Association of Remote Patient-Reported Outcomes and Step Counts With Hospitalization or Death Among Patients With Advanced Cancer Undergoing Chemotherapy: Secondary Analysis of the PROStep Randomized Trial

BACKGROUND

Patients with advanced cancer undergoing chemotherapy experience significant symptoms and declines in functional status, which are associated with poor outcomes. Remote monitoring of patient-reported outcomes (PROs; symptoms) and step counts (functional status) may proactively identify patients at risk of hospitalization or death.

OBJECTIVE

The aim of this study is to evaluate the association of (1) longitudinal PROs with step counts and (2) PROs and step counts with hospitalization or death.

METHODS

The PROStep randomized trial enrolled 108 patients with advanced gastrointestinal or lung cancers undergoing cytotoxic chemotherapy at a large academic cancer center. Patients were randomized to weekly text-based monitoring of 8 PROs plus continuous step count monitoring via Fitbit (Google) versus usual care. This preplanned secondary analysis included 57 of 75 patients randomized to the intervention who had PRO and step count data. We analyzed the associations between PROs and mean daily step counts and the associations of PROs and step counts with the composite outcome of hospitalization or death using bootstrapped generalized linear models to account for longitudinal data.

RESULTS

Among 57 patients, the mean age was 57 (SD 10.9) years, 24 (42%) were female, 43 (75%) had advanced gastrointestinal cancer, 14 (25%) had advanced lung cancer, and 25 (44%) were hospitalized or died during follow-up. A 1-point weekly increase (on a 32-point scale) in aggregate PRO score was associated with 247 fewer mean daily steps (95% CI -277 to -213; P<.001). PROs most strongly associated with step count decline were patient-reported activity (daily step change -892), nausea score (-677), and constipation score (524). A 1-point weekly increase in aggregate PRO score was associated with 20% greater odds of hospitalization or death (adjusted odds ratio [aOR] 1.2, 95% CI 1.1-1.4; P=.01). PROs most strongly associated with hospitalization or death were pain (aOR 3.2, 95% CI 1.6-6.5; P<.001), decreased activity (aOR 3.2, 95% CI 1.4-7.1; P=.01), dyspnea (aOR 2.6, 95% CI 1.2-5.5; P=.02), and sadness (aOR 2.1, 95% CI 1.1-4.3; P=.03). A decrease in 1000 steps was associated with 16% greater odds of hospitalization or death (aOR 1.2, 95% CI 1.0-1.3; P=.03). Compared with baseline, mean daily step count decreased 7% (n=274 steps), 9% (n=351 steps), and 16% (n=667 steps) in the 3, 2, and 1 weeks before hospitalization or death, respectively.

CONCLUSIONS

In this secondary analysis of a randomized trial among patients with advanced cancer, higher symptom burden and decreased step count were independently associated with and predictably worsened close to hospitalization or death. Future interventions should leverage longitudinal PRO and step count data to target interventions toward patients at risk for poor outcomes.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04616768; https://clinicaltrials.gov/study/NCT04616768.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.1136/bmjopen-2021-054675.

Association between oral targeted cancer drug net health benefit, uptake, and spending

BACKGROUND

Targeted cancer drugs (TCDs) have revolutionized oncology but vary in clinical benefit and patient out-out-pocket (OOP) costs. The ASCO Value Framework uses survival, toxicity, and symptom palliation data to quantify the net health benefit (NHB) of cancer drugs. We evaluated associations between NHB, uptake, and spending on oral TCDs.

METHODS

We conducted a retrospective cohort study of patients aged 18-64 years with an incident oral TCD pharmacy claim in 2012-2020 in a nationwide de-identified commercial claims dataset. TCDs were categorized as having high (>60), medium (40-60), and low (<40) NHB scores. We plotted the uptake of TCDs by NHB category and used standard descriptive statistics to evaluate patient OOP and total spending. Generalized linear models evaluated the relationship between spending and TCD NHB, adjusted for cancer indication.

RESULTS

We included 8,524 patients with incident claims for eight oral TCDs with nine first-line indications in advanced melanoma, breast, lung, and pancreatic cancer. Medium- and high-NHB TCDs accounted for most TCD prescriptions. Median OOP spending was $18.78 for the first 28-day TCD supply (IQR $0.00-$87.57); 45% of patients paid $0 OOP. Median total spending was $10,118.79 (IQR $6,365.95-$10,600.37) for an incident 28-day TCD supply. Total spending increased $1,083.56 for each 10-point increase in NHB score (95% CI $1,050.27-$1,116.84, p < .01 for H0=$0).

CONCLUSION

Low-NHB TCDs were prescribed less frequently than medium- and high-NHB TCDs. Total spending on oral TCDs was high and positively associated with NHB. Commercially insured patients were largely shielded from high OOP spending on oral TCDs.

Validation of Patient-Reported Outcomes in Patients With Nonmetastatic Breast Cancer Receiving Comprehensive Nodal Irradiation in the RadComp Trial

PURPOSE

Our purpose was to evaluate the measurement properties of patient-reported outcome (PRO) measures used in the ongoing RadComp pragmatic randomized clinical trial (PRCT).

METHODS AND MATERIALS

The deidentified and blinded data set included 774 English-speaking female participants who completed their 6-month posttreatment assessment. Eleven PRO measures were evaluated, including the Trial Outcome Index from the Functional Assessment of Cancer Therapy-Breast (FACT-B), Satisfaction with Breast Cosmetic Outcomes, the BREAST-Q, and selected Patient-Reported Outcomes Measurement Information System (PROMIS) measures. PROs were measured at 3 timepoints: baseline, completion of radiation therapy (RT), and 6 months post-RT. Ten variables were used as validity anchors. Pearson or Spearman correlations were calculated between PROs and convergent validity indicators. Mean PRO differences between clinically distinct categories were compared with analysis of variance methods (known-groups validity). PRO change scores were mapped to change in other variables (sensitivity to change).

RESULTS

Most correlations between PROs and validity indicators were large (≥0.5). Mean score for Satisfaction with Breast Cosmetic Outcomes was higher (better) for those with a lumpectomy compared with those with a mastectomy (P < .001). Mean scores for the FACT-B Trial Outcome Index and for PROMIS Fatigue and Ability to Participate in Social Roles and Activities were better for those with good baseline performance status compared with those with poorer baseline performance status (P < .05). At completion of RT and post-RT, mean scores for Satisfaction with Breast Cosmetic Outcomes and BREAST-Q Radiation were significantly different (P < .001) across categories for all Functional Assessment of Chronic Illness Therapy -Treatment Satisfaction - General items. There were medium-sized correlations between change scores for FACT-B Trial Outcome Index, Fatigue, Anxiety, and Ability to Participate in Social Roles and change scores in the Visual Analog Scale.

CONCLUSIONS

For patients with nonmetastatic breast cancer receiving radiation in the RadComp PRCT, our findings demonstrate high reliability and validity for important PRO measures, supporting their psychometric strength and usefulness to reflect the effect of RT on health-related quality of life.

Cancer Treatment Before and After Physician-Pharmacy Integration

Importance

Integration of pharmacies with physician practices, also known as medically integrated dispensing, is increasing in oncology. However, little is known about how this integration affects drug use, expenditures, medication adherence, or time to treatment initiation.

Objective

To examine the association of physician-pharmacy integration with oral oncology drug expenditures, use, and patient-centered measures.

Design, Setting, and Participants

This cohort study used claims data from a large commercial insurer in the US to analyze changes in outcome measures among patients treated by pharmacy-integrating vs nonintegrating community oncologists in 14 states between January 1, 2011, and December 31, 2019. Commercially insured patients were aged 18 to 64 years with 1 of the following advanced-stage diagnoses: breast cancer, colorectal cancer, kidney cancer, lung cancer, melanoma, or prostate cancer. Data analysis was conducted from May 2023 to March 2024.

Exposure

Treatment by a pharmacy-integrating oncologist, ascertained by the presence of an on-site pharmacy or nonpharmacy dispensing site.

Main Outcomes and Measures

Oral, intravenous (IV), total, and out-of-pocket drug expenditures for a 6-month episode of care; share of patients prescribed oral drugs; days' supply of oral drugs; medication adherence measured by proportion of days covered; and time to treatment initiation. The association between an oncologist's pharmacy integration and each outcome of interest was estimated using the difference-in-differences estimator.

Results

Between 2012 and 2019, 3159 oncologists (745 females [27.1%], 2002 males [72.9%]) treated 23 968 patients (66.4% female; 53.4% aged 55-64 years). Of the 3159 oncologists, 578 (18.3%) worked in practices that integrated with pharmacies (with a low rate in 2011 of 0% and a high rate in 2019 of 31.5%). In the full sample (including all cancer sites), after physician-pharmacy integration, no significant changes were found in oral drug expenditures, IV drug expenditures, or total drug expenditures. There was, however, an increase in days' supply of oral drugs (5.96 days; 95% CI, 0.64-11.28 days; P = .001). There were no significant changes in out-of-pocket expenditures, medication adherence, or time to treatment initiation of oral drugs. In the breast cancer sample, there was an increase in oral drug expenditures ($244; 95% CI, $41-$446; P = .02) and a decrease in IV drug expenditures (-$4187; 95% CI, -$8293 to -$80; P = .05).

Conclusions and Relevance

Results of this cohort study indicated that the integration of oncology practices with pharmacies was not associated with significant changes in expenditures or clear patient-centered benefits.

Trends in low-value cancer care during the COVID-19 pandemic

OBJECTIVE

To assess the association between the onset of the COVID-19 pandemic and change in low-value cancer services.

STUDY DESIGN

In this retrospective cohort study, we used administrative claims from the HealthCore Integrated Research Environment, a repository of medical and pharmacy data from US health plans representing more than 80 million members, between January 1, 2016, and March 31, 2021.

METHODS

We used linear probability models to investigate the relation between the onset of the COVID-19 pandemic and 4 guideline-based metrics of low-value cancer care: (1) conventional fractionation radiotherapy instead of hypofractionated radiotherapy for early-stage breast cancer; (2) non-guideline-based antiemetic use for minimal-, low-, or moderate- to high-risk chemotherapies; (3) off-pathway systemic therapy; and (4) aggressive end-of-life care. We identified patients with new diagnoses of breast, colorectal, and/or lung cancer. We excluded members who did not have at least 6 months of continuous insurance coverage and members with prevalent cancers.

RESULTS

Among 117,116 members (median [IQR] age, 60 [53-69] years; 72.4% women), 59,729 (51.0%) had breast cancer, 25,751 (22.0%) had colorectal cancer, and 31,862 (27.2%) had lung cancer. The payer mix was 18.7% Medicare Advantage or Medicare supplemental and 81.2% commercial non-Medicare. Rates of low-value cancer services exhibited minimal changes during the pandemic, as adjusted percentage-point differences were 3.93 (95% CI, 1.50-6.36) for conventional radiotherapy, 0.82 (95% CI, -0.62 to 2.25) for off-pathway systemic therapy, -3.62 (95% CI, -4.97 to -2.27) for non-guideline-based antiemetics, and 2.71 (95% CI, -0.59 to 6.02) for aggressive end-of-life care.

CONCLUSIONS

Low-value cancer care remained prevalent throughout the pandemic. Policy makers should consider changes to payment and incentive design to turn the tide against low-value cancer care.

Protocol for a pragmatic stepped wedge cluster randomized clinical trial testing behavioral economic implementation strategies to increase supplemental breast MRI screening among patients with extremely dense breasts

BACKGROUND

Increased breast density augments breast cancer risk and reduces mammography sensitivity. Supplemental breast MRI screening can significantly increase cancer detection among women with dense breasts. However, few women undergo this exam, and screening is consistently lower among racially minoritized populations. Implementation strategies informed by behavioral economics ("nudges") can promote evidence-based practices by improving clinician decision-making under conditions of uncertainty. Nudges directed toward clinicians and patients may facilitate the implementation of supplemental breast MRI.

METHODS

Approximately 1600 patients identified as having extremely dense breasts after non-actionable mammograms, along with about 1100 clinicians involved with their care at 32 primary care or OB/GYN clinics across a racially diverse academically based health system, will be enrolled. A 2 × 2 randomized pragmatic trial will test nudges to patients, clinicians, both, or neither to promote supplemental breast MRI screening. Before implementation, rapid cycle approaches informed by clinician and patient experiences and behavioral economics and health equity frameworks guided nudge design. Clinicians will be clustered into clinic groups based on existing administrative departments and care patterns, and these clinic groups will be randomized to have the nudge activated at different times per a stepped wedge design. Clinicians will receive nudges integrated into the routine mammographic report or sent through electronic health record (EHR) in-basket messaging once their clinic group (i.e., wedge) is randomized to receive the intervention. Independently, patients will be randomized to receive text message nudges or not. The primary outcome will be defined as ordering or scheduling supplemental breast MRI. Secondary outcomes include MRI completion, cancer detection rates, and false-positive rates. Patient sociodemographic information and clinic-level variables will be examined as moderators of nudge effectiveness. Qualitative interviews conducted at the trial's conclusion will examine barriers and facilitators to implementation.

DISCUSSION

This study will add to the growing literature on the effectiveness of behavioral economics-informed implementation strategies to promote evidence-based interventions. The design will facilitate testing the relative effects of nudges to patients and clinicians and the effects of moderators of nudge effectiveness, including key indicators of health disparities. The results may inform the introduction of low-cost, scalable implementation strategies to promote early breast cancer detection.

TRIAL REGISTRATION

ClinicalTrials.gov NCT05787249. Registered on March 28, 2023.

Protocol to evaluate sequential electronic health record-based strategies to increase genetic testing for breast and ovarian cancer risk across diverse patient populations in gynecology practices

BACKGROUND

Germline genetic testing is recommended by the National Comprehensive Cancer Network (NCCN) for individuals including, but not limited to, those with a personal history of ovarian cancer, young-onset (< 50 years) breast cancer, and a family history of ovarian cancer or male breast cancer. Genetic testing is underused overall, and rates are consistently lower among Black and Hispanic populations. Behavioral economics-informed implementation strategies, or nudges, directed towards patients and clinicians may increase the use of this evidence-based clinical practice.

METHODS

Patients meeting eligibility for germline genetic testing for breast and ovarian cancer will be identified using electronic phenotyping algorithms. A pragmatic cohort study will test three sequential strategies to promote genetic testing, two directed at patients and one directed at clinicians, deployed in the electronic health record (EHR) for patients in OB-GYN clinics across a diverse academic medical center. We will use rapid cycle approaches informed by relevant clinician and patient experiences, health equity, and behavioral economics to optimize and de-risk our strategies and methods before trial initiation. Step 1 will send patients messages through the health system patient portal. For non-responders, step 2 will reach out to patients via text message. For non-responders, Step 3 will contact patients' clinicians using a novel "pend and send" tool in the EHR. The primary implementation outcome is engagement with germline genetic testing for breast and ovarian cancer predisposition, defined as a scheduled genetic counseling appointment. Patient data collected through the EHR (e.g., race/ethnicity, geocoded address) will be examined as moderators of the impact of the strategies.

DISCUSSION

This study will be one of the first to sequentially examine the effects of patient- and clinician-directed strategies informed by behavioral economics on engagement with breast and ovarian cancer genetic testing. The pragmatic and sequential design will facilitate a large and diverse patient sample, allow for the assessment of incremental gains from different implementation strategies, and permit the assessment of moderators of strategy effectiveness. The findings may help determine the impact of low-cost, highly transportable implementation strategies that can be integrated into healthcare systems to improve the use of genomic medicine.

TRIAL REGISTRATION

ClinicalTrials.gov. NCT05721326. Registered February 10, 2023. https://www.

CLINICALTRIALS

gov/study/NCT05721326.

Prostate Advanced Radiation Technologies Investigating Quality of Life (PARTIQoL): A Phase III Randomized Clinical Trial of Proton Therapy vs. IMRT for Low or Intermediate Risk Prostate Cancer

PURPOSE/OBJECTIVE(S)

Prostate cancer is the most common non-cutaneous cancer diagnosed among men in the United States, and the majority of patients are diagnosed with localized disease. Men with localized prostate cancer have several treatment options including external beam radiotherapy with either photons or protons. Proton beam therapy (PBT) has certain dosimetric advantages and the potential to reduce treatment-associated morbidity and improve oncologic outcomes, but current PBT is significantly more costly than intensity-modulated radiotherapy (IMRT). The PARTIQoL trial (NCT01617161) is the first multicenter phase 3 randomized trial comparing protons to photons in the treatment of localized prostate cancer.

MATERIALS/METHODS

Patients with low or intermediate risk prostate cancer (Stage T1c-T2c, PSA < 20, Gleason score ≤ 7) are randomized to receive either PBT or IMRT, with targeted recruitment efforts for minority populations. A companion registry study has concurrently enrolled patients who declined randomization or whose insurance denied coverage for PBT. Patients are stratified by clinical site, age, use of rectal spacer, and fractionation schedule (conventional fractionation: 79.2 Gy in 44 fractions vs moderate hypofractionation: 70.0 Gy in 28 fractions). Participants are followed longitudinally to assess patient-reported outcomes (PROs) of bowel, urinary, and erectile function for 60 months after completion of radiotherapy (with an option for additional follow up through 10 years). Participants may also participate in correlative studies, including serial CT imaging during treatment and analyses of biopsy tissue, blood and urine specimens. The primary objective is to compare PROs of bowel function using the EPIC score at 24 months following completion of radiation. Secondary objectives are to assess treatment-related differences in urinary and erectile functions, adverse events, efficacy endpoints (biochemical control, metastasis-free survival, disease-specific survival, and overall survival), health state utilities, perceptions of care, late effects, cost-effectiveness, association between radiotherapy dose distribution and PROs, and to identify biomarkers of radiation response and toxicity.

RESULTS

The randomized trial has completed accrual, with 450 patients enrolled at 27 sites between June 2012 and November 2021. 20.3% of patients enrolled are non-white. Accrual on the companion registry is active, with 354 patients enrolled as of February 2023.

CONCLUSION

Follow-up for the primary endpoint on the randomized trial will be reached in 2024. The PARTIQoL randomized clinical trial will rigorously assess the clinical benefits of PBT relative to IMRT and results will inform decision making by patients, providers, policymakers, and payers.

Detection of Medication Taking Using a Wrist-Worn Commercially Available Wearable Device

PURPOSE

Medication nonadherence is a persistent and costly problem across health care. Measures of medication adherence are ineffective. Methods such as self-report, prescription claims data, or smart pill bottles have been used to monitor medication adherence, but these are subject to recall bias, lack real-time feedback, and are often expensive.

METHODS

We proposed a method for monitoring medication adherence using a commercially available wearable device. Passively collected motion data were analyzed on the basis of the Movelet algorithm, a dictionary learning framework that builds person-specific chapters of movements from short frames of elemental activities within the movements. We adapted and extended the Movelet method to construct a within-patient prediction model that identifies medication-taking behaviors.

RESULTS

Using 15 activity features recorded from wrist-worn wearable devices of 10 patients with breast cancer on endocrine therapy, we demonstrated that medication-taking behavior can be predicted in a controlled clinical environment with a median accuracy of 85%.

CONCLUSION

These results in a patient-specific population are exemplar of the potential to measure real-time medication adherence using a wrist-worn commercially available wearable device.

Estimating Pack-Year Eligibility for Lung Cancer Screening Using 2 Yes or No Questions

This cross-sectional study describes the development and testing the accuracy of using 2 yes or no questions to estimate pack-year eligibility for lung cancer screening.

Comparing Risk for Second Primary Cancers After Intensity-Modulated vs 3-Dimensional Conformal Radiation Therapy for Prostate Cancer, 2002-2015

Importance

Compared with 3-dimensional conformal radiotherapy (3DCRT), intensity-modulated radiotherapy (IMRT) can spare nearby tissue but may result in increased scatter radiation to distant normal tissue, including red bone marrow. It is unclear whether second primary cancer risk varies by radiotherapy type.

Objective

To evaluate whether radiotherapy type (IMRT vs 3DCRT) is associated with second primary cancer risk among older men treated for prostate cancer.

Design, Setting, and Participants

In this retrospective cohort study of a linked database of Medicare claims and Surveillance, Epidemiology, and End Results (SEER) Program population-based cancer registries (2002-2015), male patients aged 66 to 84 diagnosed with a first primary nonmetastatic prostate cancer from 2002 to 2013, as reported to SEER, and who received radiotherapy (IMRT and/or 3DCRT without proton therapy) within the first year following prostate cancer were identified. The data were analyzed from January 2022 through June 2022.

Exposure

Receipt of IMRT and 3DCRT, based on Medicare claims.

Main Outcomes and Measures

The association between radiotherapy type and development of a subsequent hematologic cancer at least 2 years after prostate cancer diagnosis or a subsequent solid cancer at least 5 years after prostate cancer diagnosis. Hazard ratios (HRs) and 95% CIs were estimated using multivariable Cox proportional regression.

Results

The study included 65 235 2-year first primary prostate cancer survivors (median [range] age, 72 [66-82] years; 82.2% White patients) and 45 811 5-year survivors with similar demographic characteristics (median [range] age, 72 [66-79] years; 82.4% White patients). Among 2-year prostate cancer survivors (median [range] follow-up, 4.6 [0.003-12.0] years), 1107 second hematologic cancers were diagnosed (IMRT, 603; 3DCRT, 504). Radiotherapy type was not associated with second hematologic cancers overall or any specific types evaluated. Among 5-year survivors (median [range] follow-up, 3.1 [0.003-9.0] years), 2688 men were diagnosed with a second primary solid cancer (IMRT, 1306; 3DCRT, 1382). The overall HR for IMRT vs 3DCRT was 0.91 (95% CI, 0.83-0.99). This inverse association was restricted to the earlier calendar year period of prostate cancer diagnosis (HR2002-2005 = 0.85; 95% CI, 0.76-0.94; HR2006-2010 = 1.14; 95% CI, 0.96-1.36), with a similar pattern observed for colon cancer (HR2002-2005 = 0.66; 95% CI, 0.46-0.94; HR2006-2010 = 1.06; 95% CI, 0.59-1.88).

Conclusions and Relevance

The results of this large, population-based cohort study suggest that IMRT for prostate cancer is not associated with an increased risk of second primary cancers, either solid or hematologic, and any inverse associations may be associated with calendar year of treatment.

Association Between Availability of Molecular Genotyping Results and Overall Survival in Patients With Advanced Nonsquamous Non-Small-Cell Lung Cancer

PURPOSE

Current guidelines recommend molecular genotyping for patients newly diagnosed with metastatic nonsquamous (mNSq) non-small-cell lung cancer (NSCLC). The association between availability of molecular genotyping before first line (1L) therapy and overall survival (OS) is not known.

METHODS

We conducted a real-world cohort study using electronic health records in patients newly diagnosed with mNSq NSCLC. Cox proportional-hazards multivariable regression models were constructed to examine the association between OS and test result availability before 1L therapy, adjusting for covariates. Additional analyses were conducted to assess the consistency and strength of the relationship. Multivariable logistic regression models were used to examine the association between concurrent tissue and plasma testing (v tissue alone) and result availability.

RESULTS

Three hundred twenty-six patients were included, 80% (261/326) with results available before 1L (available testing group), and 20% (65/326) without results available (unavailable testing group). With 14.2-month median follow-up, patients in the available testing group had significantly longer OS relative to the unavailable testing group (adjusted hazard ratio, 0.43; 95% CI, 0.30 to 0.62; P < .0001). The adjusted odds of availability of results before 1L therapy was higher with concurrent tissue and plasma testing (v tissue testing alone; adjusted odds ratio, 2.06; 95% CI, 1.09 to 3.90; P = .026).

CONCLUSION

Among patients with mNSq NSCLC in a real-world cohort, availability of molecular genotyping results before 1L therapy was associated with significantly better OS. Concurrent tissue and plasma testing was associated with a higher odds of availability of results before 1L therapy. These findings warrant renewed attention to the completion of molecular genotyping before 1L therapy.

Insurance and racial disparities in prior authorization in gynecologic oncology

While prior authorization aims to reduce unnecessary care, it may limit or delay medically necessary care. Delays in cancer care can impact survival and are more common in historically-marginalized populations. Our objective was to examine to what extent disparities occurred in prior authorizations for gynecologic oncology. Using electronic medical records, we performed a retrospective review of prior authorization occurrence during gynecologic oncology care and analyzed the association with patient race and insurance in a multivariate regression model. In this cohort of 1,406 patients treated at an academic gynecologic oncology practice, patients with Medicare Advantage and patients of Asian descent were more likely to experience prior authorization. Addressing insurance-mediate disparities, such as in the occurrence of prior authorization, may help reduce disparities in gynecologic cancer care.

Behavioural economic interventions to embed palliative care in community oncology (BE-EPIC): study protocol for the BE-EPIC randomised controlled trial

INTRODUCTION

Palliative care (PC) is a medical specialty focusing on providing relief from the symptoms and stress of serious illnesses such as cancer. Early outpatient specialty PC concurrent with cancer-directed treatment improves quality of life and symptom burden, decreases aggressive end-of-life care and is an evidence-based practice endorsed by national guidelines. However, nearly half of patients with advanced cancer do not receive specialty PC prior to dying. The objective of this study is to test the impact of an oncologist-directed default PC referral orders on rates of PC utilisation and patient quality of life.

METHODS AND ANALYSIS

This single-centre two-arm pragmatic randomised trial randomises four clinician-led pods, caring for approximately 250 patients who meet guideline-based criteria for PC referral, in a 1:1 fashion into a control or intervention arm. Intervention oncologists receive a nudge consisting of an electronic health record message indicating a patient has a default pended order for PC. Intervention oncologists are given an opportunity to opt out of referral to PC. Oncologists in pods randomised to the control arm will receive no intervention beyond usual practice. The primary outcome is completed PC visits within 12 weeks. Secondary outcomes are change in quality of life and absolute quality of life scores between the two arms.

ETHICS AND DISSEMINATION

This study has been approved by the Institutional Review Board at the University of Pennsylvania. Study results will be disseminated in peer-reviewed journals and scientific conferences using methods that describe the results in ways that key stakeholders can best understand and implement.

TRIAL REGISTRATION NUMBER

NCT05365997.

Long-term Effect of Machine Learning-Triggered Behavioral Nudges on Serious Illness Conversations and End-of-Life Outcomes Among Patients With Cancer: A Randomized Clinical Trial

Importance

Serious illness conversations (SICs) between oncology clinicians and patients are associated with improved quality of life and may reduce aggressive end-of-life care. However, most patients with cancer die without a documented SIC.

Objective

To test the impact of behavioral nudges to clinicians to prompt SICs on the SIC rate and end-of-life outcomes among patients at high risk of death within 180 days (high-risk patients) as identified by a machine learning algorithm.

Design, Setting, and Participants

This prespecified 40-week analysis of a stepped-wedge randomized clinical trial conducted between June 17, 2019, and April 20, 2020 (including 16 weeks of intervention rollout and 24 weeks of follow-up), included 20 506 patients with cancer representing 41 021 encounters at 9 tertiary or community-based medical oncology clinics in a large academic health system. The current analyses were conducted from June 1, 2021, to May 31, 2022.

Intervention

High-risk patients were identified using a validated electronic health record machine learning algorithm to predict 6-month mortality. The intervention consisted of (1) weekly emails to clinicians comparing their SIC rates for all patients against peers' rates, (2) weekly lists of high-risk patients, and (3) opt-out text messages to prompt SICs before encounters with high-risk patients.

Main Outcomes and Measures

The primary outcome was SIC rates for all and high-risk patient encounters; secondary end-of-life outcomes among decedents included inpatient death, hospice enrollment and length of stay, and intensive care unit admission and systemic therapy close to death. Intention-to-treat analyses were adjusted for clinic and wedge fixed effects and clustered at the oncologist level.

Results

The study included 20 506 patients (mean [SD] age, 60.0 [14.0] years) and 41 021 patient encounters: 22 259 (54%) encounters with female patients, 28 907 (70.5%) with non-Hispanic White patients, and 5520 (13.5%) with high-risk patients; 1417 patients (6.9%) died by the end of follow-up. There were no meaningful differences in demographic characteristics in the control and intervention periods. Among high-risk patient encounters, the unadjusted SIC rates were 3.4% (59 of 1754 encounters) in the control period and 13.5% (510 of 3765 encounters) in the intervention period. In adjusted analyses, the intervention was associated with increased SICs for all patients (adjusted odds ratio, 2.09 [95% CI, 1.53-2.87]; P < .001) and decreased end-of-life systemic therapy (7.5% [72 of 957 patients] vs 10.4% [24 of 231 patients]; adjusted odds ratio, 0.25 [95% CI, 0.11-0.57]; P = .001) relative to controls, but there was no effect on hospice enrollment or length of stay, inpatient death, or end-of-life ICU use.

Conclusions and Relevance

In this randomized clinical trial, a machine learning-based behavioral intervention and behavioral nudges to clinicans led to an increase in SICs and reduction in end-of-life systemic therapy but no changes in other end-of-life outcomes among outpatients with cancer. These results suggest that machine learning and behavioral nudges can lead to long-lasting improvements in cancer care delivery.

Trial Registration

ClinicalTrials.gov Identifier: NCT03984773.

Creating research-ready partnerships: the initial development of seven implementation laboratories to advance cancer control

BACKGROUND

In 2019-2020, with National Cancer Institute funding, seven implementation laboratory (I-Lab) partnerships between scientists and stakeholders in 'real-world' settings working to implement evidence-based interventions were developed within the Implementation Science Centers in Cancer Control (ISC3) consortium. This paper describes and compares approaches to the initial development of seven I-Labs in order to gain an understanding of the development of research partnerships representing various implementation science designs.

METHODS

In April-June 2021, members of the ISC3 Implementation Laboratories workgroup interviewed research teams involved in I-Lab development in each center. This cross-sectional study used semi-structured interviews and case-study-based methods to collect and analyze data about I-Lab designs and activities. Interview notes were analyzed to identify a set of comparable domains across sites. These domains served as the framework for seven case descriptions summarizing design decisions and partnership elements across sites.

RESULTS

Domains identified from interviews as comparable across sites included engagement of community and clinical I-Lab members in research activities, data sources, engagement methods, dissemination strategies, and health equity. The I-Labs use a variety of research partnership designs to support engagement including participatory research, community-engaged research, and learning health systems of embedded research. Regarding data, I-Labs in which members use common electronic health records (EHRs) leverage these both as a data source and a digital implementation strategy. I-Labs without a shared EHR among partners also leverage other sources for research or surveillance, most commonly qualitative data, surveys, and public health data systems. All seven I-Labs use advisory boards or partnership meetings to engage with members; six use stakeholder interviews and regular communications. Most (70%) tools or methods used to engage I-Lab members such as advisory groups, coalitions, or regular communications, were pre-existing. Think tanks, which two I-Labs developed, represented novel engagement approaches. To disseminate research results, all centers developed web-based products, and most (n = 6) use publications, learning collaboratives, and community forums. Important variations emerged in approaches to health equity, ranging from partnering with members serving historically marginalized populations to the development of novel methods.

CONCLUSIONS

The development of the ISC3 implementation laboratories, which represented a variety of research partnership designs, offers the opportunity to advance understanding of how researchers developed and built partnerships to effectively engage stakeholders throughout the cancer control research lifecycle. In future years, we will be able to share lessons learned for the development and sustainment of implementation laboratories.

Development of Machine Learning Algorithms Incorporating Electronic Health Record Data, Patient-Reported Outcomes, or Both to Predict Mortality for Outpatients With Cancer

PURPOSE

Machine learning (ML) algorithms that incorporate routinely collected patient-reported outcomes (PROs) alongside electronic health record (EHR) variables may improve prediction of short-term mortality and facilitate earlier supportive and palliative care for patients with cancer.

METHODS

We trained and validated two-phase ML algorithms that incorporated standard PRO assessments alongside approximately 200 routinely collected EHR variables, among patients with medical oncology encounters at a tertiary academic oncology and a community oncology practice.

RESULTS

Among 12,350 patients, 5,870 (47.5%) completed PRO assessments. Compared with EHR- and PRO-only algorithms, the EHR + PRO model improved predictive performance in both tertiary oncology (EHR + PRO v EHR v PRO: area under the curve [AUC] 0.86 [0.85-0.87] v 0.82 [0.81-0.83] v 0.74 [0.74-0.74]) and community oncology (area under the curve 0.89 [0.88-0.90] v 0.86 [0.85-0.88] v 0.77 [0.76-0.79]) practices.

CONCLUSION

Routinely collected PROs contain added prognostic information not captured by an EHR-based ML mortality risk algorithm. Augmenting an EHR-based algorithm with PROs resulted in a more accurate and clinically relevant model, which can facilitate earlier and targeted supportive care for patients with cancer.

Association of Oncologist Participation in Medicare's Oncology Care Model With Patient Receipt of Novel Cancer Therapies

IMPORTANCE

Medicare's Oncology Care Model (OCM) was an alternative payment model that tied performance-based payments to cost and quality goals for participating oncology practices. A major concern about the OCM regarded inclusion of high-cost cancer therapies, which could potentially disincentivize oncologists from prescribing novel therapies.

OBJECTIVE

To examine whether oncologist participation in the OCM changed the likelihood that patients received novel therapies vs alternative treatments.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study of Surveillance, Epidemiology, and End Results (SEER) Program data and Medicare claims compared patient receipt of novel therapies for patients treated by oncologists participating vs not participating in the OCM in the period before (January 2015-June 2016) and after (July 2016-December 2018) OCM initiation. Participants included Medicare fee-for-service beneficiaries in SEER registries who were eligible to receive 1 of 10 novel cancer therapies that received US Food and Drug Administration approval in the 18 months before implementation of the OCM. The study excluded the Hawaii registry because complete data were not available at the time of the data request. Patients in the OCM vs non-OCM groups were matched on novel therapy cohort, outcome time period, and oncologist specialist status. Analysis was conducted between July 2021 and April 2022.

EXPOSURES

Oncologist participation in the OCM.

MAIN OUTCOMES AND MEASURES

Preplanned analyses evaluated patient receipt of 1 of 10 novel therapies vs alternative therapies specific to the patient's cancer for the overall study sample and for racial subgroups.

RESULTS

The study included 2839 matched patients (760 in the OCM group and 2079 in the non-OCM group; median [IQR] age, 72.7 [68.3-77.6] years; 1591 women [56.0%]). Among patients in the non-OCM group, 33.2% received novel therapies before and 40.1% received novel therapies after the start of the OCM vs 39.9% and 50.3% of patients in the OCM group (adjusted difference-in-differences, 3.5 percentage points; 95% CI, -3.7 to 10.7 percentage points; P = .34). In subgroup analyses, second-line immunotherapy use in lung cancer was greater among patients in the OCM group vs non-OCM group (adjusted difference-in-differences, 17.4 percentage points; 95% CI, 4.8-30.0 percentage points; P = .007), but no differences were seen in other subgroups. Over the entire study period, patients with oncologists participating in the OCM were more likely to receive novel therapies than those with oncologists who were not participating (odds ratio, 1.47; 95% CI, 1.09-1.97; P = .01).

CONCLUSIONS AND RELEVANCE

This study found that participation in the OCM was not associated with oncologists' prescribing novel therapies to Medicare beneficiaries with cancer. These findings suggest that OCM financial incentives did not decrease patient access to novel therapies.

Long-term effect of machine learning–triggered behavioral nudges on serious illness communication and end-of-life outcomes among patients with cancer: A randomized clinical trial

109

Background: Early serious illness conversations (SICs) between oncology clinicians and patients are associated with improved mood, quality of life, and quality of end-of-life (EOL) care. Yet, most patients with cancer die without a documented SIC. We report on pre-specified 40-week SIC and EOL outcomes from a stepped-wedge randomized clinical trial (NCT03984773) testing the impact of clinician-directed behavioral nudges to prompt SICs among patients with cancer at high risk of mortality based on a machine learning algorithm. Methods: Our sample consisted of patients with cancer receiving care at one of 9 tertiary or community-based medical oncology clinics between June 2019 and April 2020. We identified high-risk patients using a prospectively validated electronic health record machine learning algorithm to predict 6-month mortality. The intervention consisted of: (1) Weekly emails comparing individual oncologists’ SIC rate relative to peers; (2) Weekly lists of forthcoming encounters with high-risk patients; and (3) Opt-out text messages to prompt SICs before high-risk patient encounters. Clinics were randomized in stepped-wedge fashion to receive the intervention in 4-week intervals through week 16, when all clinics received the intervention. Patients were followed through week 40. The primary outcome was SIC rates for all and high-risk patients. EOL outcomes among decedents were based on ASCO/NQF guidelines and included death in the hospital, intensive care unit admission within 30 days of death, receipt of systemic therapy within 14 days of death, hospice enrollment prior to death, and hospice length of stay. Intention-to-treat analyses were adjusted for clinic and wedge fixed effects and clustered at the oncologist-level. Results: The sample consisted of 20,506 patients and 41,021 encounters. 1,324 (6.5%) patients died by the end of follow-up. Among high-risk patients, the unadjusted SIC rate was 3.4% (59/1754) in the control period and 13.5% (510/3765) in the intervention period and remained >12% throughout follow-up. In adjusted analyses, the intervention was associated with an increase in SICs (adjusted odds ratio 2.09, 95% CI 1.53-2.87, p<0.001) and a decrease in systemic therapy at the end of life, relative to control (6.8% [72/1066]) vs 9.3% [24/258], adjusted odds ratio 0.27, 95% CI 0.12-0.63, p=0.002). There were no differences between control and intervention patients in hospice enrollment or length of stay, inpatient death, or EOL ICU utilization. Conclusions: In this randomized trial, a machine learning-based behavioral intervention led to a sustained increase in serious illness communication and reduction in EOL systemic therapy among outpatients with cancer. Machine learning and behavioral nudges can lead to long-lasting improvements in cancer care delivery. Clinical trial information: NCT03984773.

Racial and ethnic disparities in adherence and reported symptoms during routine collection of patient-reported outcomes (PROs)

6510

Background: Routine collection of PROs for patients with cancer is an evidence-based practice and critical component of high-quality cancer care, but real-world adherence and reporting patterns are poorly understood. We examined differences in PRO adherence and reported symptoms by race and ethnicity. Methods: We conducted a retrospective cross-sectional study using de-identified electronic health record data from a National Cancer Institute Comprehensive Cancer Center. Participants included adults seen in follow-up at one of two medical oncology practices (1 academic, 1 community) from June 2019 to February 2020, during which two validated PRO instruments (NCI PRO-CTCAE and PROMIS Global v1.2) were routinely administered electronically via portal or tablet at return visits. Adherence was defined for each participant as the proportion of visits in which a PRO questionnaire was completed within 30 days. Using ordinary least squares regression, we modeled patient adherence as a function of race/ethnicity and adjusted for age, sex, insurance, median area income, ECOG performance status, cancer type, cancer stage, visit site, and number of visits during the study. Among study participants completing at least one PRO questionnaire, we modeled reported symptoms from the first questionnaire using a similar approach. Results: From June 2019 to February 2020, there were 29,726 patients (mean [SD] age 60.8 [15.1] years; 18,045 [60.7%] female) seen across 111,262 medical oncology visits. Of these patients, 19,971 (67.2%) completed at least one PRO questionnaire. Adjusted mean PRO adherence and reported symptoms varied by race/ethnicity, with Black and Hispanic patients reporting significantly higher symptom burden than White patients (Table). Conclusions: In this large cohort reflecting real-world PRO collection patterns, Black and Hispanic patients were less likely than White patients to complete PRO questionnaires, but more likely to report more severe symptoms. There is urgent need to ensure equitable PRO access and implementation and to address greater reported symptom burden among minority patients. [Table: see text]

Evaluating risk for second primary cancers by radiotherapy technique in prostate cancer survivors

12005

Background: Radiotherapy-related adverse effects such as the development of subsequent neoplasms cause significant morbidity among prostate cancer survivors. Advances in radiotherapy techniques, including intensity-modulated radiation therapy (IMRT) and proton beam radiotherapy (PBRT), have aimed to reduce exposure to adjacent healthy tissues to reduce adverse effects. Initial reports based on small sample sizes and limited follow up (through 2011) have suggested reduced risks for subsequent colon and rectal cancers following IMRT compared to 3D conformal radiotherapy (CRT) for prostate cancer patients, but not for subsequent bladder cancer. We sought to extend previous reports with larger sample size and longer follow up. Methods: We conducted a retrospective cohort study within the linked database of Surveillance, Epidemiology and End Results (SEER) cancer registries and Medicare claims. The cohort included men diagnosed with first primary non-metastatic prostate cancer at ages 66-84 during 2002-2011; received initial IMRT, PBRT, or CRT; and survived without developing a second primary cancer ≥5 years after diagnosis (follow up through 2016). Cox regression models estimated risks of second primary solid tumors after IMRT and PBRT vs CRT, adjusting for age at prostate cancer diagnosis, tumor grade, race, Charlson comorbidity score, and receipt of initial prostate cancer therapy. Results: The cohort (median follow-up = 8.4 years) included 51,020 patients, of whom 19,536 received CRT alone, 29,868 received IMRT without PBRT, and 1,616 received PBRT. Compared to patients who received CRT (n = 1,348, 7.0%), both IMRT (n = 1,289, 4.3%; hazard ratio [HR] = 0.87; 95% confidence interval [CI], 0.80-0.95) and PBRT (n = 83, 5.1%; HR = 0.77; 95% CI 0.62-0.97) showed a decrease in risk of developing any second solid malignancy. In analyses by second cancer type, risks of colon cancer (HR = 0.70; 95%CI, 0.52-0.94) and bladder cancer (HR = 0.79; 95%CI, 0.65-0.97) were significantly lower after IMRT than CRT, whereas no association was observed for anorectal cancers (HR = 1.00; 95%CI, 0.65-1.53). Further investigation by time since prostate cancer diagnosis revealed a time-dependent decrease after IMRT compared to CRT in risk for bladder cancer (HRs = 0.94, 0.87, 0.61 for 5-7.4, 7.5-9.9, and 10+ years respectively) and anorectal cancer (HRs = 1.22, 0.97, 0.78), whereas the opposite trend was observed for colon cancer (HRs = 0.70, 0.68, 1.05). Conclusions: In this large cohort with increased follow-up time compared with previous reports, we observed reduced risk of colon and bladder cancer with IMRT overall, as well as time-dependent patterns for bladder and anorectal cancer that were consistent with improved tumor targeting. Further research is needed with larger sample sizes to evaluate long-term effects after PBRT. Our study supports the value of quantifying adverse effects as radiotherapy techniques evolve.

Trends in low-value oncology care during the COVID-19 pandemic

6594

Background: Low-value services, which provide minimal patient benefit while entailing costs and risks, are prevalent in cancer care. Shifts in cancer care delivery during the COVID-19 pandemic to minimize exposure provided opportunities for health systems and clinicians to prioritize higher-value over low-value oncology services. Methods: In this retrospective cohort study, we investigated the association between the COVID-19 pandemic period and low-value cancer care practices using administrative claims from the HealthCore Integrated Research Environment, consisting of ̃65 million members managed by 14 health plans across the US. We identified commercial or Medicare Advantage members diagnosed with breast, colorectal, or lung cancer between January 2015 and March 2021. Low-value cancer care practices were identified from peer-reviewed medical literature, including ASCO and ASTRO Choosing Wisely campaigns and evidence-based pathways. Five low-value practices were studied: (1) conventional fractionation instead of hypofractionation for early-stage breast cancer; (2) off-pathway systemic therapy; (3) non-guideline-based antiemetic use for minimal-, low-, or moderate-to-high-risk chemotherapies; (4) Positron Emission Tomography/Computed Tomography (PET/CT) instead of conventional CT for staging; and (5) aggressive end-of-life care (chemotherapy ≤14 days, multiple emergency department visits ≤30 days, ICU utilization ≤30 days, hospice initiation ≤3 days, and/or no hospice before death). We used linear probability models to evaluate the association between the COVID-19 period (March to December 2020) and the 5 outcomes, adjusting for patient, facility, geographic and temporal characteristics. Results: Among 204,581 members (mean age 63.1, 139,488 [68.1%] female), 83,593 (40.8%) had breast cancer, 56,373 (27.5%) had colon cancer, and 64,615 (31.5%) had lung cancer. Rates of low-value care were similar in pre-COVID vs. COVID periods: conventional radiotherapy: 22.1% vs. 9.4%; off-pathway systemic therapy: 36.7% vs. 43.2%; non-guideline-based antiemetics: 61.2% vs. 58.1%; PET/CT imaging: 39.9% vs. 41.3%; aggressive end-of-life care: 75.8% vs. 73.3%. In adjusted analyses, the COVID-19 period was associated with no changes in off-pathway therapy (adjusted percentage point difference [aPPD] 0.82, SD 0.08, p = 0.33), PET/CT imaging (aPPD 0.10, SD 0.005, p = 0.83), and aggressive end-of-life care (aPPD 2.71, SD 0.02, p = 0.16). Small changes in conventional radiotherapy (aPPD 3.93, SD 0.01, p < 0.01) and non-guideline-based antiemetics (aPPD -3.62, SD 0.006, p < 0.01), were noted. Conclusions: The shock of the COVID-19 pandemic did not meaningfully change several metrics of low-value cancer care. Broader changes to payment and incentive design should be considered to turn the tide toward higher-value cancer care.

Willingness to Travel for Cellular Therapy: The Influence of Follow-Up Care Location, Oncologist Continuity, and Race

PURPOSE

Patients weigh competing priorities when deciding whether to travel to a cellular therapy center for treatment. We conducted a choice-based conjoint analysis to determine the relative value they place on clinical factors, oncologist continuity, and travel time under different post-treatment follow-up arrangements. We also evaluated for differences in preferences by sociodemographic factors.

METHODS

We administered a survey in which patients with diffuse large B-cell lymphoma selected treatment plans between pairs of hypothetical options that varied in travel time, follow-up arrangement, oncologist continuity, 2-year overall survival, and intensive care unit admission rate. We determined importance weights (which represent attributes' value to participants) using generalized estimating equations.

RESULTS

Three hundred and two patients (62%) responded. When all follow-up care was at the center providing treatment, plans requiring longer travel times were less attractive (v 30 minutes, importance weights [95% CI] of -0.54 [-0.80 to -0.27], -0.57 [-0.84 to -0.29], and -0.17 [-0.49 to 0.14] for 60, 90, and 120 minutes). However, the negative impact of travel on treatment plan choice was mitigated by offering shared follow-up (importance weights [95% CI] of 0.63 [0.33 to 0.93], 0.32 [0.08 to 0.57], and 0.26 [0.04 to 0.47] at 60, 90, and 120 minutes). Black participants were less likely to choose plans requiring longer travel, regardless of follow-up arrangement, as indicated by lower value importance weights for longer travel times.

CONCLUSION

Reducing travel burden through shared follow-up may increase patients' willingness to travel to receive cellular therapies, but additional measures are required to facilitate equitable access.

Association of gynecologic oncology versus medical oncology specialty with survival, utilization, and spending for treatment of gynecologic cancers

BACKGROUND

We examined the association of gynecologic oncology (GYO) versus medical oncology (MEDONC) based care with survival, health care utilization and spending outcomes in women undergoing chemotherapy for advanced gynecologic cancers.

METHODS

Women with newly diagnosed stage III-IV uterine, ovarian, and cervical cancers from 2000 to 2015 were identified in SEER-Medicare. We assessed the association of provider specialty with overall survival, emergency department utilization, admissions, and spending. Outcomes were assessed using unadjusted and Inverse Treatment Probability Weighted propensity-score applied, multi-variable cox modeling, Poisson regression, and generalized models of log-transformed data.

RESULTS

We identified 7930 gynecologic cancer patients (4360 ovarian, 2934 uterine, 643 cervix). 37% were treated by GYO and 63% by MEDONC. For ovarian patients, GYO care was associated with improved OS (median OS 3.3 v. 2.9 years; HR 0.85, 95%CI 0.80, 0.91, p < .0001) and similar mean spending per month ($4015 v. $4316, mean ratio 0.97 (95% CI 0.93, 1.02), p = .19), compared to MEDONC in adjusted analyses. For uterine patients, GYO care was associated with similar OS, but decreased spending ($3573 v. $4081, mean ratio 0.87 (95% CI.81, 0.93), p < .0001), and decreased ED utilization (RR 0.76, 95% CI 0.69, 0.85, p < .0001). For cervical patients, GYO care was associated with similar OS, and similar spending. Admissions were more likely in ovarian (RR 1.23, 95%CI 1.11, 1.37, p = .0001) and cervical patients (RR 1.26, 95% CI 1.05, 1.51, p = .015) treated by GYO, in adjusted analyses.

CONCLUSIONS

GYO based care was associated with improved OS and equal spending for patients with advanced stage ovarian cancer. Uterine and cervix patients had similar OS, and less or equal spending respectively, when treated by GYO compared to MEDONC.

Long-term Clinical Outcomes in Favorable Risk Prostate Cancer Patients Receiving Proton Beam Therapy

Purpose

Long-term data regarding the disease control outcomes of proton beam therapy (PBT) for patients with favorable risk intact prostate cancer (PC) are limited. Herein, we report our institution's long-term disease control outcomes in PC patients with clinically localized disease who received PBT as primary treatment.

Methods

One hundred sixty-six favorable risk PC patients who received definitive PBT to the prostate gland at our institution from 2010 to 2012 were retrospectively assessed. The outcomes studied were biochemical failure-free survival (BFFS), biochemical failure, local failure, regional failure, distant failure, PC-specific survival, and overall survival. Patterns of failure were also analyzed. Multivariate Cox proportional hazards modeling was used to estimate independent predictors of BFFS.

Results

The median length of follow-up was 8.3 years (range, 1.2–10.5 years). The majority of patients had low-risk disease (58%, n = 96), with a median age of 64 years at the onset of treatment. Of 166 treated men, 13 (7.8%), 8 (4.8%), 2 (1.2%) patient(s) experienced biochemical failure, local failure, regional failure, respectively. Regional failure was seen in an obturator lymph node in 1 patient and the external iliac lymph nodes in the other. None of the patients experienced distant failure. There were 5 (3.0%) deaths, none of which were due to PC. The 5- and 8-year BFFS rate were 97% and 92%, respectively. None of the clinical disease characteristics or treatment-related factors assessed were associated with BFFS on multivariate Cox proportional hazards modeling (all P > .05).

Conclusion

Disease control rates reported in our assessment of PBT were similar to those reported in previous clinically localized intact PC analyses, which used intensity-modulated radiotherapy, three-dimensional conformal radiotherapy, or radical prostatectomy as definitive therapy. In addition, BFFS rates were similar, if not improved, to previous PBT studies.

Application of an automatic segmentation method for evaluating cardiac structure doses received by breast radiotherapy patients

•

Atlas-based method for contouring heart substructures on breast radiotherapy CT.

•

Excellent agreement between automatic and manual contours for most patients.

•

Dice similarity coefficient for LAD was low (0.06) because a narrow, long structure.

•

Doses derived from automatic and manual contours agree within observer variability.

•

For left breast treatment, right ventricle and LAD dose most senstive to contour shift.

Background and purpose

Quantifying radiation dose to cardiac substructures is important for research on the etiology and prevention of complications following radiotherapy; however, segmentation of substructures is challenging. In this study we demonstrate the application of our atlas-based automatic segmentation method to breast cancer radiotherapy plans for generating radiation doses in support of late effects research.

Material and methods

We applied our segmentation method to contour heart substructures on the computed tomography (CT) images of 70 breast cancer patients who received external photon radiotherapy. Two cardiologists provided manual segmentation of the whole heart (WH), left/right atria, left/right ventricles, and left anterior descending artery (LAD). The automatically contours were compared with manual delineations to evaluate similarity in terms of geometry and dose.

Results

The mean Dice similarity coefficient between manual and automatic segmentations was 0.96 for the WH, 0.65 to 0.82 for the atria and ventricles, and 0.06 for the LAD. The mean average surface distance was 1.2 mm for the WH, 3.4 to 4.1 mm for the atria and ventricles, and 6.4 mm for the LAD. We found the dose to the cardiac substructures based on our automatic segmentation agrees with manual segmentation within expected observer variability. For left breast patients, the mean absolute difference in mean dose was 0.1 Gy for the WH, 0.2 to 0.7 Gy for the atria and ventricles, and 1.8 Gy for the LAD. For right breast patients, these values were 0.0 Gy, 0.1 to 0.4 Gy, and 0.4 Gy, respectively.

Conclusion

Our automatic segmentation method will facilitate the development of radiotherapy prescriptive criteria for mitigating cardiovascular complications.

Differences in Cancer Care Expenditures and Utilization for Surgery by Hospital Type Among Patients With Private Insurance

IMPORTANCE

With rising expenditures on cancer care outpacing other sectors of the US health system, national attention has focused on insurer spending, particularly for patients with private insurance, for whom price transparency has historically been lacking. The type of hospital at which cancer care is delivered may be an important factor associated with insurer spending for patients with private insurance.

OBJECTIVE

To examine differences in spending and utilization for patients with private insurance undergoing common cancer surgery at National Cancer Institute (NCI) centers vs community hospitals.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cross-sectional study included adult patients with an incident diagnosis of breast, colon, or lung cancer who underwent cancer-directed surgery from 2011 to 2014. Mean risk-adjusted spending and utilization outcomes were examined for each hospital type using multilevel generalized linear mixed-effects models, adjusting for patient, hospital, and region characteristics. Data were collected from the Health Care Cost Institute's national multipayer commercial claims data set, which encompasses claims paid by 3 of the 5 largest commercial health insurers in the United States (ie, Aetna, Humana, and UnitedHealthcare). Data analyses were conducted from February 2018 to February 2019.

EXPOSURES

Hospital type at which cancer surgery was performed: NCI, non-NCI academic, or community.

MAIN OUTCOMES AND MEASURES

Spending outcomes were surgery-specific insurer prices paid and 90-day postdischarge payments. Utilization outcomes were length of stay (LOS), emergency department (ED) use, and hospital readmission within 90 days of discharge.

RESULTS

The study included 66 878 patients (51 569 [77.1%] women; 31 585 [47.2%] aged ≥65 years) with incident breast (35 788 [53.5%]), colon (21 378 [32.0%]), or lung (9712 [14.5%]) cancer undergoing cancer surgery at 2995 hospitals (5522 [8.3%] at NCI centers; 10 917 [16.3%] at non-NCI academic hospitals; 50 439 [75.4%] at community hospitals). Treatment at NCI centers was associated with higher surgery-specific insurer prices paid compared with community hospitals ($18 526 [95% CI, $16 650-$20 403] vs $14 772 [95% CI, $14 339-$15 204]; difference, $3755 [95% CI, $1661-$5849]; P < .001) and 90-day postdischarge payments ($47 035 [95% CI, $43 289-$50 781] vs $41 291 [95% CI, $40 350-$42 231]; difference, $5744 [95% CI, $1659-9829]; P = .006). There were no significant differences in LOS, ED use, or hospital readmission within 90 days of discharge.

CONCLUSIONS AND RELEVANCE

In this cross-sectional study, surgery at NCI centers vs community hospitals was associated with higher insurer spending for a surgical episode without differences in care utilization among patients with private insurance undergoing cancer surgery. A better understanding of the factors associated with prices and spending at NCI cancer centers is needed.

Trends in Patient Volume by Hospital Type and the Association of These Trends With Time to Cancer Treatment Initiation

Importance

Increasing demand for cancer care may be outpacing the capacity of hospitals to provide timely treatment, particularly at referral centers such as National Cancer Institute (NCI)-designated and academic centers. Whether the rate of patient volume growth has strained hospital capacity to provide timely treatment is unknown.

Objective

To evaluate trends in patient volume by hospital type and the association between a hospital's annual patient volume growth and time to treatment initiation (TTI) for patients with cancer.

Design, Setting, and Participants

This retrospective, hospital-level, cross-sectional study used longitudinal data from the National Cancer Database from January 1, 2007, to December 31, 2016. Adult patients older than 40 years who had received a diagnosis of 1 of the 10 most common incident cancers and initiated their treatment at a Commission on Cancer-accredited hospital were included. Data were analyzed between December 19, 2019, and March 27, 2020.

Exposures

The mean annual rate of patient volume growth at a hospital.

Main Outcomes and Measures

The main outcome was TTI, defined as the number of days between diagnosis and the first cancer treatment. The association between a hospital's mean annual rate of patient volume growth and TTI was assessed using a linear mixed-effects model containing a patient volume × time interaction. The mean annual change in TTI over the study period by hospital type was estimated by including a hospital type × time interaction term.

Results

The study sample included 4 218 577 patients (mean [SD] age, 65.0 [11.4] years; 56.6% women) treated at 1351 hospitals. From 2007 to 2016, patient volume increased 40% at NCI centers, 25% at academic centers, and 8% at community hospitals. In 2007, the mean TTI was longer at NCI and academic centers than at community hospitals (NCI: 50 days [95% CI, 48-52 days]; academic: 43 days [95% CI, 42-44 days]; community: 37 days [95% CI, 36-37 days]); however, the mean annual increase in TTI was greater at community hospitals (0.56 days; 95% CI, 0.49-0.62 days) than at NCI centers (-0.73 days; 95% CI, -0.95 to -0.51 days) and academic centers (0.14 days; 95% CI, 0.03-0.26 days). An annual volume growth rate of 100 patients, a level observed at less than 1% of hospitals, was associated with a mean increase in TTI of 0.24 days (95% CI, 0.18-0.29 days).

Conclusions and Relevance

In this cross-sectional study, from 2007 to 2016, across the studied cancer types, patients increasingly initiated their cancer treatment at NCI and academic centers. Although increases in patient volume at these centers outpaced that at community hospitals, faster growth was not associated with clinically meaningful treatment delays.

Willingness to travel for CAR-T: A choice-based conjoint analysis

e18503

Background: Chimeric antigen receptor T-cell therapy (CAR-T) is a novel therapy only available at specialized centers. Because patients prefer treatment close to home and provider continuity, preferences may result in unequal access. Sharing follow-up (f/u) with a local provider could increase willingness to travel, but its appeal to patients is unknown. Methods: We conducted a choice-based conjoint analysis to determine the clinical and care delivery factors that patients with diffuse-large B cell lymphoma (DLBCL) value when deciding whether to travel for CAR-T. We recruited from 13 sites within an integrated health system and asked participants to make choices between 12 pairs of treatment options that varied on 5 attributes: travel time, oncologist continuity, whether shared f/u was offered, two-year overall survival (OS), and intensive care unit (ICU) admission rate. Preferences for specific attributes were determined using a generalized estimating equation model with treatment choice as the outcome and the attributes as the variables. Each attribute’s coefficient (importance weight [IW]) represented its value to participants: higher IWs indicated greater value; negative IWs indicated undesirable attributes. Results: We invited 489 patients and 303 (62%) responded. IWs ranged from -0.56 to 1.15 (total cohort) and -2.83 to 2.50 (when modeled by race). Treatments at cancer centers ≥60 minutes away were valued less than those 30 minutes away without collaborative f/u; but were desirable if shared f/u care was offered (Table). Black respondents valued treatments ≥60 minutes away less despite shared f/u care. Participants also valued oncologist continuity (IW 1.15 [0.99, 1.32]), greater OS (IW 0.71 [0.63, 0.79] per 5 percentage point increase), and lower ICU admission rates (IW -0.21 [-0.25, -0.17] per 5 percentage point increase). Oncologist continuity was more highly valued by Black than by White respondents (IW 2.50 [1.74, 3.27] vs 1.09 [0.92, 1.25]). Conclusions: Patients with DLBCL are more likely to select CAR-T therapy at distant cancer centers if f/u care is shared locally. This could represent a strategy to expand access to CAR-T. Travel remains a barrier for Black patients despite shared f/u, suggesting the need to investigate targeted strategies that promote equitable access to CAR-T. [Table: see text]

Augmenting machine learning algorithms to predict mortality using patient-reported outcomes in oncology

1510

Background: Machine learning (ML) algorithms based on electronic health record (EHR) data have been shown to accurately predict mortality risk among patients with cancer, with areas under the curve (AUC) generally greater than 0.80. While patient-reported outcomes (PROs) may also predict mortality among patients with cancer, it is unclear whether routinely-collected PROs improve the predictive performance of EHR-based ML algorithms. Methods: This cohort study included 8600 patients with cancer who had an outpatient encounter at one of 18 medical oncology practices in a large academic health system between July 1st, 2019 and January 1st, 2020. 4692 (54.9%) patients completed assessments of symptoms, performance status, and quality of life from the PRO version of the Common Terminology Criteria for Adverse Events and the Patient-Reported Outcomes Measurement Information System Global v.1.2 scales. We hypothesized that ML models predicting 180-day all-cause mortality based on EHR + PRO data would improve AUC compared to ML models based on EHR data alone. We assessed univariate and adjusted associations between each PRO and 180-day mortality. To train the EHR-only model, we fit a Least Absolute Shrinkage and Selection Operator (LASSO) regression using 192 EHR demographic, comorbidity, and laboratory variables. To train the EHR + PRO model, we used a two-phase approach to fit a model using EHR data for all patients and PRO data for those who completed assessments. To test our hypothesis, we compared the bootstrapped AUC, area under the precision-recall curve (AUPRC), and sensitivity at a 20% risk threshold for both models. Results: 464 (5.4%) patients died within 180 days of the encounter. Decreased quality of life, functional status, and appetite were associated with greater 180-day mortality (Table). Compared to the EHR-only model, the EHR + PRO model significantly improved AUC (0.86 [95% CI 0.85-0.86] vs. 0.80 [95% CI 0.80-0.81]), AUPRC (0.40 [95% CI 0.37-0.42] vs. 0.30 [95% CI 0.28-0.32]), and sensitivity (0.45 [95% CI 0.42-0.48] vs. 0.33 [95% CI 0.30-0.35]). Conclusions: Routinely collected PROs augment EHR-based ML mortality risk algorithms. ML algorithms based on EHR and PRO data may facilitate earlier supportive care for patients with cancer. Association of PROs with 180-day mortality.[Table: see text]

Multi-site practice and physician travel burden by oncology specialty

e13513

Background: Having physicians who practice at multiple sites may increase patients’ access to care, but also burden physicians Gynecologic oncologists (GO) are increasingly practicing at more sites across a larger geographic area, but the degree to which medical oncologists (MO), surgical oncologists (SO) or radiation oncologists (RO) are also doing so is unknown. Methods: We conducted a retrospective, observational study using data from the 2020 Physician Compare National Dataset. We included GO, MO, SO, RO, as determined by self-reported specialty. Practice sites with incomplete street addresses were excluded. For each specialty, we calculated the number of practice sites per physician, geographic practice dispersion (median driving distance required to go to each practice site), and temporal practice dispersion (median travel time required to go to each practice site). We used linear regression to compare the number of practice sites, geographic practice dispersion, and temporal practice dispersion by specialty. Results: The number of physicians, mean number of practice sites, along with geographic and temporal practice dispersion by specialty are shown in the table. MO practiced at a smaller number of practice sites compared to GO (p<0.001) and RO (p<0.001). Compared to MO, SO had a smaller geographic dispersion (median driving distance 22 miles vs. 38 miles, p=<0.001) and temporal practice dispersion (median driving time 27 minutes vs. 43 minutes, p<0.001), whereas RO had a larger geographic dispersion (median RO driving distance 58 minutes, p<0.001) and temporal practice dispersion (median RO driving time 63 minutes, p<0.001). Conclusions: Oncologic specialties vary in the number of practice sites and practice dispersion per oncologist. In particular, GO and RO practice at more sites than MO, with MO practices more geographically concentrated than RO practices. While SO practice at a similar number of practice sites compared to MO, their practices are the most geographically concentrated. While these practice patterns may represent increased patient access to specialty oncology care, the impact on quality of care and physician wellness is unknown.[Table: see text]

Electronic symptom monitoring in pediatric patients hospitalized for chemotherapy

BACKGROUND

Using patient-reported outcomes for symptom monitoring in oncology has resulted in significant benefits for adult patients with cancer. The feasibility of this approach has not been established in the routine care of children with cancer.

METHODS

The Pediatric Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (Ped-PRO-CTCAE) is an item library that enables children and caregivers to self-report symptoms. Ten symptom items from the Ped-PRO-CTCAE were uploaded to an online platform. Patients at least 7 years old and their caregivers were prompted by text/email message to electronically self-report daily during a planned hospitalization for chemotherapy administration. Symptom reports were emailed to the clinical team caring for the patient, but no instructions were given regarding the use of this information. Rates of patient participation and clinician responses to reports were systematically tracked.

RESULTS

The median age of the participating patients (n = 52) was 11 years (range, 7-18 years). All patients and caregivers completed an initial login, with 92% of dyads completing at least 1 additional symptom assessment during hospitalization (median, 3 assessments; range, 0-40). Eighty-one percent of participating dyads submitted symptom reports on at least half of hospital days, and 54% submitted reports on all hospital days. Clinical actions were taken in response to symptom reports 21% of the time. Most patients felt that the system was easy (73%) and important (79%). Most clinicians found symptom reports easy to understand and useful (97%).

CONCLUSIONS

Symptom monitoring using patient-reported outcome measures for hospitalized pediatric oncology patients is feasible and generates data valued by clinicians and patients.

NCCN Guidelines Insights: Prostate Cancer, Version 1.2021

The NCCN Guidelines for Prostate Cancer address staging and risk assessment after a prostate cancer diagnosis and include management options for localized, regional, and metastatic disease. Recommendations for disease monitoring and treatment of recurrent disease are also included. The NCCN Prostate Cancer Panel meets annually to reevaluate and update their recommendations based on new clinical data and input from within NCCN Member Institutions and from external entities. This article summarizes the panel’s discussions for the 2021 update of the guidelines with regard to systemic therapy for metastatic castration-resistant prostate cancer.

Association of Utilization Management Policy With Uptake of Hypofractionated Radiotherapy Among Patients With Early-Stage Breast Cancer

Importance

Breast cancer accounts for the largest portion of cancer-related spending in the United States. Although hypofractionated radiotherapy after breast-conserving surgery is a cost-effective and convenient treatment strategy for patients with early-stage breast cancer, less than 40% of eligible women received hypofractionated radiotherapy in 2013.

Objective

To assess the association of a large commercial payer's utilization management policy with the use of hypofractionated radiotherapy among women with early-stage breast cancer and its associated cost.

Design, Setting, and Participants

A retrospective, adjusted difference-in-differences economic analysis was conducted using administrative claims data from January 1, 2012, to June 1, 2018, of women 18 years or older with early-stage breast cancer who were eligible for hypofractionated radiotherapy according to 2011 guidelines from the American Society for Radiation Oncology and were continuously enrolled in 14 geographically diverse commercial health plans covering 6.9% of US adult women. Women who received mastectomy, brachytherapy, or less than 11 or more than 40 external beam fractions of radiotherapy were excluded. A utilization management policy was used to encourage the use of hypofractionated radiotherapy among women in fully insured and Medicare Advantage (fully insured) plans. Under the new policy, claims for extended-course radiotherapy were not reimbursed for fully insured women who were eligible for hypofractionated radiotherapy. This policy did not apply to women in self-insured or Medicare supplemental insurance (self-insured) plans, allowing these groups to serve as a comparison group.

Main Outcomes and Measures

The primary outcome was use of hypofractionated radiotherapy, and the secondary outcome was the cost of this type of radiotherapy.

Results

Of 10 540 eligible women, 3619 (34.3%) were in fully insured plans and thus subject to the policy. There were no meaningful differences between the fully insured and self-insured groups in mean (SD) age at the start of radiotherapy (63.8 [8.6] vs 65.0 [8.9] years), mean (SD) Charlson Comorbidity Index score (3.0 [1.5] vs 3.2 [1.6]), or practice setting (outpatient hospital setting, 2982 of 3619 [82.4%] vs 5600 of 6921 [80.9%]). The policy was associated with an increase in use of hypofractionated radiotherapy among fully insured patients subject to the policy (adjusted percentage point difference-in-difference, 4.2%; 95% CI, 0.0%-8.4%; P = .05) and a nonsignificant decrease in radiotherapy-associated expenditures (-$2275 relative to self-insured patients; P = .09). Spillover analyses revealed a significantly higher uptake of hypofractionated radiotherapy among self-insured patients who were indirectly exposed to the policy (adjusted percentage point difference-in-difference, 8.5%; 95% CI, 3.6%-13.5%; P < .001) compared with those who were not exposed.

Conclusions and Relevance

This study suggests that a payer's utilization management policy was associated with direct and spillover increases in the use of hypofractionated radiotherapy, even after accounting for a long-term secular trend in the uptake of hypofractionated radiotherapy in the control groups. Utilization management may promote evidence-based cancer care.

Comparative Effectiveness of Proton vs Photon Therapy as Part of Concurrent Chemoradiotherapy for Locally Advanced Cancer

Importance

Concurrent chemoradiotherapy is the standard-of-care curative treatment for many cancers but is associated with substantial morbidity. Concurrent chemoradiotherapy administered with proton therapy might reduce toxicity and achieve comparable cancer control outcomes compared with conventional photon radiotherapy by reducing the radiation dose to normal tissues.

Objective

To assess whether proton therapy in the setting of concurrent chemoradiotherapy is associated with fewer 90-day unplanned hospitalizations (Common Terminology Criteria for Adverse Events, version 4 [CTCAEv4], grade ≥3) or other adverse events and similar disease-free and overall survival compared with concurrent photon therapy and chemoradiotherapy.

Design, Setting, and Participants

This retrospective, nonrandomized comparative effectiveness study included 1483 adult patients with nonmetastatic, locally advanced cancer treated with concurrent chemoradiotherapy with curative intent from January 1, 2011, through December 31, 2016, at a large academic health system. Three hundred ninety-one patients received proton therapy and 1092, photon therapy. Data were analyzed from October 15, 2018, through February 1, 2019.

Interventions

Proton vs photon chemoradiotherapy.

Main Outcomes and Measures

The primary end point was 90-day adverse events associated with unplanned hospitalizations (CTCAEv4 grade ≥3). Secondary end points included Eastern Cooperative Oncology Group (ECOG) performance status decline during treatment, 90-day adverse events of at least CTCAEv4 grade 2 that limit instrumental activities of daily living, and disease-free and overall survival. Data on adverse events and survival were gathered prospectively. Modified Poisson regression models with inverse propensity score weighting were used to model adverse event outcomes, and Cox proportional hazards regression models with weighting were used for survival outcomes. Propensity scores were estimated using an ensemble machine-learning approach.

Results

Among the 1483 patients included in the analysis (935 men [63.0%]; median age, 62 [range, 18-93] years), those receiving proton therapy were significantly older (median age, 66 [range, 18-93] vs 61 [range, 19-91] years; P < .01), had less favorable Charlson-Deyo comorbidity scores (median, 3.0 vs 2.0; P < .01), and had lower integral radiation dose to tissues outside the target (mean [SD] volume, 14.1 [6.4] vs 19.1 [10.6] cGy/cc × 107; P < .01). Baseline grade ≥2 toxicity (22% vs 24%; P = .37) and ECOG performance status (mean [SD], 0.62 [0.74] vs 0.68 [0.80]; P = .16) were similar between the 2 cohorts. In propensity score weighted-analyses, proton chemoradiotherapy was associated with a significantly lower relative risk of 90-day adverse events of at least grade 3 (0.31; 95% CI, 0.15-0.66; P = .002), 90-day adverse events of at least grade 2 (0.78; 95% CI, 0.65-0.93; P = .006), and decline in performance status during treatment (0.51; 95% CI, 0.37-0.71; P < .001). There was no difference in disease-free or overall survival.

Conclusions and Relevance

In this analysis, proton chemoradiotherapy was associated with significantly reduced acute adverse events that caused unplanned hospitalizations, with similar disease-free and overall survival. Prospective trials are warranted to validate these results.

Effect of Integrating Machine Learning Mortality Estimates With Behavioral Nudges to Clinicians on Serious Illness Conversations Among Patients With Cancer: A Stepped-Wedge Cluster Randomized Clinical Trial

Importance

Serious illness conversations (SICs) are structured conversations between clinicians and patients about prognosis, treatment goals, and end-of-life preferences. Interventions that increase the rate of SICs between oncology clinicians and patients may improve goal-concordant care and patient outcomes.

Objective

To determine the effect of a clinician-directed intervention integrating machine learning mortality predictions with behavioral nudges on motivating clinician-patient SICs.

Design, Setting, and Participants

This stepped-wedge cluster randomized clinical trial was conducted across 20 weeks (from June 17 to November 1, 2019) at 9 medical oncology clinics (8 subspecialty oncology and 1 general oncology clinics) within a large academic health system in Pennsylvania. Clinicians at the 2 smallest subspecialty clinics were grouped together, resulting in 8 clinic groups randomly assigned to the 4 intervention wedge periods. Included participants in the intention-to-treat analyses were 78 oncology clinicians who received SIC training and their patients (N = 14 607) who had an outpatient oncology encounter during the study period.

Interventions

(1) Weekly emails to oncology clinicians with SIC performance feedback and peer comparisons; (2) a list of up to 6 high-risk patients (≥10% predicted risk of 180-day mortality) scheduled for the next week, estimated using a validated machine learning algorithm; and (3) opt-out text message prompts to clinicians on the patient's appointment day to consider an SIC. Clinicians in the control group received usual care consisting of weekly emails with cumulative SIC performance.

Main Outcomes and Measures

Percentage of patient encounters with an SIC in the intervention group vs the usual care (control) group.

Results

The sample consisted of 78 clinicians and 14 607 patients. The mean (SD) age of patients was 61.9 (14.2) years, 53.7% were female, and 70.4% were White. For all encounters, SICs were conducted among 1.3% in the control group and 4.6% in the intervention group, a significant difference (adjusted difference in percentage points, 3.3; 95% CI, 2.3-4.5; P < .001). Among 4124 high-risk patient encounters, SICs were conducted among 3.6% in the control group and 15.2% in the intervention group, a significant difference (adjusted difference in percentage points, 11.6; 95% CI, 8.2-12.5; P < .001).

Conclusions and Relevance

In this stepped-wedge cluster randomized clinical trial, an intervention that delivered machine learning mortality predictions with behavioral nudges to oncology clinicians significantly increased the rate of SICs among all patients and among patients with high mortality risk who were targeted by the intervention. Behavioral nudges combined with machine learning mortality predictions can positively influence clinician behavior and may be applied more broadly to improve care near the end of life.

Trial Registration

ClinicalTrials.gov Identifier: NCT03984773.

Association Between a National Insurer's Pay-for-Performance Program for Oncology and Changes in Prescribing of Evidence-Based Cancer Drugs and Spending

PURPOSE

Cancer drug prescribing by medical oncologists accounts for the greatest variation in practice and the largest portion of spending on cancer care. We evaluated the association between a national commercial insurer's ongoing pay-for-performance (P4P) program for oncology and changes in the prescribing of evidence-based cancer drugs and spending.

METHODS

We conducted an observational difference-in-differences study using administrative claims data covering 6.7% of US adults. We leveraged the geographically staggered, time-varying rollout of the P4P program to simulate a stepped-wedge study design. We included patients age 18 years or older with breast, colon, or lung cancer who were prescribed cancer drug regimens by 1,867 participating oncologists between 2013 and 2017. The exposure was a time-varying dichotomous variable equal to 1 for patients who were prescribed a cancer drug regimen after the P4P program was offered. The primary outcome was whether a patient's drug regimen was a program-endorsed, evidence-based regimen. We also evaluated spending over a 6-month episode period.

RESULTS

The P4P program was associated with an increase in evidence-based regimen prescribing from 57.1% of patients in the preintervention period to 62.2% in the intervention period, for a difference of +5.1 percentage point (95% CI, 3.0 percentage points to 7.2 percentage points; P < .001). The P4P program was also associated with a differential $3,339 (95% CI, $1,121 to $5,557; P = .003) increase in cancer drug spending and a differential $253 (95% CI, $100 to $406; P = .001) increase in patient out-of-pocket spending, but no significant changes in total health care spending ($2,772; 95% CI, -$181 to $5,725; P = .07) over the 6-month episode period.

CONCLUSION

P4P programs may be effective in increasing evidence-based cancer drug prescribing, but may not yield cost savings.

Trends in Radiation Therapy for Bone Metastases, 2015 to 2017: Choosing Wisely in the Era of Complex Radiation

PURPOSE

Guidelines recommend short-course (≤10 fractions) external-beam radiation therapy (EBRT) for bone metastases. Stereotactic body radiation therapy (SBRT) may also improve outcomes; however, routine use is not recommended outside clinical trials. We assessed national radiation therapy trends in complex techniques for bone metastases and associated expenditures.

METHODS AND MATERIALS

Using a claims-based Medicare data set covering 84% of beneficiaries, we assessed the relative proportion of all radiation episodes represented by bone metastases. We then evaluated use of short-course and long-course (>10 fractions) EBRT, intensity modulated radiation therapy (IMRT), and SBRT for bone metastases in hospital-affiliated outpatient (OPD) or freestanding (FREE) facilities. We assessed differences using χ²d or Wilcoxon rank sum tests for categorical and continuous variables, respectively. We identified associations with modality, fractionation, and expenditures using multivariable logistic/linear regression.

RESULTS

Among 467,781 radiation episodes for 17 cancer diagnoses, the overall proportion of episodes dedicated to bone metastases (9.4%) was stable from 2015 to 2017, although treatments were increasing in the hospital-affiliated outpatient setting (P < .005). We identified 40,993 episodes for bone metastases, of which 63% were short-course EBRT, 24% were long-course EBRT, 7% were SBRT, and 6% were IMRT. Techniques more common in the hospital-affiliated outpatient setting included short-course EBRT (OPD, 69%, vs FREE, 56%) and SBRT (OPD, 9%, vs FREE, 5%). Techniques more common among free-standing centers included long-course EBRT (OPD, 19%, vs FREE, 31%) and IMRT (OPD, 4%, vs FREE, 9%). From 2015 to 2017, long-course EBRT decreased by an absolute 8%; short-course EBRT, SBRT, and IMRT increased by 4%, 2.5%, and 1%, respectively. The SBRT/IMRT uptake did not differ by setting (P = .4). Differences in expenditures between SBRT and short-course EBRT decreased by a relative 8% in professional and 12% in technical fees.

CONCLUSIONS

Approximately 1 in 4 patients received long-course EBRT, with small reductions in use largely replaced by complex treatment modalities. However, expenditures for complex modalities also decreased over time. As alternative payment models take effect, quality metrics are needed to ensure appropriate, effective, and safe delivery of complex technologies.

Validation of a Machine Learning Algorithm to Predict 180-Day Mortality for Outpatients With Cancer

Importance

Machine learning (ML) algorithms can identify patients with cancer at risk of short-term mortality to inform treatment and advance care planning. However, no ML mortality risk prediction algorithm has been prospectively validated in oncology or compared with routinely used prognostic indices.

Objective

To validate an electronic health record-embedded ML algorithm that generated real-time predictions of 180-day mortality risk in a general oncology cohort.

Design, Setting, and Participants

This prognostic study comprised a prospective cohort of patients with outpatient oncology encounters between March 1, 2019, and April 30, 2019. An ML algorithm, trained on retrospective data from a subset of practices, predicted 180-day mortality risk between 4 and 8 days before a patient's encounter. Patient encounters took place in 18 medical or gynecologic oncology practices, including 1 tertiary practice and 17 general oncology practices, within a large US academic health care system. Patients aged 18 years or older with outpatient oncology or hematology and oncology encounters were included in the analysis. Patients were excluded if their appointment was scheduled after weekly predictions were generated and if they were only evaluated in benign hematology, palliative care, or rehabilitation practices.

Exposures

Gradient-boosting ML binary classifier.

Main Outcomes and Measures

The primary outcome was the patients' 180-day mortality from the index encounter. The primary performance metric was the area under the receiver operating characteristic curve (AUC).

Results

Among 24 582 patients, 1022 (4.2%) died within 180 days of their index encounter. Their median (interquartile range) age was 64.6 (53.6-73.2) years, 15 319 (62.3%) were women, 18 015 (76.0%) were White, and 10 658 (43.4%) were seen in the tertiary practice. The AUC was 0.89 (95% CI, 0.88-0.90) for the full cohort. The AUC varied across disease-specific groups within the tertiary practice (AUC ranging from 0.74 to 0.96) but was similar between the tertiary and general oncology practices. At a prespecified 40% mortality risk threshold used to differentiate high- vs low-risk patients, observed 180-day mortality was 45.2% (95% CI, 41.3%-49.1%) in the high-risk group vs 3.1% (95% CI, 2.9%-3.3%) in the low-risk group. Integrating the algorithm into the Eastern Cooperative Oncology Group and Elixhauser comorbidity index-based classifiers resulted in favorable reclassification (net reclassification index, 0.09 [95% CI, 0.04-0.14] and 0.23 [95% CI, 0.20-0.27], respectively).

Conclusions and Relevance

In this prognostic study, an ML algorithm was feasibly integrated into the electronic health record to generate real-time, accurate predictions of short-term mortality for patients with cancer and outperformed routinely used prognostic indices. This algorithm may be used to inform behavioral interventions and prompt earlier conversations about goals of care and end-of-life preferences among patients with cancer.

NCCN Patient Advocacy Summit: Delivering Value for Patients Across the Oncology Ecosystem

As the oncology ecosystem shifts from service-based care to outcomes and value-based care, stakeholders cite concerns regarding the lack of patient experience data that are important to the patient community. To address the patient perspective and highlight the challenges and opportunities within policy and clinical decision-making to improve patient-centered care, NCCN hosted the NCCN Patient Advocacy Summit: Delivering Value for Patients Across the Oncology Ecosystem on December 11, 2019, in Washington, DC. The summit featured multidisciplinary panel discussions, keynote speakers, and patient advocate presentations exploring the implications for patient-centered care within a shifting health policy landscape. This article encapsulates and expounds upon the discussions and presentations from the summit.

Site engagement for multi-site clinical trials

Multi-site clinical trials are essential within medical practice to help drive reliable and generalizable knowledge on advancing medical treatments. Although the success of multi-site trials is significantly dependent on local clinician and site research teams, best practices for engagement of site teams, or "site engagement," has not been extensively discussed. Site engagement centers on including sites in the planning and implementation of clinical trials to promote trial enrollment, compliance, and applicability to local contexts. Using a case example from the RadComp Trial, a longitudinal, multi-site clinical trial, novel site engagement practices are provided across three major research phases. In the Planning Phase, site engagement builds partnerships and commitment by active elicitation of information on site specific processes and feedback on trial design. In the Conducting Phase, sustained engagement encourages bi-directional communication and facilitates learning networks for enhanced site performance. In the Dissemination Phase, site and community partnerships are leveraged to create locally designed dissemination plans for broader scientific reach and impact. Site engagement practices discussed in this paper can be replicated or molded for application in other multi-site clinical trials.