Can You Explain 'Relative Risk' in COVID-19?

Shantanu Nundy, MD;  Jay Bhatt, DO, MPH, MPA; and John Whyte, MD, MPH

May 21, 2021

During the pandemic, doctors and patients have confronted the reality that few decisions in medicine are cut and dried. Instead, we've been forced to consider the concept of relative risk.

Examples include: Does a negative COVID-19 test guarantee that I don't have the virus? After isolating for 10-14 days, can I still transmit the virus to other people? If I've had COVID-19 before, can I get it again? If I'm vaccinated, am I safe to fly?

COVID-19 clearly demonstrates that our toolkit is often too generic for individual decision-making. Interpreting a test result for COVID-19 depends on the type and precision of the test, the presence of symptoms, and the patient's exposure history. Guidance on isolating at home depends on housing stability, the number of household members, and the patient's job type. Even routine clinical decisions, such as whether to go to a clinic or emergency room, have been more complex because of the risk for nosocomial transmission.

Risk assessment is not new to the practice of medicine. Perhaps the best-known example is the Framingham risk score, which incorporates age, gender, smoking status, blood pressure, and cholesterol to help understand an individual's risk for coronary artery disease and to make treatment decisions. But COVID-19 revealed that our current approach to risk assessment is still crude and often does not take into account enough of a patient's individual context.

In response to these issues, we propose a set of recommendations for how the practice of risk assessment needs to evolve to mitigate the ongoing threat of COVID-19 and improve the delivery of routine care.

Expand the Scope of Risk Assessment

Risk assessment tools today are limited to narrow domains of medicine, such as the risk for coronary artery disease or osteoporosis fracture.

We need more evidence-based tools across a much broader set of conditions and situations, including those that occur outside of a care setting, such as tools to help individuals decide whether and where to seek care.

Where risk assessment tools do exist, they need become more individualized. Many screening guidelines are based on age and gender alone.

For example, the American College of Obstetricians and Gynecologists guidelines recommend screening mammograms for all women at age 40 and no later than age 50. However, we know that African American women are at higher risk for more aggressive forms of breast cancer at younger ages.

Risk scores should incorporate race and ethnicity, social determinants of health, and a wider range of biometric and genetic markers that would allow physicians to individualize care decisions. And because coronary artery risk also varies by race, the Framingham risk score should include an input for race and ethnicity.

Other risk assessment tools must take social context into account. During the pandemic we've learned that job type, housing situation, and family unit critically affect individual decision-making.

Implement Risk Assessment Into Clinical Practice

Even well-known tools like the Framingham risk calculator have poor uptake and understanding in clinical practice. Part of the challenge is that decision support lives outside the electronic health record and therefore outside the physician's clinical workflow.

Physicians also need more training. Research shows many physicians do not correctly interpret the terms "sensitivity" and "specificity."

We also need more real-world evidence about the utility of these tools in clinical practice and how to disseminate best practices for implementation.

Integrating Risk Assessment Into Public Health

Public health communication is meant to affect the largest number of people possible. But during COVID-19, we saw an erosion of trust that stemmed from guidance that people did not understand or did not know how to apply.

Greater investment will be required to build tools to help individuals understand their own risk. The US Preventive Services Task Force (USPSTF) has desktop and mobile tools for clinicians, but not for patients.

A major advantage of digital tools is their interactivity, which allows patients to enter their specific health and social risk information and get tailored education and guidance. An end-user perspective will be needed to design tools that are usable and useful to patients. Among other factors, these tools must take into consideration language needs and literacy levels.


Dr Francis Peabody famously wrote, "the secret of the care of the patient is in caring for the patient." As we look to move beyond this tragedy of COVID-19 while heeding the lessons from it, we need to more deeply understand the overall health and well-being of our individual patients, apply the best evidence to translate those individual needs and risks into guidance, and then help our patients understand what those risks and guidance mean for them.

COVID-19 is an opportunity for us to move toward a more personalized approach to healthcare that all patients need and deserve.

Shantanu Nundy, MD, is chief medical officer for Accolade in Plymouth Meeting, Pennsylvania; a lecturer on Health Policy at the George Washington Milken Institute for Public Health, and an advisor for World Bank Group in Washington, DC.

Jay Bhatt, DO, MPH, MPA, is on the faculty at the University of Illinois Chicago School of Public Health and a staff internist at Family Christian Health Center in Harvey, Illinois.

John Whyte, MD, MPH, is the chief medical officer for WebMD in Bethesda, Maryland.

For more news, follow Medscape on Facebook, Twitter, Instagram, YouTube, and LinkedIn


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: