Community Oncology Research Articles

A multi-modal EHR-based approach to screen patients for oncology clinical trials.

411 Background: Inefficiencies in patient (pt) screening may impair enrollment in oncology trials and could potentially be mitigated by a centralized trial matching infrastructure. The objective of this study was to implement a novel clinical trial screening service using electronic health record (EHR)-based structured variables, machine learning (ML) and human abstraction. Methods: To construct the model, we defined a set of inclusion and exclusion (IE) criteria unique to each trial and routinely documented in EHRs. Patients that met IE criteria via structured and ML-extracted clinical data were selected for review by trained specialists via a technology-enabled abstraction platform. Real-time notifications for eligible pts were issued via EHR alerts to research teams at participating study sites. Results: Between January and May 2024, over 80,000 pts were screened prior to a clinic visit for two trials at three sites. Using structured IE criteria alone, 98.7% and 99.6% of pts were excluded. An additional 31.9% of pts were excluded when employing ML modeling. A total of 15 (3.0%) and 5 (4.0%) pts were deemed eligible when using structured, ML, and abstraction-based criteria (Table). Seventeen percent of surfaced pts were consented within 10 days of implementation. Median time for abstractor confirmation of pt eligibility was 13.3 hours (interquartile range: 2.8, 27.4) with abstraction of highest priority pts taking a median of 3.0 and 9.9 hours, respectively. Conclusions: Our results support the implementation of a centralized, technology-enabled pt screening service for effective trial eligibility assessments across large patient populations. The value of using structured, ML- and abstraction-based data models, improved accuracy by more than 95% compared to structured data alone. The impact of multi-modal centralized screening is being evaluated across additional clinical trials in a nationwide community oncology research network. Pts, No. (% pts excluded). Trial 1 Trial 2 1. Pts with upcoming visit 37,865 44,973 2. Pts meeting structured criteria (including: age, stage, prior IV therapy) 497 (98.7) 182 (99.6) 3. Pts meeting ML model-based criteria (including: prior oral therapy) - 124 (31.9) 4. Pts meeting steps 1-3 and abstraction criteria (including: cancer diagnosis, biomarker status, prior malignancies) 15 (97.0) 5 (96.0)

The impact of oral oncology drugs in the era of value-based care.

15 Background: Oncology drug expenditures are a significant contributor to increasing total cost of care (TCOC). Oral oncolytics (OO), such as small molecule inhibitors (CDK 4/6 inhibitors, tyrosine kinase inhibitors, etc.), and immunomodulators, have revolutionized the treatment landscape for many cancers. We evaluated the patterns of OO use and spend for Medicare patients over time at a large community oncology practice participating in the Center for Medicare and Medicaid Innovation (CMMI)’s Value-based care (VBC) Program, the Oncology Care Model (OCM), with a focus on the seven cancers that are part of a successor model, the Enhancing Oncology Model. Methods: Using CMMI claims for attributed 6-month OCM episodes, we evaluated drug use and expenditures by performance period (PP) from July 2016 – June 2022 (11 PPs), with an emphasis on OO drugs. We used Part D expenditures as a proxy for OO. Results: From 2016-22, average TCOC per PP increased by 15.5% in the 7 EOM diagnoses (breast, lung, colon, prostate cancer, myeloma, lymphoma, and chronic leukemia). Average drug expenditures per PP grew 33% percent and represented 71% of total cost by PP11, up from 61% in PP1. Average expenditures per episode increased by 11.7% for Part B and 82.6% for Part D drugs (p<0.01). Increase in Part D drug expenditures was partly due to an increase in the number of unique oral drugs which grew by 32% over the evaluation period. The greatest increases in Part D drug expenditures occurred in breast cancer (87%) and lymphoma (153.7%). Table reports average OO expenditures per PP, percent change from PP1 to PP11, and the change in the % Part D drugs contributed to TCOC. Conclusions: Increases in drug expenditures represent the greatest impact on TCOC for cancer patients. OO medicines have an increasing role in the treatment landscape and represent a growing focus in VBC models. Utilization and cost of OO medication have a significant bearing on VBC performance. Diagnosis Avg. OO Expenditures PP1 Avg. OO Expenditures PP11 Avg. OO Expenditures Change Contribution of OO to TCOC Change from PP1 - 11 Breast $4,208 $7,869 87.0% 9.9% Leukemia $25,334 $41,128 62.3% 19.9% CRC $1,206 $2,158 78.9% 3.1% Lung $3,650 $3,625 -0.7% -2.5% Lymphoma $4,193 $10,637 153.7% 13.0% Myeloma $25,109 $39,298 56.5% 6.7% Prostate $18,815 $20,114 6.9% 3.2% Total $8,291 $15,140 82.6% 10.8%

Virtual addressing reality: Development of a virtual social work program to address health-related social needs.

405 Background: Assessment of Health-Related Social Needs (HRSNs) has become an integral part of understanding barriers to care and improving patients’ lives and experiences as they receive cancer care. Increased assessment, however, can result in a raised awareness of unmet needs in oncology patient populations. This tends to highlight gaps in services and necessitate additional staff to attend to these needs. Addressing HRSNs can be facilitated with lay navigators (non-licensed staff), licensed staff, clinical staff, and other members of a multi-disciplinary team. Actions can include, but are not limited to, identification of practical concerns, referrals to community-based organizations (CBOs), referral to the clinical team, or further escalation to a licensed clinical social worker (LCSW). Methods: The National Comprehensive Cancer Network (NCCN) Distress Thermometer screening tool was implemented in July 2023 where patients are screened every 3 months in a community oncology setting. A workflow was established where a clinic staff would administer the screening tool and enter the details of the screening into the EHR system. Based on established criteria, the patient would be referred to the social work team via electronic referral. Due to an increase in screening and subsequent referrals, a virtual team of (5) social workers called Patient Resource Coordinators (PRCs) was established in December 2023 to address the practical concerns and facilitate the coordination of resources based on identified needs. After the team was established, non-urgent referrals were then assigned to a PRC. The PRC could refer to an on-site LMSW/LCSW for higher level of acuity assessment, like crisis support or positive PHQ-9 scores. Results: See Table. Conclusions: The virtual PRC team is able to manage ~55% of referrals made to a social worker based on results from an HRSN screening tool. This has allowed the onsite LMSWs/LCSWs to work at the top of their license and enabled patients to receive support with care coordination and connection to resources. Over time, the referrals managed by the virtual team has increased and their connections to patients have grown. In the future, the PRC team aims to add additional LCSW services to assist patients with higher acuity needs. This will broaden the scope of services available to the patients to address their health-related social needs in a timely manner. PRC Activity Dec ’23 Jan ‘24 Feb ‘24 Mar ’24 Apr ‘24 May ’24 (*May to date at time of submission) # of Referrals Received via EHR 110 542 610 586 1216 570 # of Referrals Assigned to PRC 45 245 299 315 660 312 % of Referrals Assigned to PRC 41% 45% 49% 54% 54% 55% *Secondary Referral to SW 8 9 20 4 # of calls attempted 102 479 536 577 959 448 # of patients reached 38 180 217 270 413 277 % of patients reached 37% 38% 40% 47% 43% 62% # of practical concerns identified 31 131 426 430 742 349 Total Visit Duration (in hrs.) 36hrs 181hrs 348hrs 450hrs 836hrs 485hrs

Oncology nurses’ assessment of providing equitable care for patients with developmental disabilities.

164 Background: Access to quality health care can be challenging, especially for people with developmental disabilities (PwDD). Developmental Disabilities (DD) occur before age 22 and include physical, learning, language, or behavioral impairments. Studies of training and experience of oncology nurses (RNs) who care for PwDD are limited. The purpose of this study was to explore to what extent RNs working in a community oncology practice are prepared to provide quality patient-centered care to patients with DD. Methods: As no tool exists to capture oncology RNs experiences caring for PwDD, a survey was developed with input from a group of oncology nurse experts. Both academic IRB and Privacy Board approvals were completed prior to recruitment. A Qualtrics survey link was shared with RNs and Advanced Practice RNs via email. Questions about nurses’ background, perceptions of clinic resources and accommodation and referral to clinical trials were included in the survey. Nurses were also presented with 3 scenarios of oncology patients who have either been blind since birth, have autism or cerebral palsy. RN respondents were asked how to care for them, what challenges might be encountered, and what resources would be available to them. Results: Usable surveys were completed by 122 RNs. 77 RNs cared for PwDD in a statewide Oncology practice, 81 RNs had previously cared for PwDD in any setting and 54 RNs have friends or family members with DD. Only 7 RNs reported no experience with PwDD. Seven nurses knew of PwDD who had participated in clinical trials. Themes identified through the qualitative analysis include Accommodations, Strategies, Communication and Referrals. Conclusions: This study provides an initial step in understanding how to reduce disparities for oncology patients with DD. Key elements revealed additional training needs, standardized facility accommodation needs, and advocacy for clinical trials participation. Through a team approach and engagement with patient advocacy groups, the challenge of engaging PwDD in clinical trials will be tackled. More training and resource education for RNs and staff regarding care of PwDD will be built and disseminated through a standardized process to bridge gaps to impact delivery of care.

A community health worker–led intervention to improve patient knowledge and receipt of precision medicine: Updated ALCANCE randomized clinical trial findings.

181 Background: Inequitable access to precision medicine limits its impact. To address disparities, we co-developed the Addressing Latinx Cancer Care Equity Study (NCT04843332), a multi-level intervention with community stakeholders. ALCANCE was designed for low-income and Latinx adults with cancer in rural Monterey County, California. We provided annual training for clinicians about testing guidelines and precision medicine disparities. Trained CHWs educated patients about precision medicine and encouraged them to raise the topic with their oncologist. We tested whether the intervention improved patient knowledge and receipt of precision medicine. Methods: Patients with newly diagnosed or recurrent cancer, age ≥ 18 years, able to consent in English or Spanish, and who self-identified as a racial and ethnic minority OR low-income status OR were insured by public or agricultural company insurance OR uninsured and receiving cancer care at Pacific Cancer Care, a rural community oncology practice in Monterey County, were randomized 1:1 to usual care (control group) or usual care augmented with the multi-level intervention for 12 months (intervention group). The primary outcome was whether the multilevel intervention could increase patient knowledge of precision medicine using a validated 6 question survey from time of enrollment (baseline) to 6-months post-enrollment between groups. A secondary outcome was receipt of precision medicine. CHWs and clinicians completed semi-structured interviews. Results: 110 participants were randomized; 70% (n = 67) reported Spanish as their primary language; 81% (n = 89) identified as Hispanic. Median age was 55 years (55.4 ± 14.3, mean± SD); majority had annual household incomes (USD) ≤ $34,999 (59.0%, n = 46); majority obtained a high school diploma/GED or less (67.3%, n = 74). Most had gastrointestinal (35.5%, n = 39) and breast (24.5%, n = 27) cancers and were diagnosed with stage 3 or 4 (64.5%, n = 71) disease. Precision medicine knowledge was low at baseline across both the control and intervention groups (23.5 ± 19.3 versus 23.0 ± 17.7, respectively). Intervention group participants had greater precision medicine knowledge at 6-months (39.2 ± 23.0) than control group participants (24.2 ± 20.0) with greater increase in knowledge overtime than control group participants (p < 0.001). More intervention group participants (n=46, 79.3%) received precision medicine testing than control (n=34, 60.7%). Across both groups, 26 of those tested had an actionable mutation. Of those 26, 66.7% (n=6 of 9) in the control group and 94.1% (n=16 of 17) in the intervention group received targeted therapy (p=0.065). Interviews with the care team indicated the model was acceptable and has been sustained as part of usual care. Conclusions: CHW-facilitated interventions may reduce disparities in precision medicine cancer care. Clinical trial information: NCT04843332

Closing the distance gap: Evaluating patient travel distances in a community oncology practice.

157 Background: Patient travel distance is an important factor influencing healthcare access, health equity and outcomes. The Rural-Urban Commuting Area (RUCA) classification system categorizes geographic areas based on urbanization and commuting patterns to measure accessibility to healthcare. We study the relationship between RUCA and patient travel distance for a multi-site community-based, medical oncology practice in north-west Chicagoland area. Methods: Using patient ZIP code data and practice locations (clinics) from April 2023 to March 2024, we calculated and compared the straight-line, one-way, patient-clinic travel distance (travel distance) for about 20,000 patients at 5 clinics, using geospatial analysis techniques. We grouped clinics based on location and population characteristics, applied the RUCA classification to patient ZIP codes, and compared travel distances by RUCA categories. Results: The median travel distance for patients was 5.0 miles, with 82% of patients traveling 10 miles or less, and 89% traveling 15 miles or less. Table 1 shows % of patients traveling and travel distance by RUCA categories. Median travel distances for urban metro, suburban metro, and suburban small-town clinics were 4.8mi, 10.8mi and 20.6mi (Kruskal Wallis H-test, p <0.001). Within 50 miles of each clinic, median travel distance for patients from urban ZIP codes (4.8 mi) was lower compared to metropolitan, micropolitan, small town and all other RUCA categories. Conclusions: Preliminary findings indicate a significant association between RUCA and travel distance, with longer travel distances for patients in non-urban locales. Proximity of cancer care sites to the patient’s home reduces distance traveled, minimizing the burden on the patients and their families. Interventions addressing transportation needs (such as improving infrastructure, increasing transportation options, subsidizing healthcare travel, etc.) can be augmented with efforts to strengthen access to community-based sites (expansion of telehealth, incentivizing providers serving non-urban areas, etc.) to reduce patient and caregiver burden. Clinic Site Category % of patients traveling <5 mi <10mi <15mi <50mi >50mi A Suburban Metro 53.8% 30.7% 5.2% 9.1% 0.9% B Suburban Small Town 24.1% 40.6% 21.9% 12.3% 0.4% C Suburban Metro 45.7% 34.3% 5.01% 13.8% 1.0% D Suburban Metro 60.1% 23.9% 6.8% 7.9% 1.2% E Urban Metro 80.3% 9.7% 3.5% 5.2% 1.0% ALL SITES 54.7% 27.1% 7.2% 9.8% 0.9% Clinic Median travel distance in miles (% of patients) by patient RUCA within 50 miles of clinics Urban Metropolitan – High Commute Micropolitan – Flow to Large Urban Cluster Small Town – Flow to Small Urban Cluster All Other A 4.8 (98.2%) 27.5 (1.7%) 48.7 49.4 B 4.0 (42.3%) 8.6 (54.2%) 25.0 20.6 (3.2%) C 5.9 (95.2%) 17.7 (4.5%) 36.7 27.7 D 4.8 (97.1%) 23.7 (2.8%) 48.4 40.1 E 2.7 (98.9%) 32.3 (1.0%) 4.5 ALL 4.8 (91.2%) 10.8 (8.3%) 36.7 (0.1%) 20.6 (0.4%) 4.5

A nationwide double-blind phase III randomized clinical trial (RCT) of olanzapine vs. prochlorperazine for the treatment of refractory nausea in patients receiving moderately or highly emetogenic chemotherapy among NCI Community Oncology Research Program (NCORP) practices.

185 Background: Despite advances in antiemetics given during chemotherapy, nausea remains a clinically significant issue and greatly impairs quality of life (QOL). Current ASCO guidelines recommend olanzapine or prochlorperazine for refractory chemotherapy-induced nausea. However, there is a lack of direct empirical comparisons to establish their relative efficacy. Methods: The URCC NCORP Research Base and 21 NCORP practices enrolled 1,363 chemotherapy-naïve patients with breast cancer starting a high/moderate emetogenic chemotherapy regimen to this Phase III RCT (NCT03367572). Antiemetics were prescribed per ASCO guidelines. Screening occurred during Cycle 1, followed by randomization for Cycle 2. Those with ≥ moderate nausea (≥3 on a 1-7 scale) in Cycle 1 and willing to continue were randomized (1:1:1 ratio) into 3 arms (n=310): 1) ASCO regimen plus olanzapine (OLZ), 2) ASCO regimen plus prochlorperazine (PC), or 3) ASCO regimen plus placebo. Nausea was evaluated using a four-day home diary (4x daily), with the primary outcomes being changes in average (avg) and maximum (max) nausea. FACT-G measured QOL. Intent-to-treat primary analyses (ANOVA and Cohen’s Δ effect size) evaluated whether OLZ or PC improved nausea, testing between-arm differences in nausea change from Cycle 1 to 2 compared to placebo at a p=0.025 significance level corrected for 3 arms. Results: The median age was 50.7y, with 80.7% White/12.3% Black/5.8% Asian/4.2% Latino, and evenly distributed across groups. The change in avg nausea score was statistically significant for the OLZ (avg change = −1.07: Cohen’s Δ = 0.49) and PC arms (avg change = −0.94: Cohen’s Δ = 0.38) compared to the placebo (avg change = −0.50; p <0.001 vs OLZ, p = 0.01 vs PC). Similar but larger changes were noted for change in max nausea score in the OLZ (max change = −2.57: Cohen’s Δ = 0.86) and PC arms (max change = −2.01: Cohen’s Δ = 0.47) against the placebo (max change = −1.30; p <0.001 vs OLZ, p <0.01 vs PC). The OLZ arm had a clinically significant change in overall QOL from pre-chemo to Cycle 2 vs placebo (FACT-G mean difference = +4.91; p = 0.006) but the PC arm vs placebo did not (FACT-G mean difference = −0.87; p = 0.61). OLZ exhibited superiority over PC in terms of changes in max nausea (mean difference = −0.56; p = 0.023) and QOL (FACT-G mean difference = +5.80; p = 0.006), but not for avg nausea (mean difference = −0.13; p = 0.418). Conclusions: OLZ and PC, when added to ASCO guidelines, each significantly reduces refractory nausea. OLZ demonstrates superior max nausea control with clinically and statistically significant improvements in QOL. This study fills the gap in empirical comparisons and solidifies OLZ's position as a promising intervention for refractory nausea. Clinical trial information: NCT03367572 .

Oncology care model cost data for breast, lung, and multiple myeloma: Influence of outpatient rehabilitation therapy services.

14 Background: High value cancer care is defined by access to coordinated, appropriate healthcare services that achieve optimal outcomes at a reasonable cost. Evidence supports outpatient cancer rehabilitation (physical, occupational or speech therapy) as a mechanism to improve outcomes such as health-related quality of life, suggesting that integration of these services is beneficial for cancer survivors. However, the impact of outpatient rehabilitation services on healthcare costs is unknown. We aimed to compare cost segments between those who attended community-based outpatient rehabilitation therapy services (Rehab) versus those who did not (Non-rehab). We hypothesized that healthcare costs would be lower for the Rehab group in the following cost segments: emergency department (ED), part D drugs, home health services, skilled nursing facility services (SNF), inpatient rehabilitation facilities (IRF), and hospice. Methods: This is a retrospective cohort study of Medicare beneficiaries who received chemotherapy at a large community oncology practice that participated in the oncology care model (OCM) program from 2018 to 2021. Two data sources were linked, OCM cost data and the rehabilitation medical record. This yielded sufficient sample sizes for analysis of breast, lung, or multiple myeloma (MM) cancer types. Rehab cases were then matched to Non-rehab cases 3:1 based on cancer type, stage, and sex. Cost variables included total cost of care (TCOC) and cost segments using OCM methodology. Mann-Whitney-Wilcoxon statistic (p<.05) was used for analysis. Results: Cases (N=3,033) primarily had early-stage cancer (0 to 2, 74.8%) and were 70.4±7.3 (Rehab) or 71.6±8.5 years old at diagnosis (Non-rehab). Most had breast cancer (81.2%), followed by lung cancer (11.7%) and MM (7.2%). TCOC was $32,387 and $26,257 for the Rehab and Non-rehab groups, respectively. Costs were significantly lower for the Rehab group in five of the six segments analyzed; IRF costs were lower, but not statistically significant (Table). Conclusions: To our knowledge, this is the first study to explore the influence of real-world, outpatient cancer rehabilitation therapy services on the cost of cancer care. Community-based therapy services were associated with significantly lower costs in ED, part D drugs, home health, SNF, and hospice segments. Future studies examining rehabilitation services as a relatively low-cost supportive care option to meet patient needs and enhance high-value care are warranted. Healthcare segment costs for rehab vs. non-rehab cases. Rehab Non-rehab Between-group Difference Mean, (SD) Mean (SD) p-value ED $85 (260) $104 (372) .019 Part D drugs $4,813 (18,221) $5,675 (18,797) <.001 Home health $433 (1,600) $705 (2,118) <.001 SNF $160 (1,988) $386 (3,943) .046 Hospice $75 (770) $198 (1,529) <.001 IRF $245 (2,275) $384 (3,277) .434

Baseline characteristics of parents enrolled in ACCL20N1CD: Financial distress during treatment for pediatric acute lymphoblastic leukemia (ALL) in the US.

220 Background: Personal costs of childhood cancer can contribute to disparities in health outcomes. The purpose of ACCL20N1CD is to examine trajectories of parental costs and the associated financial distress, cost coping behaviors, and household material hardships (HMH) during treatment for pediatric ALL. Objective is to report baseline characteristics of the parents and their households. Methods: The study used a prospective cohort design with repeated measures. Setting was Children’s Oncology Group (COG) practices at National Cancer Institute Community Oncology Research Program (NCORP) institutions. Parents who use English or Spanish, and have children newly diagnosed with ALL treated at COG NCORP practices were eligible. Parents were asked to complete surveys at 3 time points, including baseline (T1; during induction and <7 days after planned induction remission evaluation). Surveys included items about (a) parent socio‐economic status, financial distress measured by 8-item Personal Wellness Scale (PFWS), and cost coping behaviors, and (b) HMH. Factors related to high PFWS scores were evaluated by logistic regression. Results: Of 104 evaluable parents, most were female (83.7%), white (65.4%), non-Hispanic (48.1%), partnered (79.8%), > high school diploma (93.3%), used English (84.6%), worked full (48.1%) or part (5.8%) time, and paid for their child’s care with public insurance (57.7%). Median household income=$50,000 (Inter-Quartile Range/IQR 20,000-100,000) for 4 (IQR 4-5) people. Not including wages lost due to caregiving, out of pocket payments for medical care + other ALL-related costs in the prior month >15% of monthly income for 51.9% of households, which is considered financially catastrophic. 24.0% of households lacked food stability; 21% had housing insecurity. Conclusions: At baseline, sizeable proportions of parents exhibit high out of pocket costs, high financial distress level, and multiple cost coping behaviors with implications for future health and individual and household quality of life.

Baseline practice-level financial screening processes and resources among Children’s Oncology Group National Cancer Institute Community Oncology Research Programs (NCORP).

34 Background: Financial toxicity is an increasingly recognized and prevalent adverse effect during cancer treatment. Systematic financial screening is recommended for families of children with cancer, and emerging data support the importance of system-level support and interventions to mitigate financial toxicity. Financial distress is a key, measurable domain of financial toxicity. Methods: To inform interventions for pediatric oncology settings, we are conducting ACCL20N1CD, a Children’s Oncology Group (COG) prospective, multilevel observational study among NCORP practices in the US, to determine the risk and trajectory for financial distress in guardians of children with newly diagnosed acute lymphoblastic leukemia. We analyzed baseline survey data from COG NCORP practices that enrolled at least one parent. Data were collected via REDCap following the practice’s first parent enrollment. We describe practice participation patterns, current approaches to financial screening, and available resources to address identified needs including navigation. Results: Of 50 eligible COG NCORP practices, 40 (80%) activated the study, and 28 (56%) enrolled at least 1 participant and thus, completed the survey. Most (61%) of these practices treated 50-99 new pediatric oncology patients annually; 10 (36%) enrolled at least 5 participants in ACCL20N1CD. Twenty (71%) reported that their practice routinely conducted financial screening in addition to insurance assessment; 15 (75%) of these 20 practices conducted financial screening at more than one timepoint. Social workers were primarily responsible for screening (16/20; 80%). Only three practices (3/28; 11%) used a standardized financial screening tool. Most practices had access to a private foundation/organization (25/28, 89%), institutional support (23/28, 82%), or local referrals to organizations (22/28, 78%) that provided financial support for patients/families. Fewer practices had patient navigators (9/28, 32%), financial navigators (8/28, 28%), or legal advocates (3/28, 11%) available to assist families. Conclusions: Though most COG practices treating children of parents enrolled in 20N1CD conduct financial screening, processes are not standardized, and resources available to address identified needs and potentially mitigate financial toxicity vary. Further, these data represent enrolling sites which may overestimate screening practices within COG NCORP practices. Variations in financial screening procedures and lack of individual-level resources to facilitate screening may impact delivery of equitable high-quality cancer care across NCORP COG practices. Clinical trial information: NCT04928599 .

Improving Surgical Care and Outcomes in Older Cancer Patients Through Implementation of a Presurgical Toolkit (OPTI-Surg)-Final Results of a Phase III Cluster Randomized Trial (Alliance A231601CD).

To assess the effect of a practice-level preoperative frailty screening and optimization toolkit (OPTI-Surg) on postoperative functional recovery and complications in elderly cancer patients undergoing major surgery. Frailty is common in older adults. It increases the risk of poor postoperative functional recovery and complications. The potential for a practice-level screening/optimization intervention to improve outcomes is unknown. Thoracic, gastrointestinal, and urologic oncological surgery practices within the National Cancer Institute Community Oncology Research Program (NCORP) were randomized 1:1:1 to usual care (UC), OPTI-Surg, or OPTI-Surg with an implementation coach. OPTI-Surg consisted of the Edmonton Frail Scale and guided recommendations for referral interventions. Patients 70 years old or above undergoing curative intent surgery were eligible. The primary outcome was 8 weeks postoperative function (kcal/wk). The key secondary outcome was complications within 90 days. Mixed models were used to compare UC to the 2 OPTI-Surg arms combined. From July 2019 to September 2022, 325 patients were enrolled in 29 practices. One hundred ninety-nine (64 UC, 135 OPTI-Surg) and 279 (78 UC, 201 OPTI-Surg) were evaluable for primary and secondary analysis, respectively. UC and OPTI-Surg patients did not significantly differ in total caloric expenditure (2.2 UC, 2.0 OPTI-Surg) after adjusting for baseline function ( P =0.53). UC and OPTI-Surg patients did not significantly differ in postoperative complications (25.6% UC, 35.3% OPTI-Surg, P =0.5). Frailty assessment was successfully performed, but the OPTI-Surg intervention did not improve postoperative function nor reduce postoperative complications compared with UC. Future analysis will explore practice-level factors associated with toolkit implementation and the differences between the coaching and noncoaching arms.

Clinician engagement strategies to increase palliative care access in a multi-site community oncology practice.

196 Background: Despite nearly 20 years since PC was recognized as a medical subspecialty, and ACS, ASCO and NCCN’s recommendations for early involvement, many oncologists continue to categorize PC synonymously with hospice and therefore refer to palliative medicine specialists very late in the disease trajectory, if at all. Each year, more oncology and PC researchers study and report the life changing impact of early PC for patients living with stage III and IV “advanced cancer” (AC). While mandating PC would have its challenges, creating workflows to encourage providers to change their behavior and attitude regarding PC, while empowering other clinical team members to directly refer, should result in patients and families receiving earlier PC services. Methods: The study included a pre-intervention period (9/1/2023 - 12/31/2023), and an intervention period (1/1/2024 - 4/30/2024). Pre-intervention referral data was collected from OncoEMR, OneAnalytics and Social Work reports. PC referrals were requested for both oncology and hematology patients, regardless of diagnosis date or stage. During the intervention period, 6 strategies were designed to increase PC referrals and were deployed equally in all 8 oncology/hematology offices. Strategies included 1) weekly MD/APP huddles with one of the goals to discuss the palliative care needs of all patients, 2) self-directed referrals from nursing, 3) self-directed referrals from social work, 4) APP referrals for advance care planning, 5) proactive chart reviews by PC APNs and direct PC referral suggestions to MD/APPs for AC patients, and 6) hardwiring timely scheduling of PC referrals. The primary end point is the percent of unique cancer (UC) patients who had a PC referral; the secondary end point is the percent of referred AC patients with early PC referral as defined by referral within 8 weeks of diagnosis. Results: During the pre-intervention period, of the 8,551 UC patients, 154 (1.8%) received PC referrals as compared to 227 (2.8%) of the 8,055 UC patients during the intervention period, yielding a 56% increase in the intervention period. In the pre-intervention period, 24 (14%) of the 177 newly diagnosed AC patients received early PC referrals compared to 32 (21%) of the 153 newly diagnosed AC patients in the intervention period, giving rise to a 50% increase. Conclusions: Strategies to increase PC referrals in a community oncology practice increased overall referrals as well as earlier referrals in newly diagnosed AC patients. As the interventions mature, the total number of PC referrals and a shift toward earlier referrals will likely be seen. Future goals will include assessing patient and family comfort with their treatment plan with PC versus without PC, as well as assessing providers' perceived value of PC over time. With attitude and behavior change, practice culture change is anticipated amid PC becoming the standard of care for all AC patients.

Development and implementation of a digital remote symptom monitoring program (RESPONSe) to support patients undergoing IV chemotherapy at a community ambulatory cancer clinic in Richmond, BC, Canada.

406 Background: We identified delays in accessing symptom management during treatment as an existing care gap at the Richmond Cancer Clinic. Patient-reported outcome measures (PROMs) have been demonstrated to communication, quality of life, and reduce symptom burden and acute care utilization. The RESPONSe program was initiated to address this gap by implementing a digital remote symptom monitoring (RSM) system paired with a supporting Symptom Management Nurse (SMN). Methods: We designed a digital RSM system to collect PROMs at specific timepoints during the first 6 months of IV chemotherapy initiation, with automated triaging logic to flag symptoms of concern that would result in a phone call from the SMN. Routine surveys include baseline, health-checks on days 3 and 10 post-chemotherapy, and pre-doctor’s-appointment; with optional health checks available anytime. Through competitive procurement, SeamlessMD was selected as the technology partner. From Aug 2023 to Jan 2024, multiple design cycles with dry run testing were held to incorporate branching logic, COSTARS oncology symptom specific questionnaires, automatic triaging flags, immediate self—management tips and links to educational resources. Concurrent development of SMN clinical workflows included in-person patient enrolment, digital dashboard monitoring and integration with the clinical chart. Results: From Jan 22-Apr 30, 2024, 32 patients have been enrolled with 42% over the age of 70. 90% of patients completed the baseline survey, 79% of patients completed at least 1 post-treatment health check and 45% completed at least 1 pre-doctors’-appointment survey. An average of 22 library resources were accessed per patient. Through the system, 19 severe and 34 moderate symptoms were identified. The dashboard prompted 63 telephone calls from the SMN, with a further 28 follow-up calls required. Early 3-month patient experience surveys (n=8) indicate high satisfaction, with 100% of patients saying they would recommend the program to others undergoing treatment. 87.5% of patients felt the system was very helpful in communicating their issues to the clinical team, with 100% reporting very high/extreme satisfaction with the support provided by the SMN. Additionally, 67% of patients felt very satisfied with the RESPONSe program overall. Conclusions: The RESPONSe program demonstrates the potential of digital RSM systems in improving patient-centered care and assisting with symptom reporting and management. Initial results suggest high engagement and patient satisfaction, showing this is a feasible approach in community oncology settings. We will be reviewing acute care utilization to assess the impact of the RESPONSe program on emergency room visits and hospitalization.

P4.17F.01 Effective EGFR Testing Rates in Advanced Stage NSCLC Patients Treated in Community Oncology Practices: Impact of a Decision Support Program

P4.17F.01 Effective EGFR Testing Rates in Advanced Stage NSCLC Patients Treated in Community Oncology Practices: Impact of a Decision Support Program

EP.12A.30 Landscape of EGFR Mutations and Real-World Experience of Therapeutic Interventions - A Community Oncology Study

EP.12A.30 Landscape of EGFR Mutations and Real-World Experience of Therapeutic Interventions - A Community Oncology Study

Quality improvement scorecards driving innovation and best practice in a large community oncology practice.

417 Background: Lean Six Sigma Visual Management dashboards, or scorecards, are visual tools designed to drive and improve quality patient care, operationalize best practices, and promote a culture of operational excellence. Practice scorecards create a disciplined problem-solving culture where root causes are identified, and systemic process-based improvements are implemented. Scorecards are created to show full practice data transparency, allowing all employees and sites to see each other’s data. Scorecards are designed so that providers can efficiently review the maximum amount of data, following a Green, Yellow, Red visual cue to signify goal adherence. Methods: Phase 1 (2018 – 2019): Clinical scorecards were created and introduced to all staff on a weekly basis focusing on a few key patient metrics (ex: pain scale, pain care plan, ACP, PHQ-9, teaching/consents, patient portal access). Phase 2 (2020 – 2022): Clinical scorecards were expanded with additional patient care initiatives. New nurse navigator scorecards created (ex: patient teaching and survivorship visits). New physician and APP scorecards created with practice-specific provider metrics. Phase 3 (2023 – 2024): Scorecards updated yearly to match practice quality initiatives. New dashboards for non-clinical employees initiated. Physician and APP compensation changed to include quality metric performance tied to financial incentive. Results: Over the course of 5 years, OHC quality scorecard metrics improved from below national averages (red) to exceeding top-tier threshold goals (green). Scorecard utilization increased across all areas of the business, expanding from one scorecard in 2018 to eight in 2024. Scorecards increased overall practice data transparency leading to increase practice quality metric performance. OHC staff utilize scorecards on a weekly, monthly, and quarterly basis to drive quality improvement adherence. Improved outcomes through full data transparency increased staff participation and ownership in quality metrics and led to achieved shared savings in managed care programs. Example of metrics and results: Patient NCCN Staging 36% (2588/7263) to 96% (10299/10749), Teaching/Consents 37% (423/1147) to 97% (3067/3160), and Oral regimen NCCN entry 69% (586/845) to 98% (1037/1055). Conclusions: Visual Management dashboards or practice scorecards are instrumental in driving and supporting practice change. Scorecards allowed OHC staff to efficiently visualize practice-wide actionable data with eye-catching colors to promote optimal clinical performance. As increased data procurement is on the rise, practices can empower their staff to drive change through data utilization. OHC’s scorecards have proven to bridge the gap to goal, facilitate transparent data sharing, and empower providers and staff toward collaboration and embracing operational excellence.

Financial needs and opioid use among head and neck cancer (HNC) survivors: Baseline findings from the HN-STAR trial (WF-1805CD).

365 Background: HNC survivors often experience post-treatment pain. Given known disparities in pain management and opioid use in the non-cancer population, we explored sociodemographic predictors of opioid use in HNC survivors enrolled in the HN-STAR trial (NCT04208490, funding: American Cancer Society and UG1CA189824). Methods: Survivors of HNC (stages I-III) who completed therapy 6-24 months prior and had no evidence of disease were enrolled throughout the NCI Community Oncology Research Program (NCORP). Survivors were randomized to a clinical informatics intervention or usual care. Baseline surveys elicited recent prescription opioid use in the prior week (yes/no). Blinded to arm assignment, we conducted bivariate analyses of recent opioid use and the following sociodemographic predictors: age, gender, marital status, race and ethnicity (non-white/Latinx vs all others), highest level of education, and being able to meet daily financial needs in the prior 4 weeks. We conducted stepwise multivariable logistic regression to evaluate the association of these factors on opioid use. Results: Among 254 survivors (mean age 63, 74% male, 25% non-white/non-Latinx, 13% reporting financial need) at 25 oncology practices, 14% (n=35) reported recent opioid use. In bivariate analyses, survivors who reported financial needs were more likely to have used opioids (odds ratio [OR] 4.35, 95% confidence interval [CI] 1.85-10.20, p<.001). In stepwise multivariable analyses, only having financial needs remained in the final model (Table). Conclusions: In a national sample of post-treatment HNC survivors during a subacute survivorship period (6-24 months after treatment completion), 1 in 7 survivors reported using opioids in the past week – more than twice the annual rate of opioid receipt in the general adult population. Several demographic factors related to opioid use were examined and not associated with recent opioid use among HNC survivors. Findings suggest that financial concerns may be a barrier to non-narcotic pain management (e.g., cognitive behavioral approaches, physical therapy) – or simply that unmet financial needs are associated with lower quality of pain management. Attention should focus on equitable pain management in this population. Clinical trial information: NCT04208490 . Predictors of opioid use in past week. Bivariate Models Estimate (standard error) p-value Age (years) -.01 (.02) .52 OR (95% CI) p-value Female .98 (.41, 2.16) .97 Latinx/non-white .71 (.27, 1.63) .45 Married .56 (.27, 1.18) .13 Some college or more .62 (.30, 1.30) .20 Had financial needs 4.35 (1.85,10.20) <.001 Final model OR (95% CI) p-value Had financial needs 4.35 (1.85,10.20) <.001

Intersectionality Between Country, Gender and Funding in Authorship for Phase III Trials Presented at the ASCO Annual Meeting 2022.

Multiple disparities have been recognized in the area of location, gender, and funding for leadership in oncology clinical trials. Understanding their intersectionality is crucial to be able to formulate policies and actions, to ensure research is representative of the global oncology community. Here, data from phase III trials presented at the ASCO Annual Meeting of 2022 (ASCO22) were analyzed. The location of institution, gender of lead and senior authors, and funding source for solid tumor phase III trial abstracts presented at the ASCO22 were analyzed. World Bank analytical grouping version 2021-2022 was used to describe regions and countries as high (HIC), upper-middle (UMIC), lower-middle (LoMIC), and low-income (LIC). Across 239 phase III abstracts, lead and senior authors respectively represented HIC institutions in 83% and 85%, UMIC in 13% and 12%, and LoMIC in 4% and 3%. No authors worked in LICs or sub-Saharan Africa. Women accounted for 29% of lead and 23% of senior authors. This distribution persisted across regions, with women as lead authors ranging from 19% (UMIC) to 31% (HIC), and as senior authors from 7% (UMIC) to 25% (HIC). Industry funded 62% of trials, academia 17%, and others 15%; 6% lacked funding. Industry funding was highest in HIC trials (66% for lead and senior authors), followed by UMICs (55% lead, 53% senior) and LoMICs (11% lead, 0% senior). Industry-sponsored trials were proportionally equally represented among female and male senior authors (63% each). There is marked intersectionality in leadership of oncology clinical trials presented at the world's largest oncology conference.

MHealth Apps in the Digital Marketplace for Pediatric Patients With Cancer: Systematic Search and Analysis.

The substantial increase in smartphone ownership has led to a rise in mobile health (mHealth) app use. Developing tailored features through mHealth apps creates a pathway to address the health care needs of pediatric patients with cancer and their families who have complex care needs. However, few apps are designed specifically to integrate with pediatric cancer care. This study reports a systematic search and analysis of mHealth apps available on the Apple App (iOS) and Google Play (Android) stores designed for pediatric cancer through a list of features that serve (1) patients, (2) caregivers, or (3) both audiences. Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we reviewed apps for pediatric patients with cancer and caregivers available as of January 30, 2024. We searched the Apple App and Google Play stores with a list of keyword combinations focusing on pediatric cancer care. The inclusion criteria were (1) specifically apps targeted toward pediatric patients with cancer, their families, or both; (2) available in either app store; and (3) available in English. Apps were assessed using the Mobile Application Rating Scale (MARS). The MARS is a quality assessment for mHealth apps, including components of engagement, functionality, aesthetics, and informational quality (5-point Likert scale items-1: low and 5: high quality). In total, 22 apps were identified and 17 of those apps were available on both platforms. The most popular features (n=12) were resource sharing, symptom tracking, reminders, care team connections, journaling, community support, medication tracking, data visualizations, and appointment tracking. Features and interfaces were designed for caregivers (n=9) more frequently than the patients (n=7) while a subset of apps created options for both users (n=6). A total of 16 apps received positive reviews (mean 4.4, SD 0.59; Min=3.1, Max=5.0). A small subset (n=3) achieved over 5000 downloads; however, the majority (n=15) had fewer than 500. More than half (n=12) of the apps were not available in English. Apps requested access to a range of device functionalities to operate (mean 2.72, SD 3.13; Min=0, Max=10). Out of 22, a total of 17 apps were publicly accessible. The mean MARS scores for the apps ranged from 1.71 (SD 0.75) to 4.33 (SD 0.82). Overall, apps scored high on functionality (mean 3.72, SD 0.54) but low on engagement (mean 3.02, SD 0.93). Our review highlights the promising yet underdeveloped potential of mHealth apps in pediatric oncology care, underscoring the need for more inclusive, comprehensive, and integrative digital health solutions. Future developments should actively involve key stakeholders from the pediatric oncology community, including patients, families, and health care professionals, to ensure the apps meet specific needs while addressing linguistic and cultural barriers.

Evidenced-Based Prior for Estimating the Treatment Effect of Phase III Randomized Trials in Oncology.

The primary results of phase III oncology trials may be challenging to interpret, given that results are generally based on P value thresholds. The probability of whether a treatment is beneficial, although more intuitive, is not usually provided. Here, we developed and released a user-friendly tool that calculates the probability of treatment benefit using trial summary statistics. We curated 415 phase III randomized trials enrolling 338,600 patients published between 2004 and 2020. A phase III prior probability distribution for the treatment effect was developed on the basis of a three-component zero-mean mixture distribution of the observed z-scores. Using this prior, we computed the probability of clinically meaningful benefit (hazard ratio [HR] <0.8). The distribution of signal-to-noise ratios and power of phase III oncology trials were compared with that of 23,551 randomized trials from the Cochrane Database. The signal-to-noise ratios of phase III oncology trials tended to be much larger than randomized trials from the Cochrane Database. Still, the median power of phase III oncology trials was only 49% (IQR, 14%-95%), and the power was <80% in 65% of trials. Using the phase III oncology-specific prior, only 53% of trials claiming superiority (114 of 216) had a ≥90% probability of clinically meaningful benefits. Conversely, the probability that the experimental arm was superior to the control arm (HR <1) exceeded 90% in 17% of trials interpreted as having no benefit (34 of 199). By enabling computation of contextual probabilities for the treatment effect from summary statistics, our robust, highly practical tool, now posted on a user-friendly webpage, can aid the wider oncology community in the interpretation of phase III trials.