Imagine this. A 35-year-old woman has just been diagnosed with breast cancer. Her doctor encourages her to consider germline genetic testing, explaining that it could help pick the best treatment for her – and may answer whether her family members are at a higher risk for cancer too. The woman talks with her friend, a woman about her age, who also has breast cancer and is seeing the same doctor. But her friend hasn’t heard anything about genetic testing. Similar age, similar diagnosis. What’s the difference?
The first woman is white; her friend is Black.
Studies show that minority patients with breast cancer are 50% to 65% less likely than white patients to get genetic testing, even if it is free or covered by insurance. Genetic testing is the first step in precision oncology – individualizing treatments based on a person’s hereditary risk or the molecular characteristics of their cancer. A doctor’s recommendation is a singular predictor of whether patients get genetic counseling and testing, surpassing even cost, culture, and patient attitudes. Today, we must reimagine how to help doctors make sound clinical decisions to achieve precision oncology’s north star: “the right treatments, at the right time, every time to the right person.”
There could be many reasons for why the doctor never spoke with the Black woman about genetic testing. Perhaps the doctor was running behind schedule that day, or maybe the patient’s diagnosis is more complicated. Possibly, the doctor relied on heuristics – or mental shortcuts – instead of taking the time to review national guidelines recommending that all women diagnosed with breast cancer under the age of 45 consider germline genetic testing, regardless of race or ethnicity.
Heuristics are known to accentuate cognitive biases and lead to suboptimal decisions. When a patient’s race is used as a heuristic, doctors may wrongly assume that genetic testing should be reserved only for white women of Ashkenazi Jewish descent (who have higher rates of breast cancer-associated BRCA1/2 gene mutations). Even if doctors don’t intend to, we propagate health inequities and structural racism when we make decisions that are shaped by these types of implicit biases.
As we recently wrote in JCO Precision Oncology, with insights from behavioral science, the electronic health record (EHR) could be used in three ways to “nudge” doctors to overcome race as a heuristic in decision-making.
First, defaults – pre-selected choices that are made unless actively changed – should be integrated into the EHR to automatically make genetics referrals or to order testing for any patient who meets criteria. This approach has already been shown to promote higher-value medication prescribing and imaging usage in cancer care. Defaults are powerful tools because they minimize the amount of time and cognitive effort that doctors need to put into making decisions, helping them overcome their reliance on heuristics.
Second, automated patient dashboards should be built in the EHR to identify which of a doctor’s many patients should get certain tests or treatments, minimizing “choice overload,” which can lead to no choice at all. Programs like this have already been built to successfully match patients to clinical trial opportunities, at large academic centers and also at community practices that serve minority and underserved patients. By relying on technology to enhance doctors’ decision-making, automated patient dashboards could be highly effective in promoting equitable precision oncology care.
Third, decision support should be set up in the EHR to ping doctors with testing and treatment recommendations when they’re seeing patients in real-time. Embedding guidance for doctors into the EHR has already been shown to enhance equity in other conditions like high blood pressure. This approach could also be helpful in the complex and rapidly moving field of precision oncology.
These three strategies could be rolled out separately, but when implemented together, they are likely to be even more impactful.
We know that technology can have its own biases, so we need to make sure behavioral science-informed EHR solutions don’t inadvertently widen existing health inequities as they get implemented. In addition, there are crucial challenges – access and affordability, health literacy, healthcare system distrust, and structural racism – that must be addressed.
Because a doctor’s recommendation matters so much in genetic testing and precision oncology, let’s start by using the EHR to make it easy for doctors to do the right thing every time.
Kelsey S. Lau-Min, MD, is a fellow in hematology/oncology at the University of Pennsylvania and an innovation fellow at the Penn Center for Cancer Care Innovation, both in Philadelphia. Carmen E. Guerra, MD, MCSE, is a physician and associate director of diversity and outreach at Penn’s Abramson Cancer Center and board scientific officer at the American Cancer Society. Katherine L. Nathanson, MD, is a cancer geneticist and deputy director at Penn’s Abramson Cancer Center. Justin E. Bekelman, MD, is a radiation oncologist, director of the Penn Center for Cancer Care Innovation, and a senior fellow at Penn’s Leonard Davis Institute for Health Economics.