TBCRC 057: Survey about willingness to participate in cancer clinical trials during the pandemic

BACKGROUND

Breast cancer patients experienced heightened anxiety during the pandemic. Also, modifications to clinical trial activities allowing for virtual platforms, local assessments, and greater flexibility were introduced to facilitate participation. We sought to evaluate the association between pandemic-related anxiety and willingness to participate in trials and how pandemic-era modifications to trial activities affect the decision to participate.

METHODS

We conducted an online survey from August to September, 2021 of patients with breast cancer assessing pandemic-related anxiety; clinical trials knowledge and attitudes; willingness to participate during and before the pandemic; and how each modification affects the decision to participate. Fisher's exact tests evaluated differences in proportions and two-sample t-tests evaluated differences in means. The association of pandemic-related anxiety with a decline in willingness to participate during compared to prior to the pandemic was modeled using logistic regression.

RESULTS

Among 385 respondents who completed the survey, 81% reported moderate-severe pandemic-related anxiety. Mean willingness to participate in a trial was lower during the pandemic than prior [2.97 (SD 1.17) vs. 3.10 (SD 1.09), (p < 0.001)]. Severe anxiety was associated with higher odds of diminished willingness to participate during the pandemic compared to prior (OR 5.07). Each of the modifications, with the exception of opting out of research-only blood tests, were endorsed by >50% of respondents as strategies that would increase their likelihood of deciding to participate.

CONCLUSIONS

While pandemic-related anxiety was associated with diminished willingness to participate in trials, the leading reasons for reluctance to consider trial participation were unrelated to the pandemic but included worries about not getting the best treatment, side effects, and delaying care. Patients view trial modifications favorably, supporting continuation of these modifications, as endorsed by the National Cancer Institute and others.

The painting of a patient: Provider representations of patient fitness and preferences in multi-disciplinary tumor board meetings

246

Background: Pancreatic cancer treatment can involve multiple treatment modalities (e.g. chemotherapy, surgery, radiation) that vary greatly in intensity and timing. The most intense treatments are associated with substantial toxicity and only recommended for medically ‘fit’ patients regardless of age. Despite evidence-based guidelines demonstrating the benefit of including the geriatric assessment (GA) and patient preferences, these may not be formally included in multidisciplinary tumor boards (MDTBs) where treatment decisions are often made. Methods: This qualitative study evaluated MDTB meetings at a single institution between November 2021 and February 2022 for inclusion of fitness and patient preferences. Discussions of patients with stage I-IV pancreatic cancer were identified for inclusion. These recordings were transcribed and analyzed using NVivo for recurring themes and exemplary quotes regarding how providers characterize patients’ fitness and present their preferences to the board. Results: Thirteen MDTB meetings including 50 individuals with stage I-IV pancreatic cancer were included. Descriptions of patient fitness largely consisted of the presentation of common demographic traits such as age and gender. Additional context, primarily focused on occupation, comorbidities, and patient attitudes, was provided when demographic information did not align with perceived fitness level; for example “He’s 60 years old. He’s actually very strong. He used to, or he does, work in landscaping”. There was no formal inclusion of GA data. Explicit references to patient preferences or agentive decision-making only occurred in 19 cases; 11 referring to the patient’s (non-)interest in clinical trials, 3 denying additional testing (tone suggesting non-compliance), and 7 specifically referring to treatment type preference (e.g. surgery vs. chemotherapy). Conclusions: At present, MDTB treatment discussions primarily rely on shared understandings of the cultural significance and meaning of various demographic traits and rarely reference patient-reported preferences or fitness when determining the best pathway for patient care. Further work is needed to understand how to better formally incorporate fitness and preferences into decision-making.

“It sounds very negative”: Patient perspectives on de-escalation of treatment concept and language

254

Background: As the combination of systemic and targeted chemotherapies is associated with severe adverse side effects and long-term health complications, there is interest in reducing treatment intensity for patients with early-stage breast cancer (EBC). Currently, this approach is being described as “de-escalation,” though there is some concern that this language and framing are not well understood or preferred by patients. Methods: We collected qualitative interview data from twenty-four patients with Stage II-III breast cancer and sixteen patient advocates. Interviews explored interest in participation in clinical trials testing reduced amounts of chemotherapy, reactions to the word “de-escalation,” and preferred ways to describe this approach and its anticipated benefits. Interviews were audio-recorded and transcribed, and researchers used qualitative content analysis to code for dominant themes. Results: Twenty-three participants (57.5%) expressed interest in participating in a trial of reduced chemotherapy. However, 60% of participants had a negative response to the word “de-escalation,” including 78% of respondents that said they were interested in the concept. To describe this approach, respondents tend to prefer words like “personalization,” “optimization,” or “reduced chemotherapy.” Participants also expressed the importance of provider-patient communication, clear and simple language, sharing the decision-making process, and continued support throughout and after treatment in their consideration of trials testing lower amounts of chemotherapy. Conclusions: Among individuals with EBC, there is significant interest in alleviating treatment-related toxicity by reducing chemotherapeutic intensity. Patients are more apt to feel comfortable participating in trials testing this approach if they are framed in terms of customizing treatment to the individual patient and added benefit—reduced toxicities, higher quality of life during treatment and lower risk of long-term complications—rather than in terms of taking treatments away or doing less than the standard of care. “De-escalation” is not an ideal word to describe these trials as many associate this word with giving up in the war against cancer. Instead, participants prefer positive, patient-centered language. This information will be useful to providers in considering how best to describe de-escalation clinical trials to eligible patients in ways that avoid therapeutic misconception and facilitate the shared decision-making process regarding treatment.

Interest and enrollment in clinical trials by race and ethnicity, rurality, and insurance status in patients with ovarian cancer

83

Background: Enrollment in Phase III clinical trials in gynecologic cancer patients has decreased by approximately 90% since 2011. Efforts to increase enrollment are needed. Previous research showed that not all eligible patients are approached, and those who are do not consistently enroll. Data regarding enrollment by patient characteristics, such as race/ethnicity, rurality, and insurance status is limited. We aimed to identify how these characteristics affect enrollment in ovarian cancer patients. Methods: We conducted retrospective chart review for patients with incident ovarian cancer presenting to the University of Alabama at Birmingham from 1/2017-3/2020. We abstracted patient race, ethnicity, Rural-Urban Commuting Area (RUCA; rural vs urban) and Area Deprivation Index (ADI; most vs least disadvantaged) based on Census tract codes, insurance status, eligibility for available trials, and trial participation from medical records. Patient interest in participation was abstracted from a patient-reported outcomes database. We calculated descriptive statistics and estimated enrollment as a multivariate function of age, race, ethnicity, insurance, RUCA and ADI using binomial logistic regression. We reported associations as odds ratios with 95% confidence intervals. Results: Of 156 patients, 25% were Black, Indigenous, or Persons of Color (BIPOC). 19% lived in a rural area. Mean age was 62 (SD 11.7). Most (95%) patients were insured; 49% Medicare, 40% private insurance, and 6% Medicaid. 126 (81%) were eligible for a trial during their treatment course. Of 102 patients who completed the question on clinical trial interest, 58% were interested; 42% were not. Ultimately, 36% of the 102 enrolled in a trial including 47% of those initially interested and 21% of those not. 39% of white patients (n = 117) initially expressed interest in a trial compared to 33% of BIPOC (n = 39); 48% of white patients ultimately enrolled vs 23% BIPOC. Of patients living in urban vs rural areas with known interest, patients in urban areas had higher interest (44% vs 10%) and higher enrollment (44% vs 31%). Among insurance types, interest and enrollment differed (Medicare (n = 76) 33% and 1%, Private (n = 63) 46% and 46%, Medicaid (n = 9) 33% and 22%, no insurance (n = 8) 25% and 36%). In our adjusted analysis, BIPOC patients had lower odds of enrolling onto clinical trials compared to white patients (OR 0.32, 95% CI 0.13-0.76). Additionally, as age increased by 1 year, odds of enrollment decreased (OR 0.96, 95% CI 0.92-0.99). Conclusions: BIPOC identity and older age were associated with lower rates of clinical trial enrollment. Comprehensive eligibility screening and early introduction could improve enrollment, particularly among BIPOC and older patients. These efforts have potential to improve enrollment as a greater percentage of patients ultimately enrolled on trial than initially expressed interest.

Associations between insurance status and the cancer clinical trial enrollment process

85

Background: Most patients with cancer experience multi-leveled barriers to clinical trial participation, potentially including financial concerns due to the complexity surrounding trial-related insurance coverage. Our study sought to understand the association between insurance status and cancer clinical trial eligibility, offer, and enrollment. Methods: This retrospective cohort study included patients with breast or ovarian cancer receiving a therapeutic cancer drug at the University of Alabama at Birmingham between January 2017 and February 2020. Available clinical trials and eligibility criteria were abstracted from OnCore and ClinicalTrials.gov. Patient trial eligibility, offer from provider, demographics, and clinical characteristics were abstracted from electronic medical records. Patient trial enrollment was determined via OnCore. Odds of clinical trial eligibility, offer, and enrollment by insurance status (private, public [Medicaid, Medicare]) were estimated using logistic regression models. Models estimating odds of trial offer and enrollment contained only eligible patients. Models were adjusted for patient age at diagnosis, race and ethnicity, rural-urban residence, Area Deprivation Index, cancer type, and cancer stage (early, late). Results: A total of 513 patients with breast (71%) or ovarian (29%) cancer were included in our analyses. Median age at diagnosis was 60 (interquartile range: 49-67) years; the majority were White (69%) and had early stage cancer (65%). Half of patients had private insurance (54%), and 46% of patients had public insurance (38% Medicare, 8% Medicaid). Patients with private insurance more often had early stage cancer compared to patients with public insurance (73% vs 57%). Almost two-thirds of patients (65%) were eligible for clinical trial enrollment. Of eligible patients (n = 333), 68% were offered a trial and 47% enrolled onto a trial. In adjusted analyses, patients with public vs private insurance had similar odds of clinical trial eligibility (odds ratio [OR] 0.95, 95% confidence intervals [CI] 0.61-1.48), being offered to participate (OR 1.23, 95% CI 0.71-2.14), and clinical trial enrollment (OR 1.13, 95% CI 0.68-1.89). Conclusions: Our results suggest oncologists do not assess trial eligibility or offering a trial based on insurance status, and patients do not differentially participate based on their insurance coverage. Further research is needed to understand implications of trial participation (e.g., out-of-pocket and time costs) for patients covered by differing insurance.

Development of a patient-reported outcome measure (PROM) screening strategy for early palliative care needs in outpatients with advanced cancer

203

Background: There is a critical need to develop standardized and feasible methods to monitor patients for unmet palliative needs and direct timely referral to specialty palliative care. We developed a patient reported outcome measure (PROM) measurement strategy to screen patients for multidimensional palliative care needs. Methods: Guided by evidence-based frameworks for early palliative care in oncology, we identified 8 key domains for PROM monitoring that are meaningful and actionable in clinical care and appropriate for direct patient report. We conducted a systematic search for PROMs assessing these key domains using the Palliative Care Research Cooperative (PCRC) Measurement Core resources and the Grid-Enabled Measures Database. PROMs for each domain were compared for content coverage, psychometric properties, proprietary availability, Spanish translation, and attributes (response options, length, literacy demand). Results: We selected 13 PRO items for weekly monitoring (Symptom PROs) and 11 PRO items for monthly monitoring (Palliative PROs) (Table). We did not identify any validated PROMs to assess caregiver burden from the patient perspective. Validated PROMs in short formats for spiritual/existential needs are limited. Conclusions: Existing PROMs are limited in capturing the multi-dimensionality of palliative care needs for patients with cancer, particularly for spiritual needs and patient-reported caregiver burden. Future work will focus on piloting identified PROMs to monitor patients for early palliative care needs and determining thresholds to trigger referrals to specialty palliative care.[Table: see text]

Identification of target population in the implementation of navigator-delivered home ePRO for patients with cancer receiving treatment

351

Background: One key challenge of practice transformation activities, such as remote symptom monitoring (RSM) using electronic patient reported outcomes (ePROs), is identification of patients starting treatment. In real-world settings, reliance on referrals is likely to miss patients. We describe the difficulties encountered in patient identification and the subsequent changes implemented in protocol to remediate this. Methods: We conducted two PDSA cycles focused on identification and engagement of patients for RSM at the Mitchel Cancer Institute (MCI). Target patient capture was > 75%. Modifications to the patient identification process were documented. Schedules of physicians participating in the RSM program were reviewed from 6/2021 – 5/2022 to identify eligible patients. Patients were considered eligible if they were starting chemotherapy, targeted therapy, or immunotherapy. Patients seeking a second opinion were excluded. Patient demographics, cancer type, cancer stage, and PROs were abstracted from electronic health records and the PRO platform (Carevive). Initial clinic roll-out was conducted in gynecologic oncology, with expansion to breast and thoracic oncology in 10/2021 and 3/2022, respectively. The proportion of eligible patients approached per month was reported.Results: In the first PDSA cycle, the eligibility criteria was defined. Although clinical trials included advanced disease, non-clinical staff screening expressed concern about determining advanced vs. early-stage disease. Thus, inclusion criteria was broadened to include all patients starting treatments. From 6/2021 –8/2021, navigators identified patients by screening patients who presented for chemo-education visits. The navigation team approached 23 patients during this period. However, this process didn’t identify all eligible patients as not all patients beginning treatment received chemo-education visits. In PDSA Cycle 2, the process for new patient contact from initial call for appointment through treatment was reviewed. The implementation team screened all patients in a physician’s schedule a week prior to the office visit as well as on the day of visit. This updated process identified all eligible patients starting either intravenous or oral chemotherapy. The recruitment process was modified to screen the physician schedules rather than chemo educator visits. From 9/2022-5/22, the proportion of eligible patients identified and approached remained high at 100%. This methodological screening process helped the navigation team identify all eligible patients in an efficient manner and they reported comfort in expanding to additional disease teams. Conclusions: Systematic screening of physician schedules can be successfully leveraged for patient identification and reduce time spent manually screening for eligible patients by non-clinical navigators. Clinical trial information: NCT04809740.

Associations between patient-perceived cancer curability and advance directive completion based on race and ethnicity and cancer type

200

Background: Despite Advance Care Planning (ACP) recommendations for patients with cancer for over 20 years, uptake remains low, and many such patients lack Advance Directives (AD). Our previous data showed a lack of association between patient-perceived cancer incurability and AD completion, however less is known about whether this relationship remains for key subgroups where perceptions may differ, such as BIPOC (Black, Indigenous, or People of Color) patients and patients with incurable hematologic malignancies. Methods: This cross-sectional study used Patient Reported Outcomes (PRO) from surveys administered during routine outpatient care at the University of Alabama at Birmingham from 12/2016 to 08/2021. Patients self-reported AD completion and perception of curability. Demographics consisting of age, sex, race and ethnicity, and marital status, and clinical characteristics, namely cancer type, cancer stage (stage 0-III grouped as early vs. IV/progression/recurrence as late) and phase of care (initial being the first 12 months, survivorship starting after this, and end of life being the last 6 months of life) were abstracted from the electronic medical record. Descriptive statistics were calculated using frequencies and percentages for categorical variables and median and interquartile ranges (IQR) for continuous variables. Likelihood ratios (LR) and 95% confidence intervals (CI) were estimated using a modified Poisson regression with robust error variance to evaluate the relationships between patient-perceived incurability and AD completion. Models were subset for White patients (n = 879), BIPOC patients (n = 330), and patients with solid organ malignancies (n = 987). Models were adjusted for demographics and clinical characteristics. Results: Of 1209 patients, 73% were White, and 82% had solid organ malignancies, most commonly gynecologic (32%), breast (17%), and gastrointestinal (13%). The sample was predominantly female (70%) with early-stage disease (60%) and a median age of 66 (IQR 58-72). AD completion was 46%, 32%, and 41% for patients who were White, BIPOC, or had solid organ malignancies, respectively. In adjusted analyses, patient-perceived incurability was not associated with AD completion for any of the patient subgroups (White LR 1.06, 95% CI 0.89-1.35; BIPOC LR 0.93, 95% CI 0.60-1.44; solid organ LR 1.10, 95% CI 0.88-1.36). Conclusions: Patient-perception of incurability does not appear to be associated with AD completion even in subgroup analyses based on race and ethnicity or cancer type.

Sociodemographic difference in patients who enroll and decline remote symptom monitoring (RSM)

268

Background: Remote symptom monitoring (RSM) using patient-reported outcomes has been shown to reduce symptom burden and hospitalizations in clinical trials. However, little is known about how willing patients are to participate in remote symptom monitoring in real-world settings, particularly for vulnerable patient populations. This study aims to compare characteristics of cancer patients enrolled vs. patients who declined enrollment into RSM. Methods: This prospective study used data that assessed the characteristics of patients who enrolled vs. patients who declined enrollment into RSM. Inclusion criteria included participants’ age ≥18 with cancer who received chemotherapy, targeted therapy, or immunotherapy at the University of Alabama at Birmingham. Race and ethnicity (Black or African American, White, Asian, other and unknown), sex, cancer type (breast, gastrointestinal [GI], genitourinary [GU], gynecological [GYNX], head and neck, leukemia, lymphoma, melanoma, myeloma and other), urban/rural residence, Area Deprivation Index (ADI), and insurance type (Medicaid, Medicare, none, other and private) were abstracted from electronic medical records (EMR) and PRO platform (Carevive). Descriptive statistics were calculated using frequencies and percentages for categorical variables and medians and interquartile ranges for continuous variables. Differences in enrollment status characteristics were calculated using measures of effect size such as Cramer’s V. Results: Of the 307 patients, two thirds of patients were female (71%); 25% were Black or African American and 66% were White patients; 15% lived in an area of higher disadvantage. For insurance, 46%, 26%, 10%, 8%, and 9% of patients had Private, Medicare, Medicaid, other insurance, and no insurance, respectively. The proportion of patients who declined enrollment was higher for males than females (22% vs. 10%), Black or African American than White (18% vs 13%); and having Medicare than private insurance (22% vs. 10%). Compared to those who enrolled, patients who declined enrollment were more often to be male (V:0.2), Black or African American (V:0.1); and have Medicare insurance (V:0.2). Patients enrolled vs. declined in RSM had similar ADI scores (V:0.01). Conclusions: This study demonstrates that potentially vulnerable patients, including Black patients and those with public insurance, have lower RSM engagement. Future analysis is needed to understand participation barriers and how to better engage diverse populations to ensure optimal healthcare delivery to all patients.

Evaluating nurses' time to response by severity and cancer stage in a remote symptom monitoring program

341

Background: Remote symptom monitoring (RSM) using electronic patient reported outcomes (ePROs) allow for patients with cancer to communicate symptoms to their clinical team between clinic visits. Prior randomized control trials of RSM focused on advanced cancer, and less data are available for patient with early stage cancers. The University of Alabama at Birmingham (UAB) implemented RSM for early stage (I-III) and advanced stage (IV) patients on active treatment. This study evaluates nurses’ real-world response time to alerts by varying severity and by patients cancer stages. Methods: This study included women with stage I-IV breast cancer who received care at UAB from October 2020 through May 2022. The program was first implemented in the breast clinic allowing for larger patient numbers with early and advanced stage breast cancer. A composite score for symptom severity is automatically calculated in the Carevive® platform for moderate, severe, or worsening symptoms using patient responses for frequency, severity, and interference. The nurse receives an alert if a symptom is moderate or severe. Surveys with at least one severe alert were categorized as severe and response time was categorized as optimal if the survey was closed within 48 hours (goal time for phone message follow-up). Odds ratios (OR), predicted probabilities, and 95% confidence intervals (CI) were estimated using a patient nested logistic regression evaluating time to response comparing surveys with at least one severe alert notification to those with no severe, adjusting for age at enrollment, race, cancer stage, provider who closed the surveys, and quarter from study start and date. An interaction between severity and cancer stage was evaluated. Results: Of 137 patients included in this study, 64% were White; 86% were diagnosed with early-stage breast cancer. The median age at diagnosis was 54 (27-79). Of 802 surveys included, 38% reported at least one severe symptom and 70% had an optimal response time. Similar results were seen when stratified by early vs. advanced stage with 39% and 38% reporting at least one severe alert and 68% and 71% an optimal response time, respectively. In our adjusted analysis, when compared with surveys that had no severe alerts, surveys with at least one severe alert had similar odds of having an optimal response time (OR, 1.29; 95%CI, 0.88, 1.89). No significant interaction between severity and stage was observed on the odds of optimal response time. Conclusions: Response times to alerts were similar regardless of the severity of the alert and cancer stage, suggesting alert management is incorporated into routine workflows and not prioritized based on disease or alert severity. Additional research is needed to understand factors contributing to non-optimal response times.

What’s missing? Diagnostic workup for breast cancer in Sudan

110

Background: In Sudan, healthcare access may limit diagnostic workup for breast cancer. This study evaluates the relationship between geography and ethnicity and completeness of documentation of diagnostic workup (staging and receptor testing) in Sudan. Methods: This retrospective study used data abstracted from patients with breast cancer receiving cancer care at Sudan’s largest cancer center (Radiation and Isotopes Center Khartoum [RICK]) in 2017. The patient’s age at diagnosis, sex, breast cancer stage, ethnic subgroup (further categorized as Arab and non-Arab), regions of origin and residence (Central, Northeastern, Western, and Khartoum [where RICK is located]), and receptor status from pathology reports were abstracted from paper medical records. Complete diagnostic workup was defined as having both receptor testing and staging. Descriptive statistics were calculated using frequencies and percentages for categorical variables and median and interquartile range (IQR) for continuous variables. Odds ratios (OR) and 95% confidence intervals (CI) were estimated to evaluate complete diagnostic workup on ethnic group, origin, and residence using binomial logistic regression models (excluding non-Sudanese patients and those with missing demographics). Results: Of 240 patients included, 237 were female, median age was 53 (IQR 43-62). Most often patients were Arab (68%), originated from Northeastern and Khartoum regions (both 28%) and lived in the Khartoum region (53%). Overall, 49% patients were missing receptor testing and/or staging, with modest differences by geographic region and ethnicity (Table). In adjusted analyses, non-Arab patients had similar odds of having complete diagnostic workup when compared to Arab patients (OR 1.22; 95% CI 0.70-2.10). Patients originating from and residing in regions outside the Khartoum region had similar odds of complete diagnostic workup when compared to patients originating from and residing in the Khartoum region. Conclusions: Almost half of breast cancer patients had incomplete diagnostic workup, regardless of region of origin, region of residence, and ethnic group. This highlights a substantial systems-based quality gap in care delivery, warranting efforts to improve completeness in diagnostic workup for all patients with breast cancer in Sudan.[Table: see text]

Academic and community physician perspectives on breast cancer biomarker use in clinical trials

349

Background: Biomarkers are regularly utilized to select treatment within cancer clinical trials. However, there remains a lack of understanding regarding physician perspectives on what data is needed for physicians to comfortably use these markers to escalate or de-escalate chemotherapy. Methods: Semi-structured qualitative interviews with medical oncologists from different academic and community-based cancer centers were conducted to investigate perspectives on the utilization of biomarkers to de-escalate chemotherapy. Key topics explored included: (1) physician preference for biology-based (e.g. genomic profiles) vs. response-based (e.g. complete pathologic response) biomarkers, and (2) importance of personal familiarity with biomarkers. Interviews were audio-recorded and transcribed. Two independent coders analyzed transcripts using a constant comparative method in NVivo to identify major themes. Analysis was stratified by practice-type to elucidate differences between oncologists at academic and community practices. Results: Of the 39 participating physicians, 51% practiced in an academic setting and 49% practiced in a community setting. The majority of physicians (67% overall, 77% community, 59% academic) did not have a preference for biology-based vs. response-based biomarkers, if the data is equally strong and clinical use is appropriate for the clinical context (e.g. patient subtype). Many physicians were reassured by achieving a real-time therapeutic response, with 23% of physicians preferring response-based biomarkers. One physician stated, “I am still more comfortable with a real-time, well-validated biomarker, response marker, than I am with an overall predictive marker for a population”. In contrast, 10% (all academic) preferred biology-based biomarkers. One physician commented “I think the biology is probably more attractive because that potentially allows you to avoid treatment, whereas pathCR they've already had to get treatment to get there”. The majority of academic physicians (55%) felt that strong data was more important than personal familiarity with regards to implementation of novel biomarkers, as noted by one who stated, “As long as there's good data, I don't care.” 15% of community physicians shared a similar view. The majority of community physicians (54%) voiced familiarity to be more important in their comfort with biomarker use as noted by one physician who stated, “I think things I’m already familiar with, I'm more inclined to feel good about”. 18% of academic physicians held a similar perspective. Conclusions: Academic and community physicians’ perspectives regarding use of novel biomarkers overlap, with multiple factors playing a role in how these biomarkers are used in decision-making. Future research is needed to understand the impact of biomarker selection on clinical trial enrollment.

Trajectory of symptoms reported in remote symptom monitoring over the course of oncology treatment for gynecologic cancers

270

Background: Patients now have the ability to utilize electronic patient reported outcomes (ePROs) for remote symptom monitoring (RSM). This analysis seeks to better understand trajectory of reported symptoms during treatment for patients with gynecologic cancer participating in RSM. Methods: We approached patients with gynecological cancer initiating treatment at the Mitchell Cancer Institute (MCI) between 7/1/21-4/30/2022. Patients were eligible if they were starting chemotherapy, targeted therapy, or immunotherapy for a new cancer. Patients seeking a second opinion were excluded. Enrolled patients received symptom survey (PRO-CTCAE questions) via text or email once per week. Initially, only severe alerts were forwarded to the clinical care team; moderate alerts were forwarded to clinical teams once they were comfortable with alert management. Patients completed symptom assessments for 24 weeks or until withdrawal. Patient age at enrollment, race, sex, cancer type, cancer stage, and PROs were abstracted from electronic health records and the PRO platform (Carevive). Descriptive statistics were calculated using frequencies and percentages for categorical variables and median and interquartile ranges (IQR) for continuous variables. Results: A total of 60 female patients with gynecological cancer were enrolled; 33% were Black or African American and 67% were White; median age was 61 years (IQR 53-68). Seventy-eight percent (47/60) of patients reported 379 symptoms with at least one moderate or severe alert during this time period; 32% considered moderate and 68% considered severe. Overall, the most frequently reported symptom was pain (29%). At baseline (week 0), 14% and 41% of 56 patients reported moderate symptoms and severe symptoms, respectively. Symptom burden decreased over time with 4% and 7% of 27 patients who completed a survey at 12 weeks reporting moderate and severe symptoms. Specific symptom trajectories followed similar patterns. Conclusions: In our sample, patients reported the majority of symptoms during the first three months of treatment. Symptom trajectory decreased with time, suggesting symptoms are being effectively monitored and addressed by the clinical teams engaging in RSM. Future research is needed to understand if symptom improvement translates to increased quality of life, decreased hospitalizations, and increased survival for patients, as well as lessen the burden of call volume on the clinical team.

Patients' perspectives on late diagnosis of breast cancer in northern Tanzania: The role of traditional healers

116

Background: In Tanzania majority of women are diagnosed with advanced stage of breast cancer. Factors influencing delay in diagnosis resulting in advance stage have not being investigated in the region, particularly as it relates to rural and urban patients. This study aims to explore the experience of breast cancer diagnosis amongst rural and urban patients. Methods: Women diagnosed and confirmed with breast cancer in outpatients setting in a Cancer Care Centre were identified by clinic nurse and introduced to the study. Semi-structured interviews were conducted and transcribed verbatim. Thematic coding using a grounded theory approach was done by two independent researchers using NVivo 12 Mac. Results: Twenty patients (10 rural and 10 urban) participated in the interviews. The average age was 56; 5 (25%) were married, 11 (55%) had primary education, and 10 (50%) were not employed. The majority (70%) had stage IV breast cancer, 15% had stage III and 15% had stage II breast cancer. Seventeen respondents (85%) sought care from traditional healers prior to diagnosis and treatment at the cancer center. Women largely described this pattern of care due to family or community recommendations and pressures to first seek care with traditional healers, as noted by one woman “... neighbours who took me to the traditional healer they told me that, it is the same healer who treated the man who cured from cancer. During my visit to the healer the man who get cured used his medication also I use to see him attending to this traditional healer.’’. All the participants regretted this decision at time of interview due to ineffective and costly treatment which ultimately delayed their hospital presentation and ability to receive quality treatment. One women stated, “...t he medicine cost me one thousand and fifty thousand Tanzanian shillings [75$]. The traditional healer initially want a patient to pay one hundred thousand Tanzanian shillings [50$] then the rest of the money to be paid once a patient complete the dose. I paid only hundred thousand but when used and found there are no any good progress didn’t continue to take it...’’ Rural patients emphasized. ‘‘... no they cannot cure cancer. They just waste people’s time there, while the disease is growing. I am saying this because when I call the traditional healer and tell him that, I am feeling sicker, instead of telling me to go and see him for the change of treatment. Instead he tells me go to hospital it means they cannot cure cancer. They took our money, waste our time and when disease goes bad they tell you to go to the hospital. I don’t belief on them at all.’’. Conclusions: Traditional healers are a critical part of the cancer delivery system in Northern Tanzania yet may contribute to delays in cancer care. Culturally sensitive interventions targeting these providers are necessary to promote early detection, decrease delay in presentation, and improve timely access to care.

Peeling back the curtain: The impact of patient and provider race on clinical trial enrollment

99

Background: Optimization clinical trials testing less intense treatments are becoming more prominent in oncology due to the availability of better prognostic tools and targeted therapies. In addition to previously documented barriers, these trials are likely to face new barriers from engaging racially diverse populations due to the potential of an increased recurrence risk with reducing treatment. However, little is known about the role of race in decision-making for optimization clinical trials amongst physicians and patients. Methods: This qualitative study included a subset analysis on the influence of race in decision-making for participation in trials testing less chemotherapy. This analysis is part of a larger study, which included semi-structured interviews with patients, patient advocates, and physicians assessing barriers and facilitators to trial participation. Interviews were transcribed, and four coders evaluated transcripts for key themes and exemplary quotes using NVivo. Results: 79 participants (24 patients with breast cancer, 16 patient advocates, and 39 physicians) participated; 30% of patients and patient advocates and 26% of physicians were BIPOC (Black, Indigenous, and People of Color). Several key barriers traditionally associated with Black race were noted amongst both patients and physicians, including aggressive biology (e.g. triple negative breast cancer), younger age, socioeconomic challenges, and lack of trust in physicians and clinical trials. One physician noted, “Taking someone who already has a mistrust of medical care and talking to them about a trial of cutting medical care back, it’s challenging.” While some physicians explicitly acknowledged the role of race in decision-making, often linking race to these barriers, the majority of physicians independently highlighted these barriers while denying the explicit impact of race. Black patients noted similar barriers including emphasizing the role of having triple negative breast cancer, being young, the influence of financial strain, and medical mistrust. One Black patient commented, “I was a triple negative, and that kind was more prone to African American women, usually we don’t really survive from it as well as other races do.” Another Black woman commented, “I had a lot of family and friends that were worried that I was going to be a “guinea pig”. In contrast, White patients heavily emphasized the role of trust in their physicians when making decisions. A White woman stated the following, “I would have done whatever they (doctors) told me was the best thing to do.”. Conclusions: Factors associated with Black race can play both an overt and subconscious role in patient and provider decision-making about participation in optimization clinical trials. Multi-level interventions are needed to address these specific barriers to ensure representative participation in clinical trials for all patient populations.

Adaptation of remote symptom monitoring using electronic patient-reported outcomes for implementation in real-world settings

272

Background: Despite evidence of clinical benefits, widespread implementation of remote symptom monitoring has been limited. We describe a process of adapting a remote symptom monitoring intervention developed in a research setting to a real-world clinical setting at two cancer centers. Methods: This formative evaluation assessed core components and adaptations to improve acceptability and fit of remote symptom monitoring using Stirman’s Framework for Modifications and Adaptations. Implementation outcomes were evaluated in pilot studies at the two cancer centers testing technology (Phase I) and workflow (Phase II and III) using electronic health data; qualitative evaluation with semi-structured interviews of clinical team members; and capture of field notes from clinical teams and administrators regarding barriers and recommended adaptations for future implementation. Results: Core components of remote symptom monitoring included electronic delivery of surveys with actionable symptoms, patient education on the intervention, a system to monitor survey compliance in real-time, the capacity to generate alerts, training nurses to manage alerts, and identification of personnel responsible for managing symptoms. In the pilot studies, while most patients completed > 50% of expected surveys, adaptations were identified to address barriers related to workflow challenges, patient and clinician access to technology, digital health literacy, survey fatigue, alert fatigue, and data visibility. Conclusions: Using an implementation science approach, we facilitated adaptation of remote symptom monitoring interventions from the research setting to clinical practice and identified key areas to promote effective uptake and sustainability.

Nursing strategies to improve alert closure for remote symptom monitoring

421

Background: For successful remote symptom monitoring using patient-reported outcomes, nurses should respond to alerts in a timely fashion. Where clinical trials utilized research staff for alert management, the shift to standard-of-care delivery necessitates that this responsibility be added as a task to an already strained nursing workforce. Little is known about strategies to engage nurses to improve timeliness of alert management. Methods: In this quality improvement initiative, we aimed to improve timeliness of alert closures generated by moderate or severe symptoms within a remote symptom monitoring program. Optimal closure was defined as < 48 hours, which was consistent with institutional requirements for response to patient phone calls. A continuous quality improvement approach, with multiple Plan Do Study Act (PDSA) cycles was conducted. Data was captured from the electronic medical record and PRO platform (Carevive). Descriptive statistics included frequencies and percentages. The proportion of alerts closed each month < 48 hours, 48-72 hours, 3-7 days, and > 7 days were reported overall and by disease team (i.e., major cancer types). Surveys not closed were considered > 7 days. The timing of strategies to improve nursing engagement were documented and evaluated for impact on alert closure. Results: From June 1, 2021-May 31, 2022, 1121 moderate or severe alerts were generated from 234 patients. Disease teams had variable remote symptom monitoring start dates: breast, leukemia, and limited gynecologic (prior to 6/2021); myeloma and gastrointestinal (7/2021); genitourinary (10/2021); head and neck (12/2021); melanoma (2/2022); and Lymphoma (4/2022). In 6/2021, the overall alert closure at < 48 hours, 48-72 hours, 3-7 days, and > 7 days was 57%, 4%, 14%, and 25% respectively (n = 28). To improve alert closures, several key strategies were deployed to improve alert closure times including disease-specific reporting and meetings with nursing leadership (10/2021); identification of a nurse champion, creation of “cheat sheets” to remind nurses how to close alerts, and individualized calls with nurses with open alerts (1/2022), and inclusions of requirement to close alerts in nursing newsletters (2/2022). Overall, alert closure less than 48 hours improved to 61% by 12/2021 (n = 97) and to 69% by 5/2022 (n = 167). Disease group alert closure varied, with higher closure more commonly in teams with greater duration of use, such as breast cancer team with an alert closure of 85% < 48 hours in May 2022. Conclusions: Key nursing engagement strategies improve alert closure for remote symptom monitoring programs implemented in real-world settings.

The effect of modifications to clinical trial activities implemented during the COVID-19 pandemic on willingness to participate in clinical trials: The TBCRC 057 survey

97

Background: In order to maintain safety, clinical trial activities have been modified during the COVID-19 pandemic. As part of the TBCRC 057 survey, we assessed how pandemic-related modifications to trial activities affect breast cancer patients’ willingness to participate in clinical trials. Methods: US residents with breast cancer were eligible to complete the online survey 8/6/21 – 9/30/21. Respondents rated whether each of 11 modifications to clinical trial activities would affect their decision to participate in a trial during or after the pandemic. Items evaluated modifications that involved changing the location of trial activities to closer to home, switching trial activities to telemedicine and making the trial schedule more flexible and convenient. Response options were “much less likely to participate”, “somewhat less likely to participate”, “would not affect my decision whether or not to participate”, “somewhat more likely to participate” and “much more likely to participate”. Current trial participants were asked to consider how modifications would affect their decision to participate in another trial. Results are reported descriptively. Results: Among 385 respondents, median age was 52 (range 25-85), 88.6% were non-Hispanic White, 52.5% had metastatic disease, 93% were receiving active treatment, 48.6% received care at an academic center and 9.6% were current trial participants. Changing location of trial activities was viewed favorably, with 70.2%, 64.6% and 54.1% of respondents indicating they would be much or somewhat more likely to participate if they could complete trial blood tests, x-ray tests or doctor visits closer to home, respectively. Similarly, the option to complete trial activities electronically was viewed favorably, with 59.6%, 58.6% and 60.9% of respondents indicating they would be much or somewhat more likely to participate if they could complete trial doctor visits, consent and questionnaires via telemedicine, respectively. With regard to modifications to make the trial schedule more flexible and convenient, respondent feedback was also favorable. 71.4%, 67.7% and 82.4% of respondents indicated that requiring study site visits no more than once per 3 weeks, widening windows for trial activities and offering home delivery of oral study medications, respectively, would make them much or somewhat more likely to participate. Finally, 30.4% and 51.7% indicated that the flexibility to opt-out of research-only blood tests and biopsies, respectively, would make them much or somewhat more likely to participate. Conclusions: Patients view modifications to trial activities implemented during the pandemic favorably. Trials should be flexible and the option to conduct study activities close to home or electronically when possible should be maintained during the pandemic and beyond.

TBCRC 057: An online survey about anxiety and willingness to participate in breast cancer clinical trials during the COVID-19 pandemic

1564

Background: Enrollment in clinical trials has declined during the COVID-19 pandemic. Simultaneously, breast cancer patients have reported heightened anxiety. We assessed whether breast cancer patients’ anxiety about the pandemic affects their willingness to participate in trials. Methods: English or Spanish-speaking US residents with breast cancer were eligible to complete the online REDCap survey 8/6/21 – 9/30/21. Respondents rated their anxiety about the pandemic on an 11-point scale from 0 (no anxiety) to 10 (worst anxiety possible). Anxiety scores were categorized as no/mild (0-3), moderate (4-6) or severe (7-10). Knowledge about trials was assessed with 11 true/false items and attitudes toward trials with the Attitudes Toward Cancer Trials Scales - Cancer Treatment Subscale (ATCTS-CTS). Respondents rated their willingness to participate in a breast cancer clinical trial before and during the pandemic on 5-point scales from 0 (not at all willing) to 4 (definitely willing). Trial participants were considered “definitely willing.” Change in willingness to participate in trials during the pandemic compared to prior was defined as a binary outcome, "less willing" vs "no less willing." Means were compared via t-test and mean difference was tested via paired t-test. Multivariable logistic regression was used to model the association of anxiety and other factors with being less willing to participate in trials during compared to prior to the pandemic. Results: Among 385 respondents, median age was 52 (range 25-85), 271 (70%) were non-Hispanic White and 202 (53%) had metastatic disease. 154 (40%) received care at academic centers and 37 (10%) were current trial participants. Most rated their anxiety as moderate (43%) or severe (38%). Mean willingness to participate in a trial was lower during compared to prior to the pandemic (2.97 vs 3.10; p < 0.0001). Fifty (13%) respondents were less willing to participate in a trial during the pandemic compared to prior. After controlling for covariates, those with severe anxiety had 5.07 times odds of being less willing to participate during the pandemic compared to prior than those with no/mild anxiety (p = 0.01). For every 1-point increase in ATCTS-CTS score (indicating better attitude toward trials) there was a 3% decrease in the odds of being less willing to participate during the pandemic (p = 0.006). For every 1-point increase in the clinical trials knowledge score (indicating more knowledge) there was a 15% decrease in the odds of being less willing to participate during the pandemic (p = 0.02). Conclusions: Pandemic-related anxiety is common in breast cancer patients and is associated with being less willing to participate in trials during the pandemic compared to prior. Education about trials, including safety modifications implemented during the pandemic, may mitigate anxiety and improve willingness to participate.

Telemedicine adoption and utilization among financially distressed patients with cancer during the COVID-19 pandemic: Insights from a longitudinal nationwide survey

1596

Background: Telemedicine use during the COVID-19 pandemic among financially distressed patients with cancer, with respect to the determinants of adoption and patterns of utilization, has yet to be delineated. We sought to systematically characterize telemedicine utilization in financially distressed patients with cancer during the COVID-19 pandemic. Methods: We conducted an analysis of survey data assessing the use of telemedicine in patients with cancer during the COVID-19 pandemic collected by Patient Advocate Foundation (PAF) from May 2020 to December 2020. Primary study outcome was telemedicine utilization rate. Secondary outcomes were independent predictors of telemedicine utilization patterns, volume, and utilization preferences. Multivariate and poisson regression analyses were used to identify predictive factors. Results: Of the 1,390 respondents, 627 completed two survey waves and were included in this study. Telemedicine adoption during the pandemic was reported by 67% of patients, with most (63%) preferring video visits. Younger age (odds ratio, 6.07; 95% CI, 1.47-25.1), and higher comorbidities (odds ratio, 1.79; 95% CI, 1.13-2.65) were independent predictors associated with telemedicine adoption. Younger age (19-35 yrs.) (incidence rate ratios [IRR], 1.78; 95%CI, 24-115%) and higher comorbidities (≥3) (IRR; 1.36; 95%CI, 20-55%) were independent predictors associated with higher utilization volume. As area deprivation index increased by 10 units, the number of visits decreased by 3% (IRR 1.03, 95%CI, 1.03-1.05). Conclusions: The rapid adoption of telemedicine may exacerbate existing inequities, particularly among vulnerable financially under-resourced patients with cancer. Policy-level interventions are needed for the equitable and efficient provision of this service.

Physician perspectives on extrapolating data from trials testing less-intense treatment to underrepresented populations

521

Background: Clinical trials provide the foundation for evidence-based practices, yet trial participants are often not representative of all patients. Historically, clinical trials involved adding novel agents to standard of care to improve survival. There has been a shift to an individualized approach with testing less intense treatment, yet vulnerable patient groups are at risk for underrepresentation. Little is known about physician perspectives on implementing less intense treatment approaches for patients who are not represented in sufficient number to draw conclusions on subpopulations. Methods: Open-ended, individual qualitative interviews with medical oncologists from different cancer centers exploring their perspectives on trials that test less intense treatment for patients with cancer. Interviews were audio-recorded and transcribed. Four independent coders utilized a content analysis approach to analyze transcripts using NVivo. Major themes and exemplary quotes were extracted. Results: Of the 39 participating physicians, 61.5% felt comfortable extrapolating, 30.8% were hesitant, and 7.7% would not feel comfortable extrapolating trial outcomes to underrepresented populations. One physician noted, “We've been extrapolating for as long as I can remember and certainly that I've been in practice; so we do need to do better there, but extrapolation is only natural with what we have.” Facilitators of comfort included sentiment that “biology is biology”, such that the cancer characteristics were what mattered; the strength of the evidence from the trial overall; inclusion of subset analysis on underrepresented populations; and prior experience making decisions with limited data. Barriers to extrapolation included the potential harm over the patient’s lifetime; concerns about groups that had minimal participants; application specifically to younger patients; and extending findings to racially and ethnically diverse populations. Oncologists highlighted the need for shared decision-making when applying study results to underrepresented populations. They also expressed concerns about study findings being applied to patients who would have been ineligible in the original trials. Universally, broader inclusion in trials testing lowering chemotherapy is desired. Conclusions: The majority (92%) of physicians report that they would extrapolate clinical trial results to patients poorly represented in de-escalation trials, while expressing concerns about applicability to specific subpopulations based on tumor characteristics (e.g. stage, biology) and patient demographics (e.g. age, race). Further work is needed to increase clinical trial representation of diverse populations to safely and effectively optimize treatment for patients with cancer.

Fear of COVID-19: Effects on mental health in under-resourced patients with cancer

152

Background: Given the high risk of COVID-19 mortality, patients with cancer are not only vulnerable to physical consequences of COVID-19 infection, but also to adverse psychological outcomes, including fear of COVID-19. Without intervention, psychological distress in patients with cancer can lead to worsening symptoms, poor quality of life, and lower survival. We sought to evaluate the association between fear of COVID-19 and psychological distress for under-resourced patients with cancer during the pandemic. Methods: This observational, longitudinal survey study, fielded during early (May 20- July 11, 2020) and later pandemic (December 2-December 23, 2020), evaluated the pandemic’s impact on patients with cancer receiving Patient Advocate Foundation (PAF) services from July 2019 – April 2020. Questions focused on individual experiences with COVID-19 and psychological, emotional, physical, and material effects from the pandemic. The validated Fear of COVID-19 Scale was used to determine fear of COVID-19. Psychological distress was determined using a four-item questionnaire by Holingue et al. (score range 4 – 16). Means and 95% confidence intervals (CI) were estimated using generalized estimating equation modeling with repeated measures to assess the effect of fear of COVID-19 on psychological distress early and later in the pandemic. Models adjusted for age, sex, race/ethnicity, region, annual household income, household size, marital status, employment status, Area Deprivation Index category, Rural-Urban Commuting Code category, cases per 100,000 in county of residence, cancer type, and number of comorbidities. Results: Amongst 1199 survey respondents, 94% considered themselves high risk for COVID-19. 448 respondents completed both the first and second survey. The majority of respondents were female (72%) and age 56-75 (55%); 40% were Black, Indigenous, or Persons of Color. In adjusted models of respondents who completed the early pandemic survey, respondents with more fear of COVID-19 had a higher mean psychological distress score (10.21; 95% CI 9.38-11.03) compared to respondents with less fear (7.55; 95% CI 6.75-8.36). Among those who completed the later pandemic survey, median fear of COVID-19 decreased (20 vs 19)median distress scores remained the same (8); respondents with more fear of COVID-19 had a higher mean psychological distress score (9.98; 95%CI 9.04-10.92) compared to respondents with less fear (7.87; 95%CI 6.98-8.76). Conclusions: Fear of COVID-19 was linked to psychological distress and persisted throughout the pandemic among under-resourced patients with cancer. Timely psychosocial support is critical to meet increased care needs experienced by patients with cancer during the COVID-19 pandemic. Given these results, fear of COVID-19 could be considered as a trigger to integrate psychological interventions in patients with cancer to treat psychological distress.

Association of fear of COVID-19 with delays in care or treatment interruptions in patients with cancer

98

Background: Patients with cancer are at risk for severe COVID-19 and may be vulnerable to health care delays. Delays or interruptions in care may lead to adverse cancer outcomes. Little is known about the relationship between fear of COVID-19 and disruptions in cancer care delivery. Methods: This longitudinal survey was distributed to individuals with cancer who received services July 2019-April 2020 from Patient Advocate Foundation, a non-profit organization that provides case management and financial aid to patients with chronic illness. Data was collected twice - early pandemic (5/20/20-7/11/20) and later pandemic (12/3/20-12/23/20). Fear of COVID-19 was assessed with the Fear of COVID-19 Scale and dichotomized as more (≥22) vs less (< 22) fearful. Respondents reported delays in care or treatment interruptions due to the pandemic and reasons for delays or interruptions. Respondents rated concern about potential long-term health issues due to delays on a 5-point Likert-like scale. We estimated predicted percentages and 95% confidence intervals (CI) using logistic regression models to assess the association of fear of COVID-19 (more vs less fearful) with delay in care or treatment interruption (any vs none) at each time point. We adjusted models for age, sex, race/ethnicity, region, annual household income, marital status, employment status, household size, Area Deprivation Index category, Rural-Urban Commuting Code category, county-level COVID-cases per 100,000, cancer type and number of comorbidities. Results: Amongst the 1,199 early pandemic survey respondents, the majority were female (72%), had household income < $48,000 (73%), and had ≥1 comorbidity (60%). 448 of the early pandemic survey respondents also completed the later survey. 464 (39%) and 166 (37%) respondents were categorized as more fearful at the early and later time points respectively. 567 (47%) and 191 (43%) reported delays or interruptions at the early and later time points respectively. The most common reported reasons for delays or interruptions were hospital/provider restrictions (early: 27%, later: 19%) and patient choice (early: 13%, later: 15%). Among respondents with delays or interruptions at each time point, > 70% were at least moderately concerned about potential long-term health issues due to delays. In adjusted models, more fearful respondents had higher predicted percentages of delayed care or treatment interruptions compared to less fearful respondents early (more fearful: 56%, 95% CI 39%-72%; less fearful: 44%; 95% CI 28%-61%) and later (more fearful: 55%, 95% CI 35%-73%; less fearful: 38%; 95% CI 22%-57%) in the pandemic. Conclusions: Fear of COVID-19 is common among patients with cancer and is linked with delays in care and treatment interruptions. System-wide strategies are needed to address fear of COVID-19 and to ensure equitable, timely, and safe access to cancer care throughout the pandemic.

Patient-reported unfair treatment within the health care system

115

Background: Patients with cancer who perceive discrimination and unfair treatment from the health care system are at risk for lower health-related quality of life. This study seeks to better understand the characteristics of under-resourced patients who report unfair treatment from the health care system and providers. Methods: This cross-sectional analysis used data from a nationwide survey distributed in December 2020 by Patient Advocate Foundation (PAF), a US non-profit organization providing case management and financial aid to individuals with chronic illness. The survey was fielded via email to those who received PAF services from July 2019-April 2020. Inclusion criteria included a valid e-mail address, aged ≥ 19, and either current or previous cancer treatment. Respondents reported unfair treatment in connection to their health care. Age, sex, race/ethnicity, and annual household income were abstracted from the PAF database. The validated Group-Based Medical Mistrust Scale was used to assess respondents’ level of mistrust in medical providers as it relates to their ethnic group. Scores range from 12-60 and were categorized based on tertiles as high mistrust (scores ≥ 29), neutral (21-28), and low mistrust (≤ 20). Frequencies and percentages were calculated for categorical variables. Results: There were a total of 429 survey respondents with cancer. Most respondents were female (73%) and aged 56-75 (57%); 31% were Black, Indigenous, or Persons of Color (BIPOC). The most common cancer types were hematologic (33%) and breast (33%). Overall, 20% (n = 86) of respondents reported having received unfair treatment. Of those reporting unfair treatment, 56% reported receipt from their doctor, nurse, or health care provider, 51% insurance company, 38% the health care system, and 14% pharmacist. When asked why they felt unfairly treated, the most common responses were related to insurance status (51%), disease or condition (45%), and income (35%). Notably, unfair treatment due to race/ethnicity (6%), sex (9%), and sexual orientation/gender expression (3%) were uncommon. When compared to those who reported objective treatment, respondents reporting unfair treatment were more often unemployed/other (28% vs 11%), privately insured (38% vs 27%), having income < $23,000 (40% vs 25%), having 3+ comorbidities (40% vs 23%) and reporting more mistrust in medical providers (53% vs 27%). There did not appear to be a difference in reporting of unfair treatment by race/ethnicity. Of BIPOC respondents, 51% reported high mistrust in medical providers. Conclusions: This under-resourced population of respondents with cancer reported unfair treatment related to their finances, insurance, and disease status. Our data suggest health care-associated discrimination may occur based on socioeconomic resources. This work identifies a novel equity consideration warranting further evaluation.

Survival in the real world: A national analysis of patients treated for early-stage breast cancer

75

Background: Many patient population groups are not proportionally represented in clinical trials, including patients of color, at age extremes, or with comorbidities. It is unclear how treatment outcomes may differ for these patients compared to those well represented in trials. Methods: This retrospective cohort study included women diagnosed with early-stage (I-III) breast cancer (EBC) between 2005-2015 in the CancerLinQ Discovery electronic medical record-based dataset. Patients with comorbidities or concurrent cancer were considered unrepresented in clinical trials. Non-White patients and/or those aged <45 or ≥70 years were considered underrepresented. Patients who were White and aged 45-69 were considered well represented. Overall and EBC subtype-stratified Cox proportional hazards models estimated hazard ratios (HR) and 95% confidence intervals (CI) for five-year mortality by representation group. The overall model was adjusted for cancer stage, subtype, chemotherapy intensity, and year of EBC diagnosis. Stratified models were adjusted for cancer stage, individual treatment regimen (due to lack of chemotherapy intensity variation within subtype), and year of EBC diagnosis. Results: Of 11,770 patients, most were aged 45-69 (71%), White (72%), diagnosed with stage II (51%), or HR+HER2- EBC (56%). Unrepresented patients (7%) were categorized due to comorbidities (76%), concurrent cancer (22%), or both (2%). Underrepresented patients (45%) were categorized based on age (44%), race/ethnicity (39%), or both (17%). The remaining patients were well represented in trials (48%). In adjusted models, unrepresented patients had almost three times the hazard of death than well-represented patients (HR 2.71, 95% CI 2.08-3.52; Table). The hazard of death for underrepresented versus well-represented patients was similar (HR 1.19, 95% CI 0.98-1.45). Comparable results were seen in EBC subtype-specific models. Conclusions: Over half of patients in this study would be considered underrepresented or unrepresented in clinical trials due to age, comorbidity, or race/ethnicity. Patients considered unrepresented in trials experienced poorer survival compared to those well-represented. Trialists should ensure study participants reflect the real-world disease population to support evidence-based decision making for all individuals with cancer.[Table: see text]

Effect of limited reporting in clinical trials on the ability to guide treatment decisions for real-world patients with cancer

18

Background: National Comprehensive Cancer Network (NCCN) guidelines are derived from reviews of clinical trials. Many of these trials have strict inclusion criteria, resulting in trial samples that are not representative of the larger cancer population. We therefore sought to understand how clinical trials, referenced in the NCCN guidelines, report key patient baseline demographics relating to age, race/ethnicity, and country or geographic region. Methods: NCCN guidelines for four cancer types were reviewed: prostate, colon, breast, and lung. We abstracted race/ethnicity, age, country/geographic region, and hazard ratios (HRs) from references indexed in the NCCN Guidelines. Race/ethnicity and age information was obtained from baseline characteristics tables in reported studies. The country/geographic region from which participants were recruited was acquired from each individual trial’s National Clinical Trial (NCT) number, linked to clinicaltrial.gov or the main manuscript. Each study was also assessed for its reporting of survival outcomes based on race/ethnicity, age, and country. Results: A total of 31 studies reporting on 36 regimens were examined for this review. While all studies reported age, only 39% (n=12) included characterization of older adults (60 years or older). 52% provided information on the racial and ethnic makeup of the study sample. Countries where participants were recruited were mostly not reported in the main papers, rather they were identified from ClinicalTrials.gov. Also, while 67% of all studies (n=25) included an international sample, only 5% reported the country or geographic location in the main manuscript. Few studies reported efficacy by patient sub-population. 12 of the 31 (39%) manuscripts reported HRs by age. Of the 16 manuscripts reporting race/ethnicity, 16% included HRs by race/ethnicity. Only one study reported efficacy outcomes by country. Conclusions: There is a need to have a standardized system for reporting baseline characteristics as well as trial outcomes for clinical trials. Including information on subgroup-specific baseline and efficacy outcomes in clinical trial results is an inexpensive way of improving the quality of information available to oncologists and will aid them in making evidence-based treatment decisions for the entirety of their patient populations.

Response-guided treatment optimization in the I-SPY2 trial

222

Background: As new tools for prognostic prediction and assessment of treatment response in breast cancer are developed, there is growing interest in treatment optimization strategies to reduce toxic therapy or add additional therapy as needed. The I-SPY 2 trial utilizes a new predictive tool “PreRCB,” which includes mid-treatment MRI and core biopsy responses, to identify patients likely to have a pCR (pathologic complete response) who can de-implement, or forgo, standard-of-care anthracycline-based chemotherapy (AC) after treatment with a taxane + investigational agent. This study assesses initial de-implementation eligibility and selection within the I-SPY2 trial. Methods: All patients enrolled in I-SPY2 since September of 2020 (when the preRCB protocol was first introduced to the trial) who completed the first treatment regimen were evaluated. Patients were only considered for de-implementation of AC if the mid-treatment biopsy was routed to site clinical pathology for review for residual invasive disease. In these patients, preRCB was determined to be predictive of pCR if there was 1) no invasive cancer on the core biopsy and 2) MRI findings met subtype-specific functional tumor volume criteria. Patients meeting criteria for predicted pCR could be offered the option to go directly to surgery by their provider, foregoing pre-operative AC, with the option to receive AC post-operatively should the surgical specimen reveal residual disease. The following outcomes (n,%) were descriptively evaluated: patients participating in I-SPY2, sites with at least one patient evaluated for de-implementation using pre-RCB, patients eligible for de-implementation, patients with surgical RCB assessment, and patients who de-implemented chemotherapy. Results: From September 2020 – May 2021, 145 patients were eligible for analysis. 9/20 (45%) sites with eligible I-SPY patients had at least one patient considered for preRCB assessment. Only 46 patients (32%) had clinical reads on their biopsies, as required for preRCB evaluation. Of these, 48% (14/29) of patients met both MRI and pathology criteria to de-implement AC. However, only seven (50%) of these patients chose to proceed to surgery without preoperative AC. Of those who did not have a predicted pCR using preRBC criteria, 87.5% (28/32) received AC as recommended by the protocol and 12.5% (4/32) did not receive preoperative AC. Conclusions: These pilot results demonstrate slow uptake of this strategy by patients and clinicians. We are currently interviewing patients and physicians to identify and assess facilitators and barriers for patients and physicians to pursue response-guided treatment regimes. Understanding these motivations will create the opportunity to engage site, providers, and patients to engage in research optimizing chemotherapy treatment. Clinical trial information: NCT01042379.

Exclusion criteria of breast cancer research protocols: A descriptive analysis

82

Background: Clinical trials play an important role in advancing cancer treatments. Unfortunately, only about 3% of adults with cancer are enrolled in a clinical trial in the United States due to various barriers to enrollment. This includes restrictive eligibility criteria, which currently have no standard guidelines. The purpose of this study is to evaluate the variability of eligibility criteria. Methods: This descriptive analysis utilized all therapeutic breast protocols offered at the University of Alabama at Birmingham (UAB) between 2004-2020. Exclusion criteria (e.g., laboratory values and comorbidities) were extracted from protocols using OnCore, an online dataset used to manage clinical trials, and ClinicalTrials.gov. Laboratory values or vital signs analyzed included liver function tests, hematologic labs, Eastern Cooperative Oncology Group (ECOG) performance status, and hypertension. Comorbid conditions included congestive heart failure, cardiovascular disease, presence of central nervous system (CNS) metastases, and history of prior cancer. Comorbid conditions were further analyzed by amount of time protocols required participants to be from initial diagnosis or exacerbation-free. Results: There were a total of 102 eligible protocols. Substantial heterogeneity was observed in exclusion criteria across liver/hematologic laboratory values and demographic/comorbidity variables. Among liver laboratory values, most protocols included an upper limit of acceptable for bilirubin (78%): 9% used the institutional upper limit of normal (ULN), 2% used 1.2xULN, 3% used 1.25xULN, 56% used 1.5xULN, 6% used 2xULN, and 2% used 3xULN. Similar variability was observed in protocols that included alanine transaminase and aspartate transaminase. Among hematological labs, 82% of protocols defined a lower limit of acceptable absolute neutrophil count: 1% 500mcL, 11% used 1,000mcL, 4% used 1,200mcL, 1% used 1,250mcL, 64% used 1,500mcL, and 1% used 1,800mcL. Of the comorbid conditions, exclusion criteria varied for congestive heart failure (49%), an acute exacerbation of cardiovascular disease (80%), CNS metastases (59%) and a prior cancer (66%). While most protocols included cardiovascular disease, the allowable timeframe varied between protocols: 4% did not allow an acute exacerbation within the previous 3 months, 32% did not allow within the previous 6 months, 5% did not allow within the previous 12 months, and 38 % did not specify a time frame. Protocols including history of a prior cancer as a criterion similarly had varied definitions based on timeline. Conclusions: Substantial heterogeneity was observed among clinical trial protocols. While exclusion criteria are necessary for patient safety, there is lack of evidence for current parameters. Future research should focus on defining standardized eligibility criteria while allowing for deviation based on drug specificity.

Care at the end of life during the COVID-19 pandemic: A CancerLinQ Discovery (CLQD) analysis

310

Background: For patients with cancer approaching end-of-life (EOL) during the COVID-19 pandemic, the changes in service availability and transition to telehealth may impact care delivery. Little is known about how the COVID-19 pandemic impacted EOL encounters with the healthcare system, particularly for vulnerable patient populations. Methods: This retrospective cohort study included patients with cancer who died from 1/2019-9/2020 in the CLQD electronic health record-based dataset. Descriptive statistics were generated for demographic characteristics and for clinical encounters (telehealth/in person/no encounter) in the last 3 months of life (EOL) among patients deceased in the year. Results were stratified by age, race, ethnicity, and geographic region. Results: Among the 49,688 deceased patients, 27% were under 65, 29% were 65-74, and 44% were 75+. The majority were Non-Hispanic White, with 5% Hispanic, 12% Black. In 2020, patients who were age 75+, White, Non-Hispanic, and/or living in the Midwest or Northeast had lower rates of in person encounters and higher rates of no encounter at EOL than those <65, Black, Hispanic, and/or living in the South (Table). Telehealth use at EOL increased from 2019 to 2020, with highest use amongst those in the West and Hispanic patients (Table). Conclusions: During the pandemic, telehealth use was limited at EOL compared to the 14% reported use for all cancer patients (data not shown). Black and Hispanic patients had slightly higher provider encounters at EOL, which may be due to differences in intensity of EOL care or death capture within the database. Further research is needed to evaluate the quality of EOL care and to assess opportunities to leverage telehealth where patients are unable to access outpatient services.[Table: see text]

Patient responses to weekly electronic patient-reported outcomes

182

Background: Home-based symptom monitoring using patient reported outcomes has been shown to reduce symptom burden and hospitalizations, while improving quality of life and overall. The primary goal of this study was to evaluate the early feasibility of home-based symptom monitoring. Methods: This prospective single-arm pilot study evaluated a two-part education and technology intervention (home-based symptom monitoring) for clinicians treating Multiple Myeloma (MM) and Acute Myeloid Leukemia (AML) patients. Inclusion criteria included patients ≥ 60 who were anticipating a treatment decision. Participants completed a baseline survey and then enrolled into the PROmpt® symptom monitoring platform, which used text or email to prompt weekly symptom surveys. Patients receive an auto-generated self-management plan based on electronic patient reported outcomes (ePRO). If severe symptoms were reported (score of > 7/10), alerts were sent to clinical nurses. Outcomes for this study included proportion of patient approach who agreed to participate, patient completion rates (completion of weekly surveys), compliance rate (completion of total surveys), number of alerts generated, and type of alert. Feasibility was defined as a completion rate of > 70%. Results: Between September 1, 2020 –May 19, 2021; 114 patients were screened, 77 were approached, and 35 were enrolled (18) MM, (17) AML patients. Of non-participants, 11 were not seeking care at the institution, 10 patients were uninterested or did not have a smartphone, 9 patients were ineligible, and 9 were unsure and left with information about the study, 3 declined enrollment. The majority (80%) or participants were ages 60-74; 20% of patients were ages 75+. Over the 13-week period, AML patients completed 195/220 (compliance rate of 89%). The average completion rate was 92%. For MM, 192/233 surveys were completed (82% compliance rate). The average completion rate was 94%. For AML, 9 was the average number of completed surveys and the average number for MM was 8. Over 3 months, there were 294 moderate to severe alerts generated for AML and MM patients. For AML patients, there were 40 fatigue, 25 constipation, 21 pain, 17 decreased appetite, 11 insomnia, 11 rash, 6 anxiety, 7 dyspnea/cough, 7 diarrhea, 5 depression/sadness, 4 nausea/ vomiting, 4 mouth/ throat sores, 3 neuropathy, 3 fever, and 2 alerts for other symptoms. Within a 3 month time span for MM patients, there was 35 pain, 21 constipation, 18 fatigue, 11 rash, 10 neuropathy, 9 anxiety, 7 insomnia, 6 depression/sadness, 4 decreased appetite, 4 other symptom, and 1 nausea/vomiting alert. Conclusions: This study demonstrated early feasibility with over 80% of patient completing their surveys with a high compliance rate. Future analysis will include both final implementation outcomes as well as patient outcomes for all patients within the study.

System-wide depression screening implementation in ambulatory oncology care: Key strategies and early implementation outcomes

260

Background: Oncology Care Model (OCM) requires cancer programs to provide depression screening during each care episode to meet the quality measure benchmark of 85% screening rate. This quality improvement (QI) project aims to assess 1) key strategies to integrate system-wide depression screening into routine cancer care, and 2) early outcomes of depression screening implementation. Methods: A mixed-method study design was used to assess strategies to implement routine depression screening in a southeast comprehensive cancer center between July 2019 and December 2020. Two top high-volume hematological oncology clinics that covers sixty percent of ambulatory care participated in the depression screening implementation. Data were collected using 1) depression screening completion rate during OCM performance periods, 2) needs assessment to identify barriers and facilitators of implementation, and 3) semi-structured interviews to assess staff and provider feedback on sustainable implementation strategies. Data were analyzed using descriptive analysis for quantitative outcomes and thematic analysis for qualitative outcomes. Results: A total of 64 hematological oncology providers (n = 22) and staff (n = 42) participated in the depression screening implementation training during three OCM performance periods. Depression screening rate of total ambulatory oncology care increased from 12% (OCM-PR 5, Jul-Dec 19), to 51% (OCM-PR 6, Jan-Jun 20) to 77% (OCM-PR 7, Jul-Dec 20) after the two top-volume clinics integrated depression screening into clinic intake process. Themes emerged from needs assessment revealed multi-level implementation strategies including 1) patient education and psycho-oncological care, 2) staff training and practice modification, 3) provider education & interdisciplinary Care, 4) leadership, administration, and staffing support, and 5) clinical informatics collaboration to build the infrastructure for integrating depression screen with clinic intake in the electronic medical record (EMR). Feedback from staff and provider interviews indicated high receptiveness and buy-in, especially during the COVID-19 pandemic to improve timely identification and triage of patients with depressive symptoms across all oncology care services. Conclusions: Depression screening is a key component of quality comprehensive cancer care that aims to provide timely identification and triage of cancer patients needing follow-up psychosocial care. Early implementation outcomes revealed significant improvement in depression screening completion rate after two clinics adopted depression screening into intake process. Further investigation is needed to refine system-wide implementation strategies across all ambulatory oncology sites and to assess long-term implementation outcomes meet the psychosocial care needs of cancer patients.

Hesitancy and malignancy: Vaccine hesitancy among individuals with cancer

148

Background: The World Health Organization cited vaccine hesitancy as one of 2019’s top ten threats to global health, a threat that has been further exacerbated by COVID-19 pandemic. Existing COVID-19 vaccine hesitancy research focuses on the general population, but less is known about the specific concerns of medically vulnerable populations, including individuals with cancer. Methods: This cross-sectional analysis used data that assessed likelihood of COVID-19 vaccination (likely vs unlikely/unsure) among past or current patients with cancer from a nationwide survey administered in December 2020 by the Patient Advocate Foundation (PAF), a non-profit organization that provides case management and financial aid to patients diagnosed with a chronic illness. Inclusion criteria included previous or current cancer treatment, aged ≥ 19, and a valid e-mail address. Age, sex, race/ethnicity, and urban/rural residence were abstracted from the PAF database. Respondents self-reported education level, employment status, trust in media regarding COVID-19 pandemic, and media viewership on COVID-19 vaccine development. The Group-Based Medical Mistrust Scale assessed respondents’ level of mistrust in medical providers based on ethnicity. Likelihood of COVID-19 vaccine acceptance was evaluated using risk ratios (RR) and 95% confidence intervals (CI) from modified Poisson regression models with robust error variance. All variables were included in our model. Results: Of 429 respondents, 48% were unlikely/unsure about accepting the COVID-19 vaccine, primarily due to concerns about vaccine safety (32%) and worry about health conditions (12%). When compared to those likely to accept COVID-19 vaccine, respondents who were unlikely/unsure were more often Black, Indigenous, or People of Color (40% vs. 23%), aged 36-55 (40% vs. 29%), and female (80% vs. 65%). In adjusted analysis, Black respondents were 55% less likely to accept a COVID-19 vaccine, when compared to White respondents (RR 0.55; 95% CI 0.4-0.8). When compared to those who did not follow the media regarding COVID-19 vaccine development, those who followed the media very closely were 4.5 times more likely to accept a COVID-19 vaccine (RR 4.5; 95% CI 1.6-13.2). Respondents who reported below average trust in the media were 60% less likely to accept a COVID-19 vaccine (RR 0.6; 95% CI 0.5-0.8), compared to those who reported above average trust in the media. Conclusions: Despite being at high risk of COVID-19 morbidity and mortality, a substantial proportion of under-resourced individuals with cancer were unlikely/unsure about vaccination, exposing a significant disconnect between risk of severe disease and vaccine acceptance. Our analysis also reveals a need to assess for and debunk misinformation to increase vaccine enthusiasm among medically vulnerable populations.

Disruptions to U.S. medical oncology care during the COVID-19 Pandemic: CancerLinQ Discovery (CLQD) analysis

6547

Background: The COVID-19 pandemic disrupted all facets of healthcare delivery including cancer care. This study evaluates the disruptions to US medical oncology practice during the pandemic in terms of number and type of patients (pts) encounters to determine the impact on continuity of patient care. Methods: We conducted a retrospective cohort analysis using the CLQD electronic health record database, containing data from 2+ million pts from all 50 states. We assessed changes in the monthly proportions of visit encounter types (in-person outpatient [IPOP] and telehealth [TE]) for new and established patients (NP and EP) with an invasive malignancy, benign or in situ neoplasm, or benign hematology diagnosis having an encounter between 1/1/2018 and 9/30/2020. Results: 781,945 pts were studied. Median age on 1/1/2018 was 64 years (IQR: 53-73), 38% were female, and 58% had an invasive malignancy. From 12/2019 to 9/2020, total monthly encounters dropped from 157,964 to 90,662. Monthly IPOP visits for NP dropped from 11.2% to 7.9%, an absolute drop of 3.3% and a relative drop of 30%; TE for NP increased by 1.1% (Table). Monthly IPOP visits for EP, as a percentage of all visits, dropped from 94.4% to 86.6% from 12/2019 to 6/2020 but rebounded to 90.4% by 9/2020. Fraction of TE increased substantially during the pandemic period reaching a peak in 6/2020 (13.8% for EP and 1.6% for NP) and decreased in 9/2020 to 9.6% and 1.1% for EP and NP, respectively. Compared to non-Hispanic patients, Hispanic patients had a larger reduction in IPOP and more TE during the study period. Percentage of monthly encounters, by type, from baseline*. Conclusions: We observed a reduction in the absolute number and monthly percentage of IPOP encounters during the COVID-19 pandemic. For EP, increases in TE does not fully compensate for reductions in IPOP. The reduction in IPOP NP encounters is particularly concerning since it was not accompanied by a compensatory increase in TE. The reduction in NP is consistent with reported pandemic-associated reductions in cancer screening and suggest a notable delay in cancer diagnoses during the pandemic. Reduction in Hispanic IPOP encounters warrants further evaluation.[Table: see text]

Impact of COVID-19 pandemic on time to treatment initiation for patients with advanced cancer

1528

Background: The COVID-19 pandemic has disrupted US healthcare delivery and led to delays in life-prolonging therapy for some conditions. Its impact on diagnosis and timely care delivery for patients (pts) with cancer is unknown. We assessed the pandemic’s impact on time from advanced diagnosis to systemic treatment initiation (TTI) for pts with newly diagnosed advanced solid cancers. Methods: We performed a controlled interrupted time series analysis using the nationwide Flatiron Health electronic health record-derived de-identified database, which originated from ̃280 US cancer clinics. The study sample included pts ≥ 18 years diagnosed with advanced solid cancers from Jan 1-Jul 31 in 2019 or in 2020, excluding a 30-day period (Mar 8-Apr 7) encompassing the start of most state stay-at-home orders. We used Cox proportional hazards models to estimate standardized predicted probabilities of TTI within 30 days of advanced diagnosis before (Jan-Mar) and during (Apr-Jul) the pandemic in 2020, compared to historical controls in 2019, adjusted for age, sex, race, insurance, performance status, and cancer type. Interactions by cancer type and race examined heterogeneity of effects. Results: The study included 12,977 pts (median age 69 yrs [IQR 61-77]; 47.4% female; 59.4% non-Hispanic white). At the time of analysis, fewer advanced cancer diagnoses were recorded in 2020 (Jan-Mar 2,409; Apr-Jul 3,027) than in 2019 (Jan-Mar 2,910; Apr-Jul 4,631). Compared to Apr-Jul 2019, pts diagnosed with advanced cancer during the COVID-19 period were more likely to have de novo (vs recurrent) disease (67.3% vs 56.8%). In adjusted models, the COVID-19 period was associated with an increased probability of treatment within 30 days (adjusted difference-in-differences +5.2 percentage points [ppts]). TTI improvements were not observed for pts with advanced breast cancer or Black pts, but effect differences across subgroups were not statistically significant (Table). Conclusions: Among pts diagnosed with advanced cancer, the COVID-19 pandemic was associated with shorter time to systemic therapy initiation. These treatment patterns may reflect the fewer advanced cancer diagnoses and higher proportion of de novo cancers observed during this period. Longer follow-up and data maturity are needed to understand the impact of the pandemic on clinical outcomes.[Table: see text]

Clinical trial representativeness and treatment intensity in a real-world sample of women with early-stage breast cancer

6584

Background: Early stage breast cancer (EBC) treatment is used in women of all ages, races, and health states. However, as clinical trials often do not represent real-world populations, the extent to which evidence-based treatments are prescribed to populations not well represented in these trials is not known. This study evaluated treatment intensity for patients traditionally well represented, underrepresented, and unrepresented in clinical trials. Methods: This retrospective cohort study used the nationwide de-identified electronic health record derived Flatiron Health database for patients diagnosed with EBC between 2011-2020. We categorized treatments as either high- (AC-TH [doxorubicin, cyclophosphamide followed by paclitaxel or docetaxel, trastuzumab]; ACT [paclitaxel or docetaxel, doxorubicin, cyclophosphamide]; TCH [paclitaxel or docetaxel, carboplatin, trastuzumab]; TCHP [paclitaxel or docetaxel, carboplatin, trastuzumab, pertuzumab]) or low-intensity (AC [doxorubicin, cyclophosphamide]; TC [paclitaxel or docetaxel, cyclophosphamide]; TH [paclitaxel or docetaxel, trastuzumab]). Unrepresented patients often have one or more comorbidities and/or prior cancer; underrepresented patients are typically Black, Indigenous, people of color, or of age extremes ( < 45, 70+); well represented patients are White and between the ages of 45-69. Odds ratios (OR), predicted proportions, and 95% confidence intervals (CI) from a two-level (patients nested in practice) hierarchical logistic regression model evaluated associations between receipt of high-intensity chemotherapy and patient characteristics of clinical trial representation (age, race/ethnicity, presence of comorbidity). Results: Our study included 970 patients with EBC with 13%, 45%, and 41% characterized as unrepresented, underrepresented, and well represented in clinical trials, respectively. In the adjusted model, those aged ≥ 70 vs 45-69 had lower odds of receiving a high-intensity treatment (OR 0.40, 95% CI 0.26-0.60), while those aged < 45 vs 45-69 had higher odds of receiving high-intensity treatment (OR 1.82, 95% CI 1.10-3.01). The predicted proportion of patients receiving a high-intensity treatment was 87% (95% CI: 80%-92%) for patients aged < 45, 79% (95% CI: 74%-84%) for patients aged 45-69, and 60% (95% CI: 50%-70%) for patients aged ≥ 70. Neither race/ethnicity nor comorbidity status were associated with odds of receiving high-intensity chemotherapy. Conclusions: Over half of the EBC population is not well represented in clinical trials. Age was associated with differential treatment intensity, despite a lack of evidence that these differences are appropriate. Widening clinical trial eligibility criteria is one way to better understand survival outcomes, identify potential toxicities, and ultimately make evidence-based treatment decisions using a more diverse sample.

Real-world quality of life (QoL) in black, indigenous and people of color (BIPOC) treated with palbociclib (PAL) and endocrine therapy for hormone receptor–positive (HR+)/human epidermal growth factor receptor 2–negative (HER2–) advanced breast cancer (ABC): A subgroup analysis from POLARIS

1071

Background: Racial disparities in breast cancer incidence, mortality, and care are well documented. PAL plus endocrine therapy is indicated for patients (pts) with HR+/HER2− ABC. Findings from the PALOMA clinical trials have shown that pts receiving PAL maintained stable QoL; however, limited QoL data are available from real-world settings for BIPOC receiving PAL. Methods: POLARIS is a noninterventional, prospective, primarily US-based study in pts with HR+/HER2– ABC receiving PAL. QoL was assessed with the European Organisation for Research and Treatment of Cancer Quality-of-Life Questionnaire Core 30 (EORTC QLQ-C30) at baseline, monthly for the first 3 mo of treatment (Tx) with PAL, and then every 3 mo. In this interim analysis, we report Tx patterns and QoL assessments at baseline and at 6 mo and 12 mo in BIPOC from POLARIS. Results: Of 1280 pts treated with PAL as November 10, 2020, 233 were included in the BIPOC subgroup of whom 159 (68.2%) completed PAL Tx for ≥6 mo and 112 (48.1%) for ≥12 mo. In the BIPOC cohort, 59.2% of pts were black, 35.2% Hispanic, 3.4% American Indian or Alaskan native, 2.1% Pacific Islander. PAL in combination with letrozole/anastrozole was received by 116 pts, 94 received PAL plus fulvestrant, 13 received PAL plus exemestane, and 10 received PAL plus another Tx; 175 pts (75.1%) received PAL as first-line Tx. Mean EORTC QLQ-C30 global health QoL and functional scales scores remained stable over the first 12 mo of PAL Tx, without any changes at or above the 10-point threshold considered clinically meaningful, and were similar to those previously reported in an earlier analysis of the entire POLARIS population (Table and Rocque et al SABCS 2019). Symptom scales scores, including nausea and vomiting, pain, dyspnea, insomnia, appetite loss, constipation, and diarrhea, also remained stable over 12 mo. Conclusions: In this subgroup analysis, PAL had no significant adverse impact on QOL in BIPOC with HR+/HER2– ABC, consistent with previous findings from the total POLARIS study population. Pfizer (NCT03280303). Clinical trial information: NCT03280303 .[Table: see text]

Effect of rurality and neighborhood disadvantage on clinical trial interest and decision-making style in patients with cancer living in the Deep South

110

Background: Patients living in rural or disadvantaged settings are historically underrepresented in clinical trials. This study sought to understand associations between neighborhood characteristics and both interest in clinical trial participation and decision-making style preference. Methods: This cross-sectional study used patient-reported outcome data from patients with cancer treated at the University of Alabama at Birmingham from January 2017 to May 2019. Rural-Urban Commuting Area Codes (RUCA) scores were used to determine rurality of patient residence. Area Deprivation Index (ADI) values (range 0-100) were used to identify patients living in the most disadvantaged (top 15%) census block groups. The Control Preferences Scale captured decision-making preference. Likelihood of interest in clinical trial participation by rurality and neighborhood disadvantage was estimated using risk ratios (RR) and 95% confidence intervals (CI) from modified Poisson regression models. Multinomial regression was used to calculate RRs and 95% CIs estimating likelihood of preferred decision-making style by rurality and neighborhood disadvantage. Models were adjusted for age, sex, race, cancer type, cancer stage, ECOG performance status, and phase of care. Results: Of 1005 patients with cancer, mean age was 67 (SD 11), 68% were female, and 74% white. Gynecologic cancer (32%) was the most prevalent diagnosis, followed by hematologic (20%) and breast (15%) cancer. Of this sample, 16% of patients lived in a rural setting and 18% lived in a disadvantaged neighborhood. Interest in clinical trial participation was no different for patients living in rural vs. urban (RR 0.93, 95% CI 0.73-1.17) or disadvantaged vs. non-disadvantaged neighborhoods (RR 0.88, 95% CI 0.69-1.13). Patients living in rural vs. urban settings trended toward increased likelihood of preferring physician- to patient-driven decision-making (RR 1.67, 95% CI 0.95-2.94). Patients living in disadvantaged vs. non-disadvantaged neighborhoods trended toward increased likelihood of preferring physician- to patient-driven decision-making (RR 1.39, 95% CI 0.82-2.35). Conclusions: Though clinical trial participation interest was similar, patients with cancer living in rural vs. urban settings trended toward increased likelihood of preferring physician- vs. patient-driven decision-making. Opportunities exist for providers to engage historically underrepresented patients for trial participation.

The relationship between treatment intensity and characteristics of patients with early-stage breast cancer

125

Background: Clinical trials are used to generate standard-of-care, yet often do not reflect patient populations treated in real-world settings. Elderly patients or patients of color who are often underrepresented in trials, which may impact what types of treatments are prescribed. This study examines how patient characteristics are associated with treatment intensity in early stage breast cancer. Methods: This retrospective cross-sectional study included women with a stage I-III breast cancer from American Society of Clinical Oncology’s CancerLinQ database treated by chemotherapy from 2005-2019. Seven standard-of-care regimens were characterized by intensity. For patients with ER+/- HER2- breast cancer, low-intensity regimens were Taxol and Cyclophosphamide or Adriamycin and Cyclophosphamide; while Taxol, Adriamycin, and Cyclophosphamide was considered high intensity. For patients with HER2+ breast cancer, the low intensity regimen was Taxol and Herceptin; while Adriamycin and Cyclophosphamide followed by Taxol and Herceptin; Taxol, Carboplatin, and Herceptin; or Taxol, Carboplatin, Herceptin, and Pertuzumab were considered high intensity. A model estimating the likelihood of intensity was calculated using log-binomial regression, in order to produce relative risks. The models were adjusted for patient demographics and cancer stage. Results: Of 24,383 patients, 51% had ER+HER2-, 20% ER-HER2-, and 29% HER2+ breast cancer. Most patients were White (60%), age 40-69 (80%), had stage II breast cancer (39%), and received higher intensity treatment (65%). Adjusted for the other covariates, patient who were Black were more likely to receive high-intensity treatment than patients who were White (61% vs 58%; RR 1.05, 95%CI 1.02-1.06. Additionally, older adults were more likely to receive low-intensity treatment, with 42% of patients over 70 receiving low intensity treatment, and 29% of patients between the ages 40 and 69 received low intensity treatment (RR 1.5, 95% CI 1.44 -1.54). Conclusions: Differences in treatment intensity were observed for patients with differing demographic characteristics. Further research is needed to determine lack of representation in clinical trials impacts on prescribing patterns, regimen intensity, and survival.

Quantifying treatment preferences and their association with financial toxicity in women with breast cancer

141

Background: This study sought to understand treatment preferences and their association with financial toxicity in breast cancer patients served by Patient Advocate Foundation (PAF). Methods: This cross-sectional study used survey data from a nationwide sample of women with breast cancer who received assistance from PAF. Choice-based conjoint analysis elicited patient preferences and trade-offs. Latent class analysis segmented respondents into distinct preference groups. The Comprehensive Score for Financial Toxicity (COST) tool captured financial toxicity (range 0-44, lower scores indicate worse financial toxicity). Cramer’s V determined magnitude of relationships in bivariate associations. COST score differences by preference archetype was estimated by least square means and naïve 95% confidence intervals (CI) from adjusted generalized linear models. Results: Of 220 respondents (65% response rate), median age was 58 years (interquartile range [IQR] 49-66) and 60% had household incomes < $40,000. Most respondents were diagnosed with early stage cancer (91%), with 41% diagnosed within the past 2 years; 38% had recurred. Almost two-thirds (61%) were on active treatment. Treatment choice was most affected by preferences related to affordability and impact on activities of daily living (ADLs). Two distinct treatment preference archetypes emerged. The “Cost-Prioritizing Group” (75% of respondents) was most concerned about affordability, impact on ADLs, and burdening care partners. The “Functional Independence-Prioritizing Group” (25% of respondents) was most concerned about ability to work, physical side effects, and interference with important life events. Cost- vs. functional independence-prioritizing respondents were more often diagnosed with an early stage cancer (88% vs. 78%; V = .22), white (78% vs. 56%; V = .21), or privately insured (45% vs. 36%; V = .12). Functional independence- vs. cost-prioritizing respondents more often had household incomes < $40,000 (76% vs. 54%; V = .20), identified as Hispanic/Latino (20% vs. 9%; V = .15), or had Medicaid (15% vs. 7%; V = .12). COST scores were similar between archetypes in adjusted models (Cost-Prioritizing COST = 12, 95% CI 9-14; Functional Independence-Prioritizing COST = 11, 95% CI 9-13). Conclusions: Patients with breast cancer prioritized affordability or maintaining functional independence when making treatment decisions. Because of this variability, preference evaluation during treatment decision-making could optimize patients’ treatment experiences.

Financial toxicity among breast cancer survivors with health insurance

12073

Background: Cancer treatment and its sequelae have been associated with financial toxicity in breast cancer survivors, particularly those who have no health insurance. However, the prevalence of financial toxicity in the insured survivors, and the underlying factors are not well understood. Methods: Breast cancer survivors attending a survivorship clinic (University of Alabama at Birmingham) completed a survey assessing demographics, financial toxicity (i.e., material resources; food/housing/energy insecurity), and health-related quality of life (HRQL: SF-36). Clinical characteristics were abstracted from medical records. A multivariable logistic regression model was developed to understand factors associated with financial toxicity; the model included survivor age, race, socioeconomic status, insurance type, marital status, cancer stage, time since diagnosis, current medications, and physical and mental domains of HRQL. Results: The 368 participants (1% male; 67% white, 25% African American, 8% other) were a median of 61y of age (range, 33-86y) and 4.3y post-diagnosis (1-34y) at survey completion; 90% had stage 0-II disease; 34% were single (not currently married/partnered); type of health insurance included private/military (57%), Medicare (39%), and Medicaid/self-pay (4%). Overall, 31% reported financial toxicity; 26% endorsed not being able to live at current standard of living > 2 mo. if they lost all current sources of income; 6% endorsed energy insecurity, 5% endorsed food insecurity, and 4% endorsed housing insecurity. In a multivariable model, financial toxicity was associated with age ≤60y at survey (Odds Ratio [OR] 5.1; 95% confidence interval [CI] 2.0-13.3); household income < $50K/y (OR 5.3; 95%CI 2.5-11.2); being single (OR 2.6; 95%CI 1.3-5.4); and lower physical (OR 2.6; 95%CI 1.2-5.4) and mental (OR 2.2; 95%CI 1.2-4.3) HRQL. Cancer stage, race, time from diagnosis, and insurance type were not associated with financial toxicity. The prevalence of financial toxicity among survivors who were single, ≤60y at survey, and with household income < $50k/y was 79.3%, compared with 6.7% among those who were older, married/partnered, and with higher income. Conclusions: Financial toxicity is prevalent among insured breast cancer survivors several years after cancer diagnosis, and is exacerbated among the younger survivors who are single, with low household income, and endorse poorer physical and mental quality of life. These findings inform the need to develop interventions to mitigate financial toxicity among at-risk breast cancer survivors.

Implementation of the Institute of Medicine treatment plan in oncology practice

e19233

Background: Participation in the Center for Medicare and Medicaid Services (CMS) value-based payment reform, The Oncology Care Model, requires that every beneficiary has a documented 13-point Institute of Medicine (IOM) treatment plan (TP) when commencing antineoplastic therapy. The intent of this document is to enhance shared decision making between the patient and care team by providing transparent treatment recommendations and engaging patients and caregivers in meaningful discussion. There is limited discussion in the literature on how to adapt the CMS recommendations to diverse practice settings while maintaining fidelity to the intent of the TP. Methods: We compare how three clinically and geographically unique OCM participating institutions implemented the TP in their respective institutions within the domains of the Consolidated Framework for Implementation Research (CFIR). Settings include a community cancer institute in the northeast, an academic hospital setting in the southeast, and a large community cancer network in the southern United States. Results: We identified similar themes in implementation including engaging stakeholders, leveraging information technology, structuring the TP, development of clinic processes and considering scalability. We also describe adaptations unique to the culture and setting of each site. Conclusions: Although studies have shown patients do not feel informed of their diagnosis, there are currently many approaches to improving shared decision making including utilizing the 13 points of the IOM TP as mandated by the OCM. We provide practical strategies for incorporation of the TP into clinical care with lessons from diverse settings. As shown by the wide variability in implementing shared decision making, further research is needed to optimize illness understanding. Additionally, optimal implementation of CMS’s IOM TP would ideally include concrete metrics measuring impact on shared decision making, illness understanding, or patient satisfaction.

Utilizing visualization to qualitatively evaluate electronic health record-derived database limitations

317

Background: Electronic health record (EHR) databases are a promising platform for clinical research using real-world data. However, information on potential limitations of these data sources is lacking. We sought to understand how data visualization might be used to identify data inconsistencies and the applicability of previously validated claims-based algorithms used to identify patients with metastatic breast cancer (MBC). Methods: This retrospective study utilized ASCO’s CancerLinQ Discovery database derived from EHR data. Subjects included women ≥18 years treated for MBC diagnosed ≥1980. Subjects with MBC were identified using two billing codes for metastasis on separate dates following primary breast cancer diagnosis. Treatment course sequences were visualized. Patients were represented by a horizontal bar on the Y-axis. Treatments were displayed using colored bars (blue: chemotherapy, red: endocrine therapy, green: HER2 targeted, orange: novel therapy) with time of treatment on the X-axis. Visualizations were qualitatively evaluated, and treatment patterns inconsistent with clinical practice were identified. Results: We identified 4,760 women treated for MBC using billing codes for primary breast cancer diagnosis and distant metastasis. Most patients (96%) had a primary breast cancer diagnosed in 2000 later. Treatment patterns inconsistent with clinical practice identified using the visualization technique included: 1% of patients received adjuvant chemotherapy continuously for ≥1.5 years, suggesting missed coding for metastatic disease; 5% of patients did not receive any treatment in the year following metastasis, suggesting the billing code may have been used in workup and not for confirmed metastatic disease. Among patients with MBC, 50% identified as HR+ across all records had not received hormone therapy, while 39% identified as HR- across all records received hormone therapy. Conclusions: Because previously validated algorithms may not translate well to EHR databases, quality auditing should always be performed. The proposed data visualization can be used for improving algorithms, qualitatively identifying errors, and avoiding biased or inaccurate results.

Treatment decision-making and financial toxicity in metastatic breast cancer

97

Background: Shared decision-making (SDM), a process where patients partner with their physician to incorporate personal preferences into treatment decisions, is a tenet of high-value healthcare. It is unknown if high-value care associated with SDM manifests in the form of decreased out-of-pocket costs. Therefore, this study analyzes the relationship between patient preference for SDM and financial toxicity in metastatic breast cancer (MBC). Methods: This cross-sectional study utilized surveys of women age ≥ 18 with MBC who received care at two academic hospitals in Alabama between 2017 and 2019. SDM preference and financial toxicity were measured using the Control Preferences Scale and the Comprehensive Score for Financial Toxicity (COST) tool (11-item scale from 0-44, with lower scores indicating worse FT), respectively. Patient demographic and clinical data were abstracted from the electronic medical record. Effect sizes were calculated using Cohen’s d or Cramer’s V. Differences in financial toxicity by SDM preference were estimated using mixed models clustered by site and treating medical oncologist. Results: In 79 women with MBC, 41% preferred SDM, 33% preferred provider-driven decision making, and 22% preferred patient-driven decision making. Patients preferring SDM were more often college educated (48% vs. 40%; V = .15), higher income (52% vs. 44%; V = .09), and privately insured (47% vs. 41%; V = .11). Overall median COST score was 23 (interquartile range 16-30), which varied modestly by SDM preference. After adjusting for patient demographic and clinical characteristics, similar financial toxicity levels were found in patients who preferred SDM (COST 22, 95% confidence interval [CI] 19-25), patient-driven decision making (COST 22, 95% CI 18-26), and provider-driven decision making (COST 24, 95% CI 20-27). Conclusions: Similar levels of financial toxicity were found in patients with differing decision–making preferences regarding their MBC treatment, which may be secondary to lack of discussions about cost. Further research is needed to determine if and how financial toxicity is being identified or included within decision-making.

Changes in employment and insurance for patients with cancer receiving safety net services

104

Background: Under-resourced patients with cancer often face financial burdens due, not only to costs of treatment, but also from side effects that prevent individuals’ ability to work, which impact employment status and may cause insurance coverage loss. Financial assistance may be sought from safety net programs, which provide both material support and financial counseling. However, knowledge of the impact of cancer on employment and insurance in a population seeking safety net services is limited. Methods: This observational, cross-sectional study uses data on safety net services from a nationwide survey conducted in July 2017 and distributed by the Patient Advocate Foundation (PAF). The survey respondents included patients with cancer who received services from PAF from July 2016 to June 2017. Descriptive statistics were calculated using frequencies for categorical variables. Results: A total of 508 patients with cancer completed the survey. Most patients had a diagnosis of breast cancer (47%), followed by myeloma (13%), and prostate cancer (8%). The majority of patients reported that their illness affected their employment (67%); by either job loss (13%), income loss (24%), or inability to work as usual (27%). Of these patients, 27% lost their insurance coverage. Those able to enroll in a new insurance plan reported having more expensive rates (40%) and fewer covered services (36%) compared to their previous coverage. The most commonly utilized governmental safety net services were Social Security Disability Insurance (19%) and Medicaid (12%). Non-governmental safety net services such as financial assistance from non-profits (27%) and free medication from drug companies (13%) were also frequently used. Beyond their insurance coverage, cancer patients still needed assistance paying for diagnostic tests (18%), clinic visit fees (23%), and prescription drugs (15%) from the safety net program. Conclusions: Cancer patients commonly experience financial burden due to losses in employment and insurance, resulting in need for safety net programs. Further work is needed to identify approaches to reducing the adverse financial impact of cancer care.

Patient and physician perceptions: Importance and documentation of patient priorities in treatment planning and decision-making

224

Background: The National Academies of Medicine prioritizes patient’s needs, values, and preferences when patient’s and care teams make treatment decisions. However, the collection of this patient information is not part of the formal clinical intake process, nor are pertinent questions included in EHR templates. Methods: Surveys were conducted of cancer patients and cancer physicians to understand perceptions on whether patients’ quality of life priorities (e.g work, hobbies, key events, household responsibilities) are discussed, documented and reflected in treatment plans. Physicians were recruited via a market research panel. Patients were recruited from the Cancer Care data base via an email request to complete an online survey. Results: From February to October, 2018, 310 cancer patients and 109 cancer physicians completed surveys. Cancer patients were mostly female, 91% had been diagnosed in the last year, 61% had breast cancer, and 55% were stage 1 or 2. Among physicians, 88% self-identified as trained in oncology, 33% practice at academic centers, 51% in community practices, 34% from the Northeast, 21% the Southeast, and 25% the Midwest. Most patients (62%) said it was very/extremely important their doctor know their priorities. Most physicians (66%) report they most of the time/always know patients’ personal quality of life priorities before finalizing treatment plans, and 62% say this information has a large/major impact on recommendations. However, only 40% of patients report having this conversation before treatment started. Although 76% of doctors said they discussed what is important to patients most of the time or always, only 60% report it is documented, usually in the social history or notes fields. Only 36% of patients are sure this information is entered into their medical record. Conclusions: Cancer patients want their doctors to know what is important to them, and physicians agree. However, documentation is often relegated to narrative notes fields in the EHR. Considering the key role patients’ priorities should play in treatment planning, there need to be standardized collection tools so this information can be timely collected and shared.

Evaluating the early impact of adding a financial counselor to an Oncology Care Model pilot program

67

Background: The Oncology Care Model (OCM) emphasizes the importance of patient costs. Our practice identified a need for patient financial counseling (Plan) and added a financial counselor to the care team (Do). Our objective is to describe the financial counseling process and evaluate (Study) the impact of adding a financial counselor to address financial burden among cancer patients. Methods: New treatment orders trigger a task to the financial counselor who mails a letter including personalized insurance benefit information (i.e., deductible, out of pocket maximum, amount met to date), estimated out of pocket cost for prescribed chemotherapy regimen, and additional financial resources. The health care team can also place referrals for identified needs. Data on patient characteristics, referral patterns, and services delivered was abstracted from the electronic medical record. Results: The financial counselor contacted 157 cancer patients from 4/2018-4/2019, including 134 (85%) OCM patients. The average age was 68 years old (range 22-91); 93 (59%) patients were female; 39 (25%) patients identified as black. Out of 32 patients with documented income, the average annual household income was $25,000. Most patients had Medicare Part D coverage (67/73, 92%) and secondary insurance (59/77, 77%). The financial counselor contacted patients by mail (52/77, 68%), phone (13/77, 17%), in-person at clinic (7/77, 9%), e-mail (1/77, 1%), or using multiple methods (4/77, 5%). Actions taken included mailing insurance benefit and estimated out of pocket cost letters (60/72, 83%), referring to social workers or lay navigators (5/72, 7%), or reviewing insurance benefits (4/72, 6%). Co-pay applications through foundation grants were submitted by the financial counselor for 27 patients of which most (23/27, 85%) were approved. Conclusions: Financial counselors provide diverse services to cancer patients to reduce financial burden, including to patients without secondary insurance who were noted to have additional financial needs. Patient satisfaction surveys and accounting for the dollar amount returned to patients and the health system represent next steps for our new program (Act).

Oncologic services through Project Access and other safety net care coordination programs

126

Background: Since the 1996 formation of Project Access, a nationwide program linking uninsured, low-income, or underserved patients to specialty care, the US has experienced a growth of third-party, safety net care coordination programs that provide sub-specialty care at no or reduced cost. These programs pre-negotiate with local physicians and hospitals to offer pro bono services to patients. Little is known about the provision of oncologic services within these programs. This study evaluated the current landscape of oncologic services within US safety net care coordination programs. Methods: This study reviewed websites via an online search from 04/22/19 - 04/28/19 using keywords “Project Access”, “healthcare”, and individual state names. The websites were reviewed for information on location, services offered, and patient eligibility criteria. Where online data on oncologic services was unavailable, individual programs were called on 05/08/19 for additional information. Five programs were unreachable by phone or e-mail. Results: Websites of 29 safety net care coordination programs in 22 states were identified, serving only approximately 40 counties. Online, one program highlighted "limited" access to oncologic care, while another offered chemotherapy at a reduced cost ($5000/round). When contacted by phone, 55% offered access to oncologic services and 48% assisted with chemotherapy costs. One program provided chemotherapy only if the cancer diagnosis was made after the patient was enrolled in their system. Virtually all programs who offered chemotherapy did so through affiliated large hospital systems. A commonly reported reason for lack of oncology care was a reluctance of oncologists to commit to potential long-term agreements of free care. Online, one program offered PET scans. By phone, 89% offered some imaging. Two programs (7%) explicitly required US citizenship for care. Conclusions: Third-party care coordination centers are increasing in number and provide a novel, and potentially unrecognized, approach to increasing oncology service access. Further research should identify barriers contributing to the relative lack of oncologic services in these programs compared to other specialties.

Understanding palbociclib practice patterns in a real-world setting

200

Background: Real-world practice patterns often differ from treatment in clinical trials. We assessed real-world standard-of-care treatment with palbociclib (PAL) in the context of previously reported PALOMA trial results. Methods: POLARIS is a prospective, observational study of PAL in patients (pts) with hormone receptor–positive, human epidermal growth factor receptor 2–negative advanced breast cancer. Baseline demographics, clinical characteristics, initial treatment dose, dose modifications, dose delays, and adverse events (AEs) during the first 6 months (mo) of treatment were analyzed. Results: 412 pts enrolled at 92 US sites had at least 6 mo of PAL treatment; 73% received PAL in the first-line setting and 27% in second or later line. While a majority of pts received the recommended dose of 125 mg/d; 6% and 1% started at 100 mg and 75 mg, respectively. Physician-reported reasons for choosing a lower dose were: comorbidities (31%), pt age (21%), past treatment (10%), patient preference (3%), and other (34%). Selected pt and clinical characteristics, dosing, and treatment outcomes in POLARIS and in PALOMA-2 are shown in Table. Conclusions: In a real-world data set of 6 mo of PAL treatment, most pts started at the recommended dose, with tolerability and safety outcomes consistent with those reported in clinical trials. Differing populations, treatment patterns, and outcome reporting in real world vs trial settings underscore the need to study real-world practices and outcomes. Clinical trial information: NCT03280303

[Table: see text]

Health insurance literacy, status, and financial toxicity in women receiving treatment for metastatic breast cancer

96

Background: Though uninsurance rates declined after the Affordable Care Act, the number of underinsured, or individuals who spend > 10% of their income on out-of-pocket (OOP) medical costs, continues to rise. In patients with metastatic breast cancer (MBC), underinsurance may lead to financial toxicity (FT), or patient-level financial burden and distress, since diagnosis and treatment is extremely costly. This study explores health insurance literacy and the association between FT and health insurance status in women receiving treatment for MBC. Methods: This cross-sectional study utilized survey data collected from 2017-2019 in women age ≥18 receiving treatment for MBC at two academic medical centers in Alabama. FT was measured by the Comprehensive Score for Financial Toxicity (COST) tool (11-item scale from 0-44, with lower scores indicating worse FT). Health insurance status and OOP costs were self-reported. Effect sizes were calculated using Cohen’s d or Cramer’s V. Mixed and generalized linear models clustered by site and treating medical oncologist estimated the association between FT and health insurance status. Results: In 81 women with MBC, median COST score was 24 (interquartile range [IQR] 17-30), 44% had private insurance, 40% Medicare, and 16% Medicaid. Though 25% and 33% of surveyed patients did not know their health insurance premium or deductible cost, respectively, privately insured patients more often knew the cost of their premiums (97%; V = 0.58) and deductibles (81%; V = 0.33) compared to publicly insured patients. In adjusted models, FT levels did not differ significantly based on health insurance type (private insurance COST 21, 95% confidence interval [CI] 18-25; Medicaid COST 23, 95% CI 17-29; Medicare COST 24, 95% CI 20-27). However, risk of severe FT (COST ≤13) was 147% higher for privately insured patients versus Medicare beneficiaries (risk ratio 2.47, 95% CI 1.44-4.21). Conclusions: Despite higher levels of health insurance literacy, privately insured patients receiving treatment for MBC may be at increased risk of severe FT. Further research is needed to understand causes of underinsurance in patients with MBC, which could lead to cancer-related FT.

Navigating costs of care in women with breast cancer: Examining racial differences in non-treatment costs and financial toxicity in under-resourced populations struggling to afford medical care

156

Background: Patients with cancer struggle to afford needed medical care alongside daily financial obligations due to rising healthcare costs. This may be more pronounced among minorities who are less likely to seek resources to mitigate their financial distress. This study sought to identify racial differences in non-treatment costs for under-resourced women with breast cancer who sought assistance from Patient Advocate Foundation (PAF). Methods: This cross-sectional study utilized secondary survey data collected from breast cancer patients receiving case management services from PAF in 2018. Respondents answered questions describing their financial distress and COmprehensive Score for financial Toxicity (COST) tool (0-44 with lower scores indicating worse toxicity). Descriptive statistics were calculated using means and standard deviations (SD) for continuous variables and frequencies for categorical variables. Two sample t-tests were used for bivariate comparisons between racial groups. Results: Of 267 breast cancer patients surveyed, 54% were Caucasian, 29% were African American (AA), and 83% indicated a household income of < $48,000. Cohorts expressed strong dissatisfaction with their financial situation with AA impacted more acutely (78% vs 56%) and acknowledged inability to pay for treatment costs (83% vs 58%). Compared to Caucasians, AAs were more often concerned with transportation costs (33% vs 16%) and with day-to-day living expenses (83% vs 59%). Younger (≤55 years) AA respondents were twice as often unable to meet monthly expenses (60% vs. 27%). Older AA respondents ( > 55 years) reported greater distress than older Caucasians (74% vs 57%), while younger Caucasians reported greater distress than their AA counterparts (72% vs 65%). COST scores differed significantly between Caucasians (mean 13, SD 9) and AAs (mean 11, SD 8; p = 0.04). Conclusions: While the impacts of medical care costs were felt by all survey respondents, under-resourced AA breast cancer patients may be at higher risk for household material hardships as financial resources are diverted toward essential healthcare costs.