Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Health Economics Research in Cancer Treatment: Current Challenges and Future Directions

Health Economics Research in Cancer Treatment: Current Challenges and Future Directions Abstract The National Cancer Institute Division of Cancer Control and Population Science hosted a virtual conference on the Future of Cancer Health Economics Research and included a presentation from a workgroup that considered current challenges and future directions in health economics research centered on cancer treatment. The workgroup identified 3 broad categories of focus: data limitations, opportunities for training for clinicians and health economists interested in collaboration, and the need for prospective economic study of cancer treatment. Within these areas of focus, the workgroup recommended the following: improvement of the availability of key economic measures in data available to researchers, creation of more comprehensive datasets robust to insurance type or coverage, development of cancer care health economics research-focused symposia, instituting clear mechanisms to support integration of economic analyses alongside clinical trials, development of standardized methods to measure the cost of cancer care to health-care systems and patients, and development of standardized evaluations that include measures of social determinants of health. Background Cancer treatment lies at the center of the health economics research framework (Figure 1) (1). Health economics research that focuses on cancer treatment evaluates how patients, health-care providers, insurers, and governments make decisions with respect to cancer treatment using economic theory, models, and empirical methods that can involve analyses of observational data sources such as insurance payments or primary data collection. More comprehensive and timely observational data are needed. Additionally, prospective evaluations of economic factors alongside clinical trials of novel therapies are also needed. Given the increasing expense of cancer treatments (2,3) and growth in the number of cancer patients and survivors (4), health economics research of cancer treatment is critically important to inform patients, health-care providers and payers, and the public about the relative economic value of cancer treatments. Figure 1. Open in new tabDownload slide Cancer health economics research framework. (Adapted from Halpern et al. A framework for cancer health economics research. Cancer. 2021;127:994-996.1 QALY = quality adjusted life year. Figure 1. Open in new tabDownload slide Cancer health economics research framework. (Adapted from Halpern et al. A framework for cancer health economics research. Cancer. 2021;127:994-996.1 QALY = quality adjusted life year. Cancer treatment is diverse and evolves quickly. Multiple types of cancer treatment (surgery, radiation therapy, oral and intravenous chemotherapy, immunotherapy, molecularly targeted therapy, cellular therapy, nonradiotherapy locally ablative treatment, etc.) exist in combinations and sequences that vary by cancer site, stage, and molecular characteristics. Furthermore, novel treatments such as molecularly targeted and immune-modulating therapies are rapidly introduced, with thousands of potential treatments in the development pipeline (5). Even for treatments with extensive existing literature such as radiotherapy, treatment intensity, and duration are often varied to balance convenience, quality of life, and effectiveness. The rapid pace of innovation and change occurs in an environment with constrained national health-care resources and can result in high cost burdens for patients (6). To make the best health-care decisions regarding cancer treatment, patients, providers, insurers, and payers need access to reliable trustworthy information about the economic value and effectiveness of cancer treatments. In light of these issues, a workgroup was convened to develop a subsequent panel presentation for the National Cancer Institute (NCI)–hosted conference on the Future of Cancer Health Economics Research about the current challenges and recommendations for future directions in health economics research investigating cancer treatment. This document summarizes discussions leading up to the workshop as well as the presentation and discussion from December 2, 2020. The workgroup, led by Dr Robin Yabroff, included Drs Deborah Schrag, Henry J. Henk, Stephanie B. Wheeler, Ya-Chen Tina Shih, Marc Smaldone, and James Yu. The workgroup identified 3 broad categories of focus: data limitations, opportunities for training for clinicians and health economists interested in collaboration, and the need for prospective economic analysis of treatments being evaluated in clinical trials. Data Limitations in Conducting Health Economics Research Focused on Cancer Treatment Critical to any health economics research of cancer treatment is the availability of contemporary and comprehensive data characterizing either expenditures or resource utilization that can be converted to expenditures. Most commonly, administrative data from public and private health insurers are used to evaluate cancer treatment expenditures. Because cancer is a disease associated with aging and because the Medicare program insures almost all Americans older than age 65 years, Medicare claims are the most important source for characterizing cancer treatment. The linkage of Medicare data to NCI’s Surveillance, Epidemiology, and End Results (SEER) data makes measurement of the costs and evaluation of factors associated with cancer treatment for specific tumor and treatment types relatively straightforward, as the population-based SEER registries collect information about cancer diagnoses and survival among a representative sample of the US population. Examples of potential data resources in health economics research are shown on Table 1. Though these data are quite useful for some studies, there are critical gaps in available data, especially for researchers evaluating aspects of cancer treatments. Notably, information about patient eligibility for treatment or key measures of treatment outcomes (other than survival), including response, recurrence and location of recurrence, or health-related quality of life, are not currently available from the data sources listed (7-9). Though technically the International Classification of Diseases version 10 diagnosis codes may indicate secondary malignant neoplasms, prior work has shown these codes have limited utility for inferring stage and/or recurrence because of limited sensitivity and specificity (7–9). In addition, there is a lag in data availability. This means that economic evaluations of new treatment approaches that rely on administrative data are typically delayed by several years. The workgroup also noted key clinical and economic measures that would be helpful for health economics research (Table 2). Table 1. Characteristics of selected data sources and research resources in the US for health economics research in cancer treatment . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ Data on Medicare Part D prescription drug services are available starting in 2006. MEPS = Medical Expenditure Panel Survey; SEER = Surveillance Epidemiology and End Results. Diagnosis codes exist for secondary malignant neoplasms, although caution is advised for interpretation as disease recurrence, even with careful analysis. Open in new tab Table 1. Characteristics of selected data sources and research resources in the US for health economics research in cancer treatment . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ Data on Medicare Part D prescription drug services are available starting in 2006. MEPS = Medical Expenditure Panel Survey; SEER = Surveillance Epidemiology and End Results. Diagnosis codes exist for secondary malignant neoplasms, although caution is advised for interpretation as disease recurrence, even with careful analysis. Open in new tab Table 2. Key measures for health economic research of cancer treatment Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Open in new tab Table 2. Key measures for health economic research of cancer treatment Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Open in new tab The reliance of most economic studies of cancer treatment on existing health insurance claims data means that inferences regarding cancer treatment choice can be clouded by rapid and frequent changes in insurance benefit design or by changes in plans among working-age adults. Furthermore, the data linkages to insurance claims (eg, SEER-Medicare) provide information about cancer characteristics and treatment only for select populations (such as those enrolled in Medicare fee-for-service). Detailed information about treatment and many outcomes end when health insurance coverage ends or changes, especially for the population aged younger than 65 years who are not age eligible for Medicare coverage. For Medicaid, median enrollment in state-federal Medicaid programs is less than 10 months, which means that these data will be limited for evaluating longer-term patterns of care (10). Low-income adults in particular are at risk for changing insurance coverage due to job-related changes and loss of eligibility for Medicaid or Affordable Care Act Marketplace subsidies (11). Furthermore, the transitions in coverage can be meaningful in and of themselves (switch to disability, unemployed because of cancer, and loss of employer-sponsored coverage). In addition to a lack of comprehensive data on measures, the workgroup noted a lack of real-time data, making health economics research on contemporary practice patterns more difficult. Beyond health economics research, informing current practice with data obtained in a continuous and up-to-date manner is the goal of a learning health-care system, which is defined as a health-care system in which knowledge is continuously generated and cycled back into daily practice, which then creates more knowledge regarding best practice in cycle of continual improvement and learning (12). Health economics research of cancer treatment plays a critical part of this learning health-care system by taking observational and administrative data from real-world cancer treatment and analyzing these data to inform best practices. Timeliness of analysis is especially challenging with rapid changes in treatments. Further, as innovations in cancer treatment do not occur uniformly for all people and providers, there are trade-offs between timeliness of analysis and representativeness of the data analyzed. Finally, reliance on historical data is less useful for research to inform modern changes in benefit design (eg, high deductible health insurance plans, bundled payment, value-based payment) that may not have historical precedents. To address these limitations, the workgroup prioritized improving capacity for collection of economic data from large prospective cancer cohorts and clinical trials with streamlined and standardized economic data collection at baseline and defined follow-up. This approach could facilitate understanding patient’s direct medical expenses as well as indirect expenses for alternative treatment approaches. However, the workgroup did not have a specific recommendation as to what should be included in streamlined and standardized economic data collection. Opportunities in Training of Clinicians and Health Economists The workgroup recognized a need for training and engagement of clinicians without a formal health economics background and for health services researchers and economists without formal clinical training. Health economics research that informs clinical practice requires the marriage of diverse skillsets that cross disciplinary boundaries. Economic theory, methods, and analytic techniques are not taught in medical school, nursing school, or other clinically focused training programs, and therefore, to understand and measure economic inputs and outcomes, clinicians require further education to be knowledgeable research partners. Training intensity of clinicians can vary depending on the end goal. Fellowship-level training or NCI career development awards are available for clinicians who are interested in a career in health economics. More formally, health economics master of public health programs exist for those who can commit the time and effort required for completion. Short courses and intensive workshops can equip physician researchers with the skills necessary, but partnerships with formally trained economists or research methodologists are optimal. NCI and academic medical centers can use their convening power to create opportunities for investigators with complementary skill sets to form partnerships. Given the complexity of clinical cancer care, further efforts beyond those discussed at the workshop may be needed. For health services researchers and health economists, clinical knowledge of cancer treatment is difficult to obtain. As obtaining a medical degree and completing a residency and fellowship are rarely feasible for health economics researchers, comprehensive knowledge of the rapidly changing care across the cancer control continuum will likely require partnering with clinicians. Again, focused meetings and collaborative seminars may be an optimal combination of efficiency and timeliness. Suggestions for meetings and seminars to increase collaboration included the creation of specific sessions for cancer health economics research at both clinical (such as the American Society for Clinical Oncology, American Society for Radiation Oncology, and American Association for Cancer Research) and economic and health services research (such as Academy Health and American Society for Health Economics) meetings. The Cancer Care Delivery Research committees of National Clinical Trials Network groups are potentially fertile places for collaboration between clinicians and health economists. Finally, dedicated annual or biennial meetings of cancer health economics was also recommended, to bring focused attention to health economics research. Professional organizations (such as the American Society for Clinical Oncology, Academy Health, and American Society of Health Economists) can play a critical role in facilitating these collaborations. Economic Analyses Alongside Clinical Trials Prospective clinical trials often address whether a new treatment will be an improvement over an older standard of care. Randomized controlled trials still offer most unbiased estimate of whether a treatment can work. However, randomized controlled trials are rarely designed to answer whether the new treatment has incremental economic value and often do not concern themselves with the actual cost of the new treatment. Costs can encompass both medical and nonmedical direct costs, such as the cost of treatment and the cost of transportation to the treatment, as well as indirect costs, such as premature death or the need to take early retirement. Economic analysis aims to assess all of these costs to answer the following questions: Is a new treatment worthwhile? And worthwhile to whom? When asking whether a new treatment is worthwhile, health economics research also asks the following: Are anticipated differences in economic resource utilization meaningful from a societal perspective? Will adding an economic component influence clinical practice or health policy? Is collection of good economic data feasible within the context of the overall trial design? Does the trial design have external validity from an economic perspective? To perform health economics research within a randomized study, the researcher must have data that captures baseline information (ideally including direct and indirect costs) on all participants, tracks resource utilization (especially big-ticket items like hospitalizations) for participants, and estimates medical costs from resource utilization. These data are best collected prospectively. Despite the advantages of embedding health economics research into randomized trials, the NCI Cancer Therapy Evaluation Program and Division of Cancer Prevention do not provide funding for this research, and proposals to integrate economic companions alongside clinical trials are rarely approved. Similarly, the NCI Biomarker, Imaging and Quality of Life Studies Funding Program, which is used to fund correlative science associated with NCI clinical trial concepts or protocols, cannot be used to fund comparative effectiveness or cost-focused studies. Therefore, researchers interested in health economics can be stymied in their ability to obtain funding to support these studies although the incremental costs of economic analyses are typically very small in relationship to clinical trial costs. Finally, although prospective economic data are ideal, they may be difficult to directly obtain within clinical trials. Further, prospectively collected economic data may be limited by participation bias, resulting in potentially misleading research findings if this bias is not adequately adjusted for or anticipated. Other methods to obtain economic data alongside clinical trials exist when prospective data collection is not possible within the trial itself. Data on the direct medical costs of clinical trial interventions can be obtained by measuring resource utilization from electronic health records or patient reporting. Administrative data can also provide estimates of direct medical cost accrued by patients. Indirect medical cost estimates can be found by surveying trial participants, caregivers, and providers using common data forms. These data can be collected prospectively during the course of the trial though this often requires additional funding through sources outside the NCI and National Institutes of Health. Recommendations for Health Economics Research in Cancer Treatment Fundamental questions in health economics include questions of incremental cost effectiveness, true health system and patient costs of treatment (direct and indirect costs), measurement of productivity loss (eg, days lost from work), and the influence of economic factors on treatment itself. To answer these fundamental questions, the following interventions are recommended. Improve availability of key economic measures when designing or updating existing or planned data registries (Table 2) through standardized economic data collection at baseline and follow-up. Ideally, this should include social determinants of health. Create more comprehensive datasets that include patients from all payers (eg, all payer claims database). Increase availability of real-time data to researchers to better inform contemporary care delivery within a learning health-care system. Creation of health economics–focused symposia within major cancer treatment society meetings as well as meetings dedicated to health economics research to bring focused attention to health economics research and cancer care. Provide a clear mechanism to support integration of economic analyses that rely on standardized methods and approaches to data collection alongside clinical trials of cancer treatment. Develop transparent and standardized measurement methods of the cost of care to health-care systems that facilitate comparison across treatments. Develop and deploy standard methods to estimate the economic burden of treatment on patients in terms of time, out-of-pocket costs, and productivity. Discussion In conclusion, the workgroup identified the need to address data limitations, improve collaborative opportunities for clinicians and health economists, and increase prospective economic study within and alongside clinical trials as 3 areas of focus for improving health economics research of cancer treatment. To address these areas, the workgroup recommended the following: improvement of the availability of key economic data, creation of datasets robust to insurance changes, development of focused symposia, instituting clear mechanisms to support prospective economic analyses alongside clinical trials, development of standardized methods to measure the cost of cancer care, and development of standardized evaluations that include measures of social determinants of health. Funding No funding was used for this study. Notes Role of the funder: Not applicable. Disclosures: James B. Yu, MD, MHS, reports personal fees (speaking and consulting) from Boston Scientific and personal fees (advisory board) from Galera Pharmaceuticals. Dr Schrag acknowledges funding from the National Cancer Institutes (1UM1 CA233035-01). Dr Schrag received compensation for speaking at a Pfizer satellite symposium in 2019, receives services for editorial work for JAMA, and obtained research funding from the AACR for project GENIE. Dr Yabroff serves on the Flatiron Health Equity Advisory Board. Author contributions: All authors contributed to the conceptualization and writing of the original draft, as well as reviewing and editing the manuscript. Acknowledgments: The authors would like to acknowledge Drs Henry J. Henk, Stephanie B. Wheeler, Tina Shih, and Marc Smaldone for their participation in the cancer treatment workgroup, which this manuscript summarizes. References 1 Halpern MT , Shi YT, Yabroff KR, et al. A framework for cancer health economics research . Cancer . 2021 ; 127 ( 7 ): 994 – 996 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Yang EJ , Galan E, Thombley R, et al. Changes in drug list prices and amounts paid by patients and insurers . JAMA Netw Open . 2020 ; 3 ( 12 ): e2028510 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Leighl NB , Nirmalakumar S, Ezeife DA, Gyawali B. An arm and a leg: the rising cost of cancer drugs and impact on access . Am Soc Clin Oncol Educ Book . 2021 ; 41 : 1 – 12 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 4 Mariotto AB , Enewold L, Zhao J, Zeruto CA, Yabroff KR. Medical care costs associated with cancer survivorship in the United States . Cancer Epidemiol Biomarkers Prev . 2020 ; 29 ( 7 ): 1304 – 1312 . Google Scholar Crossref Search ADS PubMed WorldCat 5 IQVIA. Global oncology trends 2021. https://www.iqvia.com/insights/the-iqvia-institute/reports/global-oncology-trends-2021. Accessed August 11, 2021 . 6 Yabroff KR , Bradley C, Shi YT. Understanding financial hardship among cancer survivors in the United States: strategies for prevention and mitigation . J Clin Oncol . 2020 ; 38 ( 4 ): 292 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Warren JL , Yabroff KR. Challenges and opportunities in measuring cancer recurrence in the United States . J Natl Cancer Inst . 2015 ; 107 ( 8 ): djv134 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Warren JL , Mariotto A, Melbert D, et al. Sensitivity of Medicare claims to identify cancer recurrence in elderly colorectal and breast cancer patients . Med Care . 2016 ; 54 ( 8 ): e47 – e54 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Chawla N , Yabroff KR, Mariotto A, McNeel TS, Schrag D, Warren JL. Limited validity of diagnosis codes in Medicare claims for identifying cancer metastases and inferring stage . Ann Epidemiol . 2014 ; 24 ( 9 ): 666 – 672.e1-2 . doi: 10.1016/j.annepidem.2014.06.099 . Epub 2014 Jul 3. Google Scholar Crossref Search ADS PubMed WorldCat 10 Ku L , Steinmetz E, Bysshe T. Continuity of Medicaid Coverage in an Era of Transition . George Washington University ; November 1, 2015 . http://www.communityplans.net/Portals/0/Policy/Medicaid/GW_ContinuityInAnEraOfTransition_11-01-15.pdf. Accessed August 11, 2021. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 11 Sommers BD , Gourevitch R, Maylone B, Blendon RJ, Epstein AM. Insurance churning rates for low-income adults under health reform: lower than expected but still harmful for many . Health Affairs . 2016 ; 35 ( 10 ): 1816 – 1824 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Institute of Medicine. Delivering High-Quality Cancer care: Charting a New Course for a System in Crisis; December 27, 2013 . https://www.ncbi.nlm.nih.gov/books/NBK202148/pdf/Bookshelf_NBK202148.pdf. Accessed August 11, 2021. © The Author(s) 2022. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © The Author(s) 2022. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Monographs Oxford University Press

Health Economics Research in Cancer Treatment: Current Challenges and Future Directions

JNCI Monographs , Volume 2022 (59) – Jul 5, 2022

Loading next page...
 
/lp/oxford-university-press/health-economics-research-in-cancer-treatment-current-challenges-and-k77t6s3fKF
Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
ISSN
1052-6773
eISSN
1745-6614
DOI
10.1093/jncimonographs/lgac009
Publisher site
See Article on Publisher Site

Abstract

Abstract The National Cancer Institute Division of Cancer Control and Population Science hosted a virtual conference on the Future of Cancer Health Economics Research and included a presentation from a workgroup that considered current challenges and future directions in health economics research centered on cancer treatment. The workgroup identified 3 broad categories of focus: data limitations, opportunities for training for clinicians and health economists interested in collaboration, and the need for prospective economic study of cancer treatment. Within these areas of focus, the workgroup recommended the following: improvement of the availability of key economic measures in data available to researchers, creation of more comprehensive datasets robust to insurance type or coverage, development of cancer care health economics research-focused symposia, instituting clear mechanisms to support integration of economic analyses alongside clinical trials, development of standardized methods to measure the cost of cancer care to health-care systems and patients, and development of standardized evaluations that include measures of social determinants of health. Background Cancer treatment lies at the center of the health economics research framework (Figure 1) (1). Health economics research that focuses on cancer treatment evaluates how patients, health-care providers, insurers, and governments make decisions with respect to cancer treatment using economic theory, models, and empirical methods that can involve analyses of observational data sources such as insurance payments or primary data collection. More comprehensive and timely observational data are needed. Additionally, prospective evaluations of economic factors alongside clinical trials of novel therapies are also needed. Given the increasing expense of cancer treatments (2,3) and growth in the number of cancer patients and survivors (4), health economics research of cancer treatment is critically important to inform patients, health-care providers and payers, and the public about the relative economic value of cancer treatments. Figure 1. Open in new tabDownload slide Cancer health economics research framework. (Adapted from Halpern et al. A framework for cancer health economics research. Cancer. 2021;127:994-996.1 QALY = quality adjusted life year. Figure 1. Open in new tabDownload slide Cancer health economics research framework. (Adapted from Halpern et al. A framework for cancer health economics research. Cancer. 2021;127:994-996.1 QALY = quality adjusted life year. Cancer treatment is diverse and evolves quickly. Multiple types of cancer treatment (surgery, radiation therapy, oral and intravenous chemotherapy, immunotherapy, molecularly targeted therapy, cellular therapy, nonradiotherapy locally ablative treatment, etc.) exist in combinations and sequences that vary by cancer site, stage, and molecular characteristics. Furthermore, novel treatments such as molecularly targeted and immune-modulating therapies are rapidly introduced, with thousands of potential treatments in the development pipeline (5). Even for treatments with extensive existing literature such as radiotherapy, treatment intensity, and duration are often varied to balance convenience, quality of life, and effectiveness. The rapid pace of innovation and change occurs in an environment with constrained national health-care resources and can result in high cost burdens for patients (6). To make the best health-care decisions regarding cancer treatment, patients, providers, insurers, and payers need access to reliable trustworthy information about the economic value and effectiveness of cancer treatments. In light of these issues, a workgroup was convened to develop a subsequent panel presentation for the National Cancer Institute (NCI)–hosted conference on the Future of Cancer Health Economics Research about the current challenges and recommendations for future directions in health economics research investigating cancer treatment. This document summarizes discussions leading up to the workshop as well as the presentation and discussion from December 2, 2020. The workgroup, led by Dr Robin Yabroff, included Drs Deborah Schrag, Henry J. Henk, Stephanie B. Wheeler, Ya-Chen Tina Shih, Marc Smaldone, and James Yu. The workgroup identified 3 broad categories of focus: data limitations, opportunities for training for clinicians and health economists interested in collaboration, and the need for prospective economic analysis of treatments being evaluated in clinical trials. Data Limitations in Conducting Health Economics Research Focused on Cancer Treatment Critical to any health economics research of cancer treatment is the availability of contemporary and comprehensive data characterizing either expenditures or resource utilization that can be converted to expenditures. Most commonly, administrative data from public and private health insurers are used to evaluate cancer treatment expenditures. Because cancer is a disease associated with aging and because the Medicare program insures almost all Americans older than age 65 years, Medicare claims are the most important source for characterizing cancer treatment. The linkage of Medicare data to NCI’s Surveillance, Epidemiology, and End Results (SEER) data makes measurement of the costs and evaluation of factors associated with cancer treatment for specific tumor and treatment types relatively straightforward, as the population-based SEER registries collect information about cancer diagnoses and survival among a representative sample of the US population. Examples of potential data resources in health economics research are shown on Table 1. Though these data are quite useful for some studies, there are critical gaps in available data, especially for researchers evaluating aspects of cancer treatments. Notably, information about patient eligibility for treatment or key measures of treatment outcomes (other than survival), including response, recurrence and location of recurrence, or health-related quality of life, are not currently available from the data sources listed (7-9). Though technically the International Classification of Diseases version 10 diagnosis codes may indicate secondary malignant neoplasms, prior work has shown these codes have limited utility for inferring stage and/or recurrence because of limited sensitivity and specificity (7–9). In addition, there is a lag in data availability. This means that economic evaluations of new treatment approaches that rely on administrative data are typically delayed by several years. The workgroup also noted key clinical and economic measures that would be helpful for health economics research (Table 2). Table 1. Characteristics of selected data sources and research resources in the US for health economics research in cancer treatment . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ Data on Medicare Part D prescription drug services are available starting in 2006. MEPS = Medical Expenditure Panel Survey; SEER = Surveillance Epidemiology and End Results. Diagnosis codes exist for secondary malignant neoplasms, although caution is advised for interpretation as disease recurrence, even with careful analysis. Open in new tab Table 1. Characteristics of selected data sources and research resources in the US for health economics research in cancer treatment . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ . SEER-Medicare . Private insurer claims . MEPS . Description SEER tumor registries linked to Medicare claims Administrative claims for patients with specific coverage type Nationally representative in-person survey with provider data collection Data characteristics  National or nationally representative Geographically defined Plan-specific, typically in multiple states √  Individual-level longitudinal data √ Average duration of coverage can vary 2 years  Population-based patients Within registry areas √  Population-based providers Within registry areas √  Population-based facilities Within registry areas √  Approximate magnitude of cancer patients receiving specific treatments annually Very large Potentially large Very small  Duration of information Medicare eligibility through death While covered 2 years Health insurance characteristics  Type of coverage Medicare fee-for-service only (about 60% of beneficiaries in 2018) Plan specific All payers and types  Benefit design Contained in Part D pharmacy only Potential for multiple benefit designs Some characteristics Social determinants of health  Social services spending, area deprivation index, and social capital  Health-care segregation Can be estimated Timeliness Multiyear delay Close to real time Multiyear delay Patient information  Age distribution Aged 65 years or older or disabled (any age) Mostly younger than 65 years Aged 18 years or older  Information about patients without cancer In cancer registry regions √ √ Cancer information  Cancer diagnosis and history Registry, procedure, or diagnosis codes Procedure or diagnosis codes Self-report, procedure, or diagnosis codes  Diagnosis date Diagnosis month  Stage at diagnosis √  Recurrence b b  Vital status and/or date of death √  Death due to cancer √ Cancer treatment information  Biomarker information and targeted treatment eligibility Oncotype diagnosis as a restricted variable; some biomarker information is available Some biomarker information is available  Functional status and physical eligibility for treatment  Receipt of surgery √ During enrollment During 2-year panel  Receipt of chemotherapy √ During enrollment  Receipt of radiation therapy √ During enrollment  Reason for change in treatment Cost domains Direct medical cost components  Hospital √ √ √  Physician and other outpatient services √ √ √  Outpatient pharmacy √ √ √  Out of pocket Part D only √ √ Indirect cost components  Productivity loss (eg, days lost from work) √  Patient time  Caregiver time Intangible costs √ Data on Medicare Part D prescription drug services are available starting in 2006. MEPS = Medical Expenditure Panel Survey; SEER = Surveillance Epidemiology and End Results. Diagnosis codes exist for secondary malignant neoplasms, although caution is advised for interpretation as disease recurrence, even with careful analysis. Open in new tab Table 2. Key measures for health economic research of cancer treatment Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Open in new tab Table 2. Key measures for health economic research of cancer treatment Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Treatment eligibility measures Functional and performance status Tumor molecular phenotype and genetic data Patient genetic data Treatment recommendations and factors affecting decision making Key treatment outcomes Health-related quality of life, treatment-specific utility estimates, and other patient-reported outcomes Treatment intent, reasons for switching, dose changes, and discontinuation Recurrence, recurrence location, and recurrence timing Survival and cause-specific mortality Other patient characteristics Granularity on race and ethnicity Sexual orientation and gender identity Social risk and social needs Nonmedical economic data Financial distress, food and housing insecurity, and other social determinants of health Productivity loss Provider characteristics •Comprehensive information about vital status •Consistent and granular area-level identifiers •Social determinants of health •Informal caregiving data Open in new tab The reliance of most economic studies of cancer treatment on existing health insurance claims data means that inferences regarding cancer treatment choice can be clouded by rapid and frequent changes in insurance benefit design or by changes in plans among working-age adults. Furthermore, the data linkages to insurance claims (eg, SEER-Medicare) provide information about cancer characteristics and treatment only for select populations (such as those enrolled in Medicare fee-for-service). Detailed information about treatment and many outcomes end when health insurance coverage ends or changes, especially for the population aged younger than 65 years who are not age eligible for Medicare coverage. For Medicaid, median enrollment in state-federal Medicaid programs is less than 10 months, which means that these data will be limited for evaluating longer-term patterns of care (10). Low-income adults in particular are at risk for changing insurance coverage due to job-related changes and loss of eligibility for Medicaid or Affordable Care Act Marketplace subsidies (11). Furthermore, the transitions in coverage can be meaningful in and of themselves (switch to disability, unemployed because of cancer, and loss of employer-sponsored coverage). In addition to a lack of comprehensive data on measures, the workgroup noted a lack of real-time data, making health economics research on contemporary practice patterns more difficult. Beyond health economics research, informing current practice with data obtained in a continuous and up-to-date manner is the goal of a learning health-care system, which is defined as a health-care system in which knowledge is continuously generated and cycled back into daily practice, which then creates more knowledge regarding best practice in cycle of continual improvement and learning (12). Health economics research of cancer treatment plays a critical part of this learning health-care system by taking observational and administrative data from real-world cancer treatment and analyzing these data to inform best practices. Timeliness of analysis is especially challenging with rapid changes in treatments. Further, as innovations in cancer treatment do not occur uniformly for all people and providers, there are trade-offs between timeliness of analysis and representativeness of the data analyzed. Finally, reliance on historical data is less useful for research to inform modern changes in benefit design (eg, high deductible health insurance plans, bundled payment, value-based payment) that may not have historical precedents. To address these limitations, the workgroup prioritized improving capacity for collection of economic data from large prospective cancer cohorts and clinical trials with streamlined and standardized economic data collection at baseline and defined follow-up. This approach could facilitate understanding patient’s direct medical expenses as well as indirect expenses for alternative treatment approaches. However, the workgroup did not have a specific recommendation as to what should be included in streamlined and standardized economic data collection. Opportunities in Training of Clinicians and Health Economists The workgroup recognized a need for training and engagement of clinicians without a formal health economics background and for health services researchers and economists without formal clinical training. Health economics research that informs clinical practice requires the marriage of diverse skillsets that cross disciplinary boundaries. Economic theory, methods, and analytic techniques are not taught in medical school, nursing school, or other clinically focused training programs, and therefore, to understand and measure economic inputs and outcomes, clinicians require further education to be knowledgeable research partners. Training intensity of clinicians can vary depending on the end goal. Fellowship-level training or NCI career development awards are available for clinicians who are interested in a career in health economics. More formally, health economics master of public health programs exist for those who can commit the time and effort required for completion. Short courses and intensive workshops can equip physician researchers with the skills necessary, but partnerships with formally trained economists or research methodologists are optimal. NCI and academic medical centers can use their convening power to create opportunities for investigators with complementary skill sets to form partnerships. Given the complexity of clinical cancer care, further efforts beyond those discussed at the workshop may be needed. For health services researchers and health economists, clinical knowledge of cancer treatment is difficult to obtain. As obtaining a medical degree and completing a residency and fellowship are rarely feasible for health economics researchers, comprehensive knowledge of the rapidly changing care across the cancer control continuum will likely require partnering with clinicians. Again, focused meetings and collaborative seminars may be an optimal combination of efficiency and timeliness. Suggestions for meetings and seminars to increase collaboration included the creation of specific sessions for cancer health economics research at both clinical (such as the American Society for Clinical Oncology, American Society for Radiation Oncology, and American Association for Cancer Research) and economic and health services research (such as Academy Health and American Society for Health Economics) meetings. The Cancer Care Delivery Research committees of National Clinical Trials Network groups are potentially fertile places for collaboration between clinicians and health economists. Finally, dedicated annual or biennial meetings of cancer health economics was also recommended, to bring focused attention to health economics research. Professional organizations (such as the American Society for Clinical Oncology, Academy Health, and American Society of Health Economists) can play a critical role in facilitating these collaborations. Economic Analyses Alongside Clinical Trials Prospective clinical trials often address whether a new treatment will be an improvement over an older standard of care. Randomized controlled trials still offer most unbiased estimate of whether a treatment can work. However, randomized controlled trials are rarely designed to answer whether the new treatment has incremental economic value and often do not concern themselves with the actual cost of the new treatment. Costs can encompass both medical and nonmedical direct costs, such as the cost of treatment and the cost of transportation to the treatment, as well as indirect costs, such as premature death or the need to take early retirement. Economic analysis aims to assess all of these costs to answer the following questions: Is a new treatment worthwhile? And worthwhile to whom? When asking whether a new treatment is worthwhile, health economics research also asks the following: Are anticipated differences in economic resource utilization meaningful from a societal perspective? Will adding an economic component influence clinical practice or health policy? Is collection of good economic data feasible within the context of the overall trial design? Does the trial design have external validity from an economic perspective? To perform health economics research within a randomized study, the researcher must have data that captures baseline information (ideally including direct and indirect costs) on all participants, tracks resource utilization (especially big-ticket items like hospitalizations) for participants, and estimates medical costs from resource utilization. These data are best collected prospectively. Despite the advantages of embedding health economics research into randomized trials, the NCI Cancer Therapy Evaluation Program and Division of Cancer Prevention do not provide funding for this research, and proposals to integrate economic companions alongside clinical trials are rarely approved. Similarly, the NCI Biomarker, Imaging and Quality of Life Studies Funding Program, which is used to fund correlative science associated with NCI clinical trial concepts or protocols, cannot be used to fund comparative effectiveness or cost-focused studies. Therefore, researchers interested in health economics can be stymied in their ability to obtain funding to support these studies although the incremental costs of economic analyses are typically very small in relationship to clinical trial costs. Finally, although prospective economic data are ideal, they may be difficult to directly obtain within clinical trials. Further, prospectively collected economic data may be limited by participation bias, resulting in potentially misleading research findings if this bias is not adequately adjusted for or anticipated. Other methods to obtain economic data alongside clinical trials exist when prospective data collection is not possible within the trial itself. Data on the direct medical costs of clinical trial interventions can be obtained by measuring resource utilization from electronic health records or patient reporting. Administrative data can also provide estimates of direct medical cost accrued by patients. Indirect medical cost estimates can be found by surveying trial participants, caregivers, and providers using common data forms. These data can be collected prospectively during the course of the trial though this often requires additional funding through sources outside the NCI and National Institutes of Health. Recommendations for Health Economics Research in Cancer Treatment Fundamental questions in health economics include questions of incremental cost effectiveness, true health system and patient costs of treatment (direct and indirect costs), measurement of productivity loss (eg, days lost from work), and the influence of economic factors on treatment itself. To answer these fundamental questions, the following interventions are recommended. Improve availability of key economic measures when designing or updating existing or planned data registries (Table 2) through standardized economic data collection at baseline and follow-up. Ideally, this should include social determinants of health. Create more comprehensive datasets that include patients from all payers (eg, all payer claims database). Increase availability of real-time data to researchers to better inform contemporary care delivery within a learning health-care system. Creation of health economics–focused symposia within major cancer treatment society meetings as well as meetings dedicated to health economics research to bring focused attention to health economics research and cancer care. Provide a clear mechanism to support integration of economic analyses that rely on standardized methods and approaches to data collection alongside clinical trials of cancer treatment. Develop transparent and standardized measurement methods of the cost of care to health-care systems that facilitate comparison across treatments. Develop and deploy standard methods to estimate the economic burden of treatment on patients in terms of time, out-of-pocket costs, and productivity. Discussion In conclusion, the workgroup identified the need to address data limitations, improve collaborative opportunities for clinicians and health economists, and increase prospective economic study within and alongside clinical trials as 3 areas of focus for improving health economics research of cancer treatment. To address these areas, the workgroup recommended the following: improvement of the availability of key economic data, creation of datasets robust to insurance changes, development of focused symposia, instituting clear mechanisms to support prospective economic analyses alongside clinical trials, development of standardized methods to measure the cost of cancer care, and development of standardized evaluations that include measures of social determinants of health. Funding No funding was used for this study. Notes Role of the funder: Not applicable. Disclosures: James B. Yu, MD, MHS, reports personal fees (speaking and consulting) from Boston Scientific and personal fees (advisory board) from Galera Pharmaceuticals. Dr Schrag acknowledges funding from the National Cancer Institutes (1UM1 CA233035-01). Dr Schrag received compensation for speaking at a Pfizer satellite symposium in 2019, receives services for editorial work for JAMA, and obtained research funding from the AACR for project GENIE. Dr Yabroff serves on the Flatiron Health Equity Advisory Board. Author contributions: All authors contributed to the conceptualization and writing of the original draft, as well as reviewing and editing the manuscript. Acknowledgments: The authors would like to acknowledge Drs Henry J. Henk, Stephanie B. Wheeler, Tina Shih, and Marc Smaldone for their participation in the cancer treatment workgroup, which this manuscript summarizes. References 1 Halpern MT , Shi YT, Yabroff KR, et al. A framework for cancer health economics research . Cancer . 2021 ; 127 ( 7 ): 994 – 996 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Yang EJ , Galan E, Thombley R, et al. Changes in drug list prices and amounts paid by patients and insurers . JAMA Netw Open . 2020 ; 3 ( 12 ): e2028510 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Leighl NB , Nirmalakumar S, Ezeife DA, Gyawali B. An arm and a leg: the rising cost of cancer drugs and impact on access . Am Soc Clin Oncol Educ Book . 2021 ; 41 : 1 – 12 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 4 Mariotto AB , Enewold L, Zhao J, Zeruto CA, Yabroff KR. Medical care costs associated with cancer survivorship in the United States . Cancer Epidemiol Biomarkers Prev . 2020 ; 29 ( 7 ): 1304 – 1312 . Google Scholar Crossref Search ADS PubMed WorldCat 5 IQVIA. Global oncology trends 2021. https://www.iqvia.com/insights/the-iqvia-institute/reports/global-oncology-trends-2021. Accessed August 11, 2021 . 6 Yabroff KR , Bradley C, Shi YT. Understanding financial hardship among cancer survivors in the United States: strategies for prevention and mitigation . J Clin Oncol . 2020 ; 38 ( 4 ): 292 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Warren JL , Yabroff KR. Challenges and opportunities in measuring cancer recurrence in the United States . J Natl Cancer Inst . 2015 ; 107 ( 8 ): djv134 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Warren JL , Mariotto A, Melbert D, et al. Sensitivity of Medicare claims to identify cancer recurrence in elderly colorectal and breast cancer patients . Med Care . 2016 ; 54 ( 8 ): e47 – e54 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Chawla N , Yabroff KR, Mariotto A, McNeel TS, Schrag D, Warren JL. Limited validity of diagnosis codes in Medicare claims for identifying cancer metastases and inferring stage . Ann Epidemiol . 2014 ; 24 ( 9 ): 666 – 672.e1-2 . doi: 10.1016/j.annepidem.2014.06.099 . Epub 2014 Jul 3. Google Scholar Crossref Search ADS PubMed WorldCat 10 Ku L , Steinmetz E, Bysshe T. Continuity of Medicaid Coverage in an Era of Transition . George Washington University ; November 1, 2015 . http://www.communityplans.net/Portals/0/Policy/Medicaid/GW_ContinuityInAnEraOfTransition_11-01-15.pdf. Accessed August 11, 2021. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 11 Sommers BD , Gourevitch R, Maylone B, Blendon RJ, Epstein AM. Insurance churning rates for low-income adults under health reform: lower than expected but still harmful for many . Health Affairs . 2016 ; 35 ( 10 ): 1816 – 1824 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Institute of Medicine. Delivering High-Quality Cancer care: Charting a New Course for a System in Crisis; December 27, 2013 . https://www.ncbi.nlm.nih.gov/books/NBK202148/pdf/Bookshelf_NBK202148.pdf. Accessed August 11, 2021. © The Author(s) 2022. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © The Author(s) 2022. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com

Journal

JNCI MonographsOxford University Press

Published: Jul 5, 2022

References