Abstract: As the cancer care landscape continues to evolve, oncologists must stay up to date on the latest treatment options for their patients while also navigating the administrative complexities of running a successful oncology practice. Additionally, the rising cost of cancer care has contributed to the shift from volume-based to value-based care, often with different payers having different requirements for reporting and reimbursement. Clinical decision support (CDS) systems were developed to help oncologists overcome some of these challenges; however, they have not yet been widely adopted in practice. Through a collaboration between the National Comprehensive Cancer Network (NCCN), Flatiron Health, and oncology practices, an NCCN Guidelines-based CDS system, Flatiron Assist,™ is being piloted as a tool to address some major challenges that oncologists continue to face. Flatiron Assist could help oncology care networks drive standardized care across practices, ease the administrative burden of value-based models, and gather real-world data to better understand costs, quality of care, and patient outcomes, while limiting the need for excessive physician documentation. We also explore persistent barriers to CDS uptake and potential strategies for increasing their utility and use in the field.
Key Words: clinical decision-support, clinical pathways, value-based care, electronic health record, reimbursement
Oncologists face a number of significant challenges in today’s cancer care landscape. They are facing increasing patient care complexities, which requires adapting prescribing habits as new drugs (including biosimilars) and biomarker tests are approved, but they must also be monitoring toxicities, rapidly responding to patient concerns and inquiries, and caring for an increasingly aging population.1 Moreover, oncologists must navigate the constantly evolving administrative complexities of running a successful oncology practice. This includes keeping up with the volume of new drugs that can be ordered, understanding and fulfilling payer requirements, managing full schedules, and measuring outcomes and total costs associated with different treatment regimens. All of these challenges are compounded by physician burnout (much of which is attributed to administrative tasks) and an impending shortage of oncologists to manage the increased workload.2,3
The result is an overwhelming amount of information and a number of complicated decisions that oncologists must make each day that affect not only the health of their patients but also the health of their practice. While the industry has developed valuable resources for oncologists that consolidate information, such as the National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guidelines) and websites such as UpToDate for staying informed on the latest oncology research,4,5 reviewing these resources adequately requires time and may become an additional burden for many oncologists.
Clinical decision support (CDS) systems are integrated tools that present existing clinical knowledge and patient-specific information to clinicians at the point of care to improve health outcomes.6,7 CDS systems can be used by oncologists and administrators to help navigate information related to treatment decisions as well as reimbursement. However, widespread adoption of existing CDS tools in oncology has yet to be realized for various reasons, including disruptions to workflow, lack of trust in the evidence, or the perception that the information that they present may already be known.8 Recently, health care technology companies and organizations such as the NCCN are partnering with oncology practices in efforts to make integration of effective CDS systems into daily point-of-care workflow a reality in cancer care.
This perspective piece aims to further explore the challenges that oncologists and administrators face and to shed light on how CDS systems can help to guide evidence-based therapeutic decisions, minimize errors, improve patient outcomes, and address critical aspects of the economics, performance, and burden of care faced by clinicians and health systems.
Clinical Decision Support: Why Now?
As oncologists face an increasingly complex treatment landscape, CDS systems play a fundamental role in providing guidance and confirmation of appropriate decision making.7 In addition to helping oncologists navigate treatment decisions, the convergence of a number of industry trends have made implementing CDS systems a rising priority across the cancer care continuum.
Ever-Evolving Requirements of Value-Based Care Models
In the fee-for-service model, oncologists were paid based on services rendered. However, the rising costs of cancer care has placed substantial financial pressures on patients, oncology practices, and health care systems and is one of several factors that has contributed to the shift from volume-based to value-based care models.9-11 Accordingly, oncology practices are being asked to navigate complex payer requirements, possibly with multiple payers and unique arrangements. The lack of uniformity among payer requirements has heretofore been a major impediment to the use of CDS systems as well as a major day-to-day challenge for oncologists and patients. Value-based payment models require oncologists to meticulously document and report on quality measures, understand patient populations and treatment patterns, and advise patients on cost of care, demanding high analytical sophistication from oncologists and administrators and shifting the burden of cost-containment to the physician and practice.
The need for data and analytic insights on cost and quality of care has become increasingly important for oncologists’ success in value-based arrangements and practices’ need to satisfy payer requirements for data collection in a way that does not disrupt the core objective of providing care to patients. Many current CDS systems have limitations when it comes to delivering these insights to practices in a dynamic and actionable manner. To make matters more complicated, much of the documentation required for value-based arrangements is unstructured data (eg, clinician’s notes or scanned documents in a patient’s record) and difficult for software systems to translate into the required reporting format.
Data Challenges
Pathway-compliance data can give physicians and payers valuable information on evidence-based decision-making at the physician- and practice-level; however, it would be suboptimal and probably not informative to compare the small volume of data generated from single practices to large sets of aggregated data from extensive networks. Often their modest size limits the ability of individual oncology practices to assess site-specific performance and extract meaningful or actionable information. Larger oncology datasets, such as those generated by clinic networks, are better suited to support this type of data analysis, but the current landscape of CDS systems and pathway tools available to oncologists has not garnered enough adoption to sufficiently meet these data needs.
The Adoption Gap
While progress has been made to improve the design and implementation of oncology pathways,12,13 the widespread adoption of CDS systems in oncology practices has been hampered due to a number of converging challenges.8,14-18 For one, many CDS systems function outside of existing electronic health record (EHR) software, which is disruptive to practice workflows. For a CDS system to be fully embraced by clinicians, it has to be intuitive, always contain the latest clinical information, and be integrated into the provider’s normal workflow. Ideally, the CDS system could be embedded within or “sit on top” of the existing EHR, so that there is no disruption to practice workflow or patient care. Additionally, many existing CDS tools, such as Clear Value Plus (McKesson) and ClinicalPath (Elsevier), require physicians to work through formal lengthy decision-trees, even in cases where oncologists know a priori which treatment is appropriate for a given patient. The perception that full decision-support processes may be time-consuming and largely redundant have been noted as possible deterrents to widespread adoption.16,17
Optimizing CDS: Pairing Quality Content With Effective Technology
Overcoming these challenges requires the collaboration and collective expertise of health care organizations, technology companies, and oncology practices. Leading organizations such as the NCCN have developed trusted treatment guidelines that incorporate “Categories of Preference” that are constantly updated based on the latest research and industry advances into their recommendation framework.19 However, in order for this trusted, high-quality content to be accessible to clinicians at the point of care, health care technology companies need to collaborate with organizations such as the NCCN and practicing oncologists to develop user-friendly platforms that incorporate clinically accurate content.
Recently, a collaboration between the NCCN, Flatiron Health (a health care technology and services company), and physician groups including OneOncology, Lahey Hospital and Medical Center, New Mexico Cancer Center, and Hematology-Oncology Associates of Central New York has produced a new CDS tool called Flatiron Assist™. This NCCN guidelines-based decision-support tool aims to help oncologists select evidence-based treatment options at the point of care with minimal disruption to their existing workflow. This tool has been in use at Lahey Hospital and Medical Center since October 2019 and was deployed at New Mexico Cancer Center, West Cancer Center (a OneOncology practice), and Hematology-Oncology Associates of Central New York in May of 2020.
Clinical experts collaborated with design specialists to develop this tool under the guiding principle of user-centered design. One of the premises in this development was that, for a substantial proportion of patients, physicians are able to select appropriate therapies independently, therefore Flatiron Assist provides alternatives to the “mandatory decision-tree” approach. The tool is flexible to users’ preferences, offering calibrated guidance, or providing a more unobtrusive interface (Figure 1). While decision-support is always available, regimens can be selected in a 20 to 30 second single-step process without multi-step decision tree navigation. This may mitigate obstacles to the use of the CDS tool, easing engagement before ordering a treatment, as opposed to retrospectively. Whether the oncologist uses the full decision-support tree or not, the platform captures the data needed to determine guideline concordance.
Another user-centric feature potentially valuable within the context of a large network (eg, OneOncology) is the capability of Flatiron Assist to allow oncologists to design preferred regimens at the practice level and then monitor compliance with the regimens regardless of the way physicians decide on the selection. In the future, this feature will allow practices to assess outcomes by regimen and therefore have real-time feedback on the impacts of their choices.
The Benefits Enabled by Partnership
The benefits of this collaboration can help to alleviate some of the major challenges that oncologists face. CDS systems can serve as a mechanism for disseminating industry recognized and accepted, evidence-based NCCN Guidelines and the NCCN Categories of Preference, enabling oncologists to have full access to broad expert consensus at the point of care. Having access to current information about, for example, the latest in biomarker testing and approved targeted therapy options would enable oncologists to select the best regimen the first time and eliminate the need for retrospective review and potential changes in therapy.
Another major benefit of CDS systems that integrate NCCN content is the potential to streamline the prior authorization process, one of the greatest administrative hassles for oncologists and front office staff. Survey data from the American Medical Association show that on average practices process 33 prior authorizations per physician per week.20 Rather than questioning each individual proposed treatment plan for a patient, a CDS system could instead incorporate an “on pathway” button that submits notification of the pathway-compliant approved treatment plan directly to the insurer. Through the close collaboration with pilot practices and with entities like the NCCN, the Flatiron Assist tool was designed to incorporate this functionality. The tool enables oncologists to leverage previously documented clinical data or add pertinent information to determine pathways compliance. The data can be automatically submitted to payers, allowing the payers to respond in real-time with an authorization. Alternatively, data can be pre-populated in prior authorization portals. In either case, this solution could save significant time for prior authorization staff at the practice and for the insurer, reduce denial rates, and accelerate time to treatment for patients.
Furthermore, CDS tools can play a key role in surfacing information about clinical trials. Practitioners are challenged with not only staying on top of the latest and best standard treatment regimens but also being aware of investigational treatment options for which a patient may be eligible. Building this capability into the CDS workflow in a way that is tailored to each organization is highly valuable for both oncologists and patients.
The Power of Network-Wide Shared Data for Success in Value-Based Arrangements
Providing access to the latest guideline-concordant treatment options, improving the prior-authorization process, and surfacing clinical trials all bring substantial benefits to physicians, administrators, and patients. In addition, the benefit of collaboratively developing a tool like Flatiron Assist is observed at a larger scale. With a number of sites across the nation, it is essential that care networks such as OneOncology or National Cancer Care Alliance have the ability to maintain and publish centralized pathways content and standardize care across all sites. Furthermore, by enabling sites across the network to use the same tool in a standardized manner, practices can gather large datasets for analytics and better understanding of costs and quality of care. These tools enable oncology networks to both drive standardization across practices and clinics and gather real-world data from those sites of care for analysis and iteration.
The overall goal of value-based care models is to reimburse oncologists for quality and value of care rather than volume of care. A key feature of value-based arrangements is bundled payments, which require accurate data on the cost of care for a given patient. To facilitate the development of bundled payments, innovative models such as MASON (Making Accountable, Sustainable, Oncology Networks) are being developed to help providers understand the actual cost of oncology care.21 MASON uses data science to compare charge data with clinical information obtained from the National Cancer Care Alliance practices to accurately assign costs to each pathway and each patient. However, the clinical data required to perform such analyses cannot be locked in unstructured data (typically in the form of physician notes).
With models like MASON as a guide, Flatiron Assist bridges this gap by collecting the oncology data (eg, stage, histology, setting, biomarkers, line of therapy) needed to create specific patient cohorts that enable the calculation of more accurate and consistent cost estimates, appropriately adjusted for clinical characteristics. These “oncology payment categories” and associated pathways can eventually be used to produce bundled payments. Over time, the same platform can be used to refine pathways to ensure the most efficacious, cost-effective regimen is recommended to all oncologists within the network.
Looking Ahead
The aspiration of every oncologist is to have the wisdom and science of the entire oncology community at his or her fingertips to assist in the most appropriate treatment selection for the patient to achieve the best outcome; they simultaneously want to ensure fair payment for these choices and therefore the financial viability of the practice to provide ongoing optimal cancer care. As oncologists continue to work with health care technology providers to implement next-generation CDS systems, they will be armed with more structured clinical data in combination with cost data to continuously optimize and refine pathways. Furthermore, through these partnerships, oncologists can access national networks of data to answer specific questions where the sample size is too small for a given cancer practice site. Organizations like the NCCN will also be able to use this type of national real-world data to inform content development, especially where there are gaps in data from clinical trials.
At the individual level, physicians will be able to use pathway compliance data to validate their own clinical decisions, target personal learning to areas where pathway compliance is lower, and be reminded of new advances in science without disrupting their workflow. Together, physicians and payers will be able to evaluate the efficacy and toxicity of a regimen in real-world patients as well as understand the financial impact of their choices.
Ultimately, patients will benefit from a better understanding of their choices for care and participate in informed shared decision-making as they undergo treatment for cancer.
References
1. DeSantis CE, Miller KD, Dale W, et al. Cancer statistics for adults aged 85 years and older, 2019. CA Cancer J Clin. 2019;69(6):452-467. doi:10.3322/caac.21577
2. Yang W, Williams JH, Hogan PF, et al. Projected supply of and demand for oncologists and radiation oncologists through 2025: an aging, better-insured population will result in shortage. J Oncol Pract. 2014;10(1):39-45. doi:10.1200/JOP.2013.001319
3. Kane L. Medscape national physician burnout & suicide report 2020: the generational divide. Updated 2020. Accessed October 8, 2020. https://www.medscape.com/slideshow/2020-lifestyle-burnout-6012460#2
4. National Comprehensive Cancer Network. NCCN guidelines®. Updated 2020. Accessed October 8, 2020. https://www.nccn.org/professionals/physician_gls/default.aspx
5. UpToDate. Updated 2020. Accessed October 8, 2020. https://www.uptodate.com/home
6. Castaneda C, Nalley K, Mannion C, et al. Clinical decision support systems for improving diagnostic accuracy and achieving precision medicine. J Clin Bioinforma. 2015;5:4. doi:10.1186/s13336-015-0019-3
7. Osheroff JA, Teich JM, Middleton B, Steen EB, Wright A, Detmer DE. A roadmap for national action on clinical decision support. J Am Med Inform Assoc. 2007;14(2):141-145. doi:10.1197/jamia.M2334
8. Khairat S, Marc D, Crosby W, Al Sanousi A. Reasons for physicians not adopting clinical decision support systems: critical analysis. JMIR Med Inform. 2018;6(2):e24. doi:10.2196/medinform.8912
9. Vokinger KN, Hwang TJ, Grischott T, et al. Prices and clinical benefit of cancer drugs in the USA and Europe: a cost-benefit analysis. Lancet Oncol. 2020;21(5):664-670. doi:10.1016/S1470-2045(20)30139-X
10. Mariotto AB, Yabroff KR, Shao Y, Feuer EJ, Brown ML. Projections of the cost of cancer care in the United States: 2010-2020. J Natl Cancer Inst. 2011;103(2):117-128. doi:10.1093/jnci/djq495
11. Schnipper LE, Davidson NE, Wollins DS, et al. American Society of Clinical Oncology statement: a conceptual framework to assess the value of cancer treatment options. J Clin Oncol. 2015;33(23):2563-2577. doi:10.1200/JCO.2015.61.6706
12. Daly B, Zon RT, Page RD, et al. Oncology clinical pathways: charting the landscape of pathway providers. J Oncol Pract. 2018;14(3):e194-e200. doi:10.1200/JOP.17.00033
13. Zon RT, Frame JN, Neuss MN, et al. American society of clinical oncology policy statement on clinical pathways in oncology. J Oncol Pract. 2016;12(3):261-266. doi:10.1200/JOP.2015.009134
14. Pawloski PA, Brooks GA, Nielsen ME, Olson-Bullis BA. A systematic review of clinical decision support systems for clinical oncology practice. J Natl Compr Canc Netw. 2019;17(4):331-338. doi:10.6004/jnccn.2018.7104
15. Bucur A, van Leeuwen J, Christodoulou N, et al. Workflow-driven clinical decision support for personalized oncology. BMC Med Inform Decis Mak. 2016;16(suppl 2):87. doi:10.1186/s12911-016-0314-3
16. Kuntz G. Contemplating the future of oncology pathways. J Clin Pathways. 2019;5(8):28-30. doi:10.25270/jcp.2019.10.00099
17. Greene RA, Bates DW, Kawamoto K, Middleton B, Osheroff J, Shahar Y. Clinical decision support models and frameworks: Seeking to address research issues underlying implementation successes and failures. J Biomed Informat. 2018;78:134-143. doi:10.1016/j.jbi.2017.12.005
18. Bates DW, Kuperman GJ, Wang S, et al. Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality. J Am Med Inform Assoc. 2003;10(6):523-530. doi:10.1197/jamia.M1370
19. National Comprehensive Cancer Network. NCCN categories of preference. Updated 2020. Accessed October 7, 2020. https://www.nccn.org/professionals/physician_gls/categories_of_consensus.aspx#nccnpreference
20. American Medical Association. 2019 AMA prior authorization (PA) physician survey. Published 2020. Accessed October 7, 2020. https://www.ama-assn.org/system/files/2020-06/prior-authorization-survey-2019.pdf
21. National Cancer Care Alliance. MASON - making accountable, sustainable, oncology networks. Updated 2020. Accessed October 7, 2020. https://nccalliance.org/innovation/mason/
