A randomized controlled trial of shared decision-making treatment planning process to enhance shared decision-making in patients with MBC

A unique opportunity exists for patients with metastatic breast cancer (MBC) to engage in shared decision-making (SDM) due to the over 50 MBC treatment regimens (single drug or combination) included in the National Comprehensive Cancer Network guidelines [1], the sequential nature of systemic treatment over the course of a patient’s illness, and the varying toxicity profiles for different treatment regimens [2, 3]. Furthermore, given the incurable nature of MBC, individuals may have different treatment priorities and preferences (i.e., maximizing survival vs. quality of life).

Though most patients with cancer desire an active or shared role in treatment decision-making [4‐6], 40% of patients experience discordance between their preferred and actual decision-making roles, with most reporting less than desired involvement [6]. Our prior qualitative work demonstrated that oncologists’ treatment selection for MBC is prominently driven by treatment efficacy and physical side effects [7]. However, treatment factors considered by patients were more varied, and included treatment cost, logistics, side effects, salience of cutting edge treatments, and impact on personal responsibilities, important life events, and daily living [7, 8]. Discussions of preferences are uncommon since few avenues exist to capture and discuss treatment preferences due to numerous barriers including workflow challenges, limited clinician time, and misaligned incentives [9]. In one study by Williams et al., 67% of oncologists reported that patients’ personal quality of life priorities and preferences had a major impact on treatment recommendations, yet only 37% of patients reported having discussions about priorities and preferences prior to treatment selection [10].

Black patients may face additional challenges engaging in decisions. Black women report feeling less prepared for engaging in decisions or experience discomfort taking an active role when talking with physicians [11]. This is consistent with our prior study in which more Black women than White women with MBC preferred physician-led decision-making (32% vs. 28%) [12]. This may be related, in part, to differential patient activation, which includes the knowledge, skill, confidence, and ability to engage with the medical team to be able to take an active role in healthcare. Black patients have overall lower levels of patient activation, but higher activation moderates the relationship between implicit bias and patient-perceived quality of care [13]. Thus, consideration of this subpopulation is critical when evaluating SDM interventions.

One approach to enhance SDM is to incorporate patient-reported data into the process for generating treatment plans, which are lay language documents which the oncology team provides to patients, caregivers, and other healthcare providers that summarize the patient’s diagnosis, planned treatments, and potential side effects [14]. In early-stage breast cancer, treatment plans improved patient knowledge about their cancer and treatments, facilitated patient–clinician communication, enhanced coordination of care, and helped patients feel prepared for cancer treatment [15, 16]. However, these early studies did not leverage the use of patient-reported data. Prior literature highlights improvements in communication between oncologists, patients, and caregivers, as well as ability to identify better patient needs during treatment [17]. Thus, inclusion of patient-reported outcomes within the treatment planning process represents an opportunity to activate patients and increase shared decision-making.

Within our institution, standard-of-care treatment plans include patient-reported data (e.g., distress assessment, illness understanding, interest in clinical trials, presence of advance directives, preferred locus of control for decisions) [18]. Treatment plans are delivered by nurses after treatment is started to reinforce education, and the patient-reported treatment planning data are rarely reviewed by oncologists prior to treatment decisions. We hypothesized that presenting patient-reported data, including treatment preferences, to oncologists before or during a treatment plan decision would improve patient activation and shared decision-making. To this end, we conducted a randomized controlled trial to test the impact on decision-making outcomes of a patient-reported data-driven, oncologist-engaged, SDM treatment planning process compared to the standard treatment planning.

Although this randomized controlled trial in women with MBC did not meet the primary outcome with a statistically significant difference, the proportion of patients reporting SDM was 4% higher in the intervention arm than the control arm. In addition, non-significant, but numerically higher patient activation was observed for patients receiving the intervention compared to control patients (31% vs. 19% in highest activation level). Among Black patients receiving the intervention, 77% perceived the treatment decision as shared, compared to 65% of Black patients in the control arm. Finally, the favorable perception of the intervention from the oncologists and use in decision-making suggest that capture of patient preferences with subsequent review of patient-reported data as part of the decision-making process may enhance treatment decision-making and warrants further exploration.

Our study in patients with MBC complements existing studies on tools for SDM in early-stage breast cancer. In a study by Hawley and colleagues, the iCanDecide decision support tool demonstrated higher decision preparation and decision quality [35]. In another study of breast surgery decision aids, improvements in knowledge about treatment options, decisional conflict, and decision satisfaction were observed [36]. At the same time, these studies focused on specific decisions (e.g., mastectomy vs. lumpectomy) and providing knowledge to patients about those options. This differs from our study that provides a unique perspective on how oncologist review of patient-reported data can facilitate the incorporation of broader patient preferences and values into the treatment planning process.

In the majority of decisions, the oncologist used and valued the patient-reported data. This study adds to prior work by demonstrating how patient-reported data can be leveraged in clinical settings. For example, the geriatric assessment improves communication, reduces chemotherapy toxicities, and improves health-related quality of life compared to standard-of-care in older adults with cancer [37‐41]. In other studies, patient-reported symptom monitoring with associated nurse management improved symptom assessment [42‐44], patient–clinician communication and satisfaction [45, 46], and survival [47]. In aggregate, all of these studies highlight the opportunity for use of patient-reported data to better provide patient-centric care.

The greatest difference between the intervention and control group was observed in patient activation. Prior work demonstrates that higher activated patients are more likely to feel their treatment plans reflect their values, and they are more likely to tolerate side effects. In contrast, less activated patients have greater difficulty understanding their diagnosis and adhering to treatment [48]. We hypothesize that the discussion about their responses during the decision-making encounter contributed to the observed higher levels of patient activation among the intervention patients.

Inclusion of patient-reported preferences to guide decisions may be particularly important for patients whose voices may be historically unheard. One systematic review found that the majority of studies assessing patient–oncologist communication indicate that Black patients have lower quality patient–oncologist communication than White patients [34]. In a study of patients with lung cancer, Black patients viewed patient–oncologist communication to be less supportive than White patients [49]. In our study, more Black patients in the intervention arm perceived SDM (77%) than those in the control arm (65%). This may signal that a systematic approach to patient-reported data capture and review by oncologists could overcome known barriers in patient communication for vulnerable populations, which should be further explored.

The treatment plan process built upon for this intervention is designed to meet Oncology Care Model requirements, a requirement included in the current payment reform demonstration project, the Enhanced Oncology Model [50]. The novelty of this intervention is oncologist engagement through their review of electronic medical record-embedded patient dashboard and incorporation of patient-reported data into decision-making visits. The current version of the dashboard, which includes patient preferences, is shown in Fig. 6. Importantly, it does not require the oncologists to complete additional work to create or deliver treatment plans. The delegation of different healthcare team member roles (e.g., oncologist reviewing patient-reported data, nurse educating patient) remains important for the financial sustainability of the treatment planning process [51]. Also, focusing the oncologist role on review of patient-reported data likely contributed to the acceptability found in this study where the treatment planning process was noted to be helpful. In the setting of a busy oncology clinic, this structured approach of reviewing a simple patient-reported outcome dashboard may be more easily incorporated into existing workflows and overcome barriers related to limited oncologist training in preference elicitation. Future practice transformation efforts should carefully consider how to best engage oncologists in meaningful and efficient ways.

This study should be considered in light of several limitations. The study was conducted at a single academic medical center. There is potential for study contamination due to trained providers treating patients in both study arms. Such contamination would minimize observed impact. Furthermore, there was turnover and staffing shortages due to the COVID-19 pandemic. Sensitivity analysis accounting for time yielded similar results (not shown). When extended to the broader oncology community, use of the patient-reported data may differ between individuals. Additionally, this study focused specifically on patients with MBC, where there are multiple available treatment options, and thus it may not be applicable to patients with other cancer types, in the curative setting, or where treatment choices are limited. The patient-reported data also included multiple instruments in addition to the decision-related questions, and it is unknown which patient-reported data were most influential and longitudinal evaluation was not captured. Finally, although the target accrual of 140 decisions was met, the exclusion of patients due to non-evaluable patients and the expectation of a 20% difference in outcomes (based on prior studies outside breast cancer) resulted in the study being underpowered to detect smaller differences in patient activation or SDM. Thus, this study is hypothesis generating only. Future studies in a larger sample are warranted to further elucidate the role of patient-reported data in decision-making.

Oncologist engagement in the treatment planning process, with oncologist review of patient preferences and inclusion of patient-reported data in decision-making, is a promising approach to improve patient participation in treatment decisions. This approach may be particularly salient for certain subgroups of patients (e.g., Black women) who may face challenges to effectively communicating preferences during treatment discussion. Further trials are needed to assess the impact of this approach for broader and more diverse populations.