From Race-Based to Precision Oncology: Leveraging Behavioral Economics and the Electronic Health Record to Advance Health Equity in Cancer Care

The senseless killings of George Floyd, Ahmaud Arbery, Breonna Taylor, and countless others have led to a national reckoning on institutional racism. In medicine, race has long been used as a heuristic, or mental shortcut, used by clinicians to process information and make decisions within the time and cognitive constraints of our healthcare system.¹ However, using race as a heuristic in medical decision making has the potential to accentuate cognitive biases and lead to deviations from rational decision making, even among the most well-intentioned of clinicians.² Furthermore, it is well recognized that race is a social, not a biological, construct; as a result, the broad race categories used in medicine are heterogeneously defined and may be incongruent with genetic ancestry, particularly among minority patients.^3,4 Because cancer is inherently a disease of genetic aberration, using these race categories to make decisions in oncology may lead to the delivery of ineffective, or even detrimental, patient care.

The rapid advancement of precision oncology, the use of molecular and genetic testing to tailor cancer therapies to individual patients, has uniquely positioned the field of oncology to move beyond race-based medicine. Since the discovery of the first targeted therapy, imatinib, in the 1990s, over 80 targeted agents have been approved by the US Food and Drug Administration for the treatment of numerous solid organ and hematologic malignancies.^5,6 Some of these precision oncology applications correlate with patient race. For instance, epidermal growth factor receptor (EGFR) mutations in non–small-cell lung cancer are found in over 60% of patients of Asian descent and strongly predict sensitivity to EGFR tyrosine kinase inhibitors.^7,8 Similarly, several founder mutations in BRCA1/2 have been observed in multiple ethnic groups, including patients with Ashkenazi Jewish ancestry, and can be used to predict sensitivity to poly (ADP-ribose) polymerase inhibitors in patients with breast, ovarian, prostate, and pancreatic cancers.^9-14 Fortunately, the tools of precision oncology enable the administration of agents such as EGFR and poly (ADP-ribose) polymerase inhibitors to be guided by genetic testing rather than race or ancestry.

However, differential use of precision oncology has the potential to widen the very racial disparities that the field seeks to eliminate.¹⁵ Indeed, multiple studies have demonstrated decreased rates of genetic counseling, germline and somatic genetic testing, targeted therapy administration, genomic database participation, and clinical trial enrollment among minorities compared with White patients.^16-22 There are many barriers to achieving the promise of precision oncology, including lack of access to and payment for testing, targeted treatment, and counseling. Additionally, one major challenge is simply a lack of awareness among minority patients and clinicians.²³ The rapid advancement of precision oncology has made it difficult for many clinicians to have the time and expertise to select appropriate genetic testing, interpret and apply test results, and educate patients on their implications.²⁴ This challenge may be especially pertinent among minority-serving physicians, who have been shown to be less likely to refer to genetics and order germline testing for diseases such as breast and colorectal cancer.²⁵

The resultant use of race as a heuristic in the setting of these time and cognitive constraints may lead to decisions that are influenced by implicit bias, unconscious stereotypes, or attitudes toward people that influence one's actions and behaviors. Much like the general population, physicians have been found to have implicit bias with a preference for White over Black patients and to associate race with their assessments of patient intelligence, risk-taking behavior, and adherence to medical advice.^26-29 In turn, these perceptions have been linked to poorer communication, biased treatment recommendations, and unequal access to quality healthcare for minority patients.^30,31

In this paper, we propose three strategies to advance health equity in precision oncology (Table 1). These recommendations leverage the principles of behavioral economics and the electronic health record (EHR) to empower clinicians to overcome race as a heuristic in their medical decision making.

TABLE 1.

EHR-Based Strategies to Advance Health Equity in Precision Oncology

BEHAVIORAL ECONOMICS AND THE ELECTRONIC HEALTH RECORD

The field of behavioral economics recognizes that humans make decisions within the constraints of bounded rationality because of time and cognitive limits such as those observed with precision oncology.¹ This results in the use of heuristics (such as race), which may in turn increase individuals' reliance on cognitive biases (including implicit bias).² To minimize the use of these heuristics and cognitive biases, choice architecture, or the context in which in which options are presented, can be redesigned to nudge individuals toward better decisions.³² Choice architecture tools include (1) using defaults to minimize the effort used to make decisions, (2) selecting the number of alternatives that balances adequate choice with the cognitive burden required to evaluate each one, and (3) using decision support to steer individuals toward more desirable selections.³³ In medicine, behavioral economic principles have been proposed^34,35 and shown^36,37 to be effective strategies to incentivize physicians to deliver high-value cancer care. We propose that behavioral economic principles should be leveraged to promote health equity in precision oncology as well.

The EHR offers a scalable, cost-effective approach to harnessing behavioral economic principles to advance health equity in precision oncology. Since the Patient Protection and Affordable Care Act mandated the use of EHRs for all patient care in 2014, the EHR has emerged as a powerful tool to decrease disparities in care.^38,39 Furthermore, the integration of precision medicine into the EHR has gained traction in recent years and is an active area of investigation at many healthcare institutions, including ours.^40-42 However, EHR-based interventions have the potential to disrupt clinical workflows negatively and increase burnout among clinicians.^43,44 As such, they must be informed by the principles of behavioral economics to maximize their impact on medical decision making while minimizing clinician burden.

THREE STRATEGIES TO ADVANCE HEALTH EQUITY IN PRECISION ONCOLOGY

Default Orders

Status quo bias is the tendency to favor the current state of affairs, thereby impeding the ability to make decisions that may make one better off. Defaults—preselected choices that are applied unless an individual actively changes them—are powerful tools that take advantage of the status quo bias to nudge decision making while minimizing cognitive effort and preserving choice. The EHR is ideally suited to integrate defaults into clinical care. This approach has been shown to optimize cancer care delivery by increasing high-value medication prescribing and reducing unnecessary daily imaging during palliative radiotherapy.^36,37 In the precision oncology space, default orders in the EHR can be used to make direct referrals to genetics for patients who meet guideline criteria on the basis of factors such as tumor type and age, clinical characteristics that can be easily identified in the EHR. Default options can also be used to initiate reflex biomarker testing after a new cancer diagnosis, an approach that has been shown to increase testing uptake and expedite treatment administration.^45,46 Integrating default genetic referral and biomarker testing orders into the EHR may help clinicians overcome their reliance on race as a heuristic in their decision making while offering the flexibility to actively modify orders in accordance with patient preferences.

Automated Patient Dashboards

Choice overload occurs when there are too many options to decide between, resulting in excess cognitive burden, decision regret, and decreased motivation to make the decision altogether.^47-49 The EHR is well equipped to solve the problem of choice overload through automated patient dashboards, which can efficiently search through all the patients cared for by a given clinician or at a particular practice to identify a more limited pool of individuals who are eligible for a given intervention. This approach has been leveraged by multiple clinical trial matching programs to match patient and tumor characteristics from the EHR to study eligibility criteria, thereby enabling oncologists to identify study opportunities for individual patients. These programs have been effective in facilitating increased clinical trial participation not only within academic institutions but also at community oncology practices, sites that are more likely to serve minority and underserved populations.^50,51 Similar automated patient dashboards should be leveraged to identify candidate patients for other precision oncology opportunities that remain underutilized by minorities, such as genetics referrals, germline and somatic testing, targeted therapy administration, and genomic database participation. Because this strategy matches patients to precision oncology care using the automation and computing power of the EHR rather than the heuristics of human decision making, it is likely to be more effective than individual clinicians in advancing health equity.

Clinical Decision Support Systems

Decision support also mitigates choice overload by guiding individuals toward more desirable decisions while preserving their freedom of choice. In precision oncology, clinical pathways have emerged as a decision support strategy for clinicians to handle the field's ever-expanding scope and complexity.⁵² However, these pathways are ineffective unless they are seamlessly integrated into routine clinical practice to maximize their usefulness. Embedding clinical decision support systems into the EHR has been shown to facilitate decision making in other areas of cancer diagnosis, treatment, and supportive care,⁵³ and it has been associated with decreased racial disparities in the management of nononcologic conditions such as hypertension.³⁹ This approach should also be leveraged to standardize precision oncology care for all patients, regardless of race.

Perhaps an even more compelling clinician decision aid is a patient's own request for a given medical intervention. Not only does this approach uphold the clinician's respect for patient autonomy and preference, but it may also prime subsequent decision making. Priming, the use of subtle stimuli to activate thoughts and ideas,⁵⁴ is a powerful tool that can be directly used by patients to support clinicians in overcoming the use of race as a heuristic. In the EHR, the use of patient portals has been shown to be similar across racial and ethnic groups although efforts must be made to ensure equal adoption at the outset.⁵⁵ These electronic portals can be harnessed to automatically send educational materials to patients about their eligibility for and benefits of precision oncology and to communicate normative feedback on its uptake among other patients like them. These types of patient communications informed by the principles of behavioral economics may empower and nudge patients to initiate conversations with their providers about genetic testing, targeted therapy, or participation in precision oncology research.

LIMITATIONS AND ALTERNATIVE APPROACHES

Several concerns may be raised about the use of behavioral economics and the EHR to advance health equity in precision oncology. First, we recognize that our proposed strategies do not fully combat the lack of access, financial challenges, health literacy or language barriers, and longstanding history of mistrust in the healthcare system that have also impeded the adoption of precision oncology among minority patients. To be successful, EHR-based strategies informed by the principles of behavioral economics must be implemented as part of a broader, multifaceted approach that addresses all these challenges. Second, the application of behavioral economics in medicine is still in its nascency, so there is a paucity of data on the impact of choice architecture redesign on subgroups within the population. Future studies should incorporate the emerging concept of precision nudging to tailor behavioral economics interventions to specific patient populations as a strategy to ensure equal effectiveness for all. Finally, the implementation of EHR-based strategies may require initial investment of financial and human resources that is not feasible at all institutions. Early adopters should share their algorithms and code to facilitate the scaling up of these interventions throughout the rest of the medical community. As implementation efforts get underway, we must monitor for and address any signs of differential uptake so that our proposed strategies do not inadvertently widen existing inequities in care.

In conclusion, in this article, we present three EHR-based strategies that leverage the principles of behavioral economics to nudge oncologists away from the use of race as a heuristic in their decision making. The thoughtful redesign of choice architecture within the EHR will empower clinicians to better individualize cancer care, advance health equity, and progress from the era of race-based to precision oncology.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/po/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Carmen E. Guerra

Leadership: Freenome, Guardant Health, Genentech

Stock and Other Ownership Interests: Crispr Therapeutics, BEAM Therapeutics, Intellia Therapeutics, Editas Medicine

Honoraria: Lundbeck (I)

Consulting or Advisory Role: Bristol-Myers Squibb (I)

Speakers' Bureau: Janssen, Pfizer (I)

Research Funding: Bristol-Myers Squibb (Inst)

Travel, Accommodations, Expenses: Lundbeck, Janssen (I)

Uncompensated Relationships: Tapestry Networks

Justin E. Bekelman

Honoraria: UnitedHealthcare, National Comprehensive Cancer Network, CVS Health, Optum, American Journal of Managed Care

Research Funding: Pfizer, United Health Group, North Carolina Blue Cross Blue Shield, Embedded Healthcare

No other potential conflicts of interest were reported.