testing. The ultimate goal of this work is to further assess impact, standardize practice, and disseminate these novel resources to the growing number of older patients (and their clinicians) who are in urgent need of tailored approaches to survivorship care. Offering multi-layered support to pragmatically individualize mammography with a reassuring and informative approach has the potential to facilitate de-implementation of routine mammography when it is unlikely to provide benefit and may cause harm.