Forehead thermometers, which were widely used to screen for COVID-19 throughout the pandemic, may be less accurate in Black patients—an oversight that could be contributing to delayed diagnoses and a greater risk of mortality.
Researchers found that, compared with oral temperature measurement, temporal temperature measurement—or the use of forehead thermometers—was about 26% less likely to identify fever in Black patients. There were no significant differences in oral and temporal temperature readings in white patients.
“If fevers are going undetected, then notifications are being missed, and these missed notification pathways could lead to delays in antibiotics and could lead to worse outcomes for Black patients,” lead study author Sivasubramanium Bhavani, MD, an assistant professor in the Department of Medicine, Division of Pulmonology, Allergy, Critical Care and Sleep Medicine at Emory University School of Medicine, told Health.
For the study, researchers from Emory University and the University of Hawai’i evaluated the temperature readings of 4,375 people—2,031 Black patients and 2,344 white patients—who were admitted to four Emory hospitals between 2014 and 2021.
Each patient had their temperature taken orally and then temporally within a one-hour window on the first day of being hospitalized. The researchers then compared the measurements in Black and white patients and found that the temporal (forehead) temperature readings were lower than the oral temperature readings in Black patients.
Fever was detected in 10.1% of Black patients with a forehead thermometer and in 13.2% of Black patients with an oral thermometer. In white patients, 10.8% had a fever with the forehead thermometer and 10.2% had a fever with the oral thermometer.
Temperature measurement is an integral part of care delivery in hospital, outpatient, and community settings. Healthcare providers typically refer to a patient’s temperature early on in the diagnostic process to inform the type of care patients receive—from triggering sepsis alerts to determining triage for patients.
Because of that, discrepancies, even small ones, likely have serious consequences on the health outcomes of people of color, as a missed fever could cause delays in treatment or missed diagnoses.
“These small differences in temperature can make a huge difference in the decision on how to treat a patient, when to call for more help, order more tests, and ultimately how to treat their disease,” Eleni Linos, MD, MPH, DrPH, a professor of dermatology and epidemiology at Stanford University Medical Center, told Health.
In the community setting, where temporal thermometers were often used by churches, schools and workplaces to detect COVID, the tool’s inaccuracies likely led to a significant number of missed fevers, and therefore, missed or delayed diagnoses as well, Dr. Bhavani added, as well as potentially avoidable infections in others.
Prevalence of Racial Biases in Medical Tools and Diagnostic Procedures
Certain medical tools, like thermometers and pulse oximeters, work by sending light or heat through the skin, and though evidence has been mixed, some data suggests that skin color likely impacts how skin emits light, heat, and radiation—and ultimately the measurements these devices produce.
Though research was done during the pandemic regarding the effectiveness (or lack thereof) of pulse oximeters on darker skin tones—namely that the devices may be inaccurate, overestimate oxygen levels, or even make people less likely to qualify for life-saving treatments—these racial biases have been known about for years.
Evidence dating back to 1976 shows that scientists were aware pulse oximeters were influenced by skin pigmentation—even leading researchers to say that the method was “not capable of making absolute measurements.”
“While the intent of device manufacturers is not to harm patients, the impact of having misleading numbers for their pulse oximeter or thermometer results can be harmful for Black patients when it comes to health outcomes,” Dylan Roby, PhD, an associate professor of health, society, and behavior at University of California Irvine’s Program in Public Health, told Health.
Issues with how healthcare providers use the devices on patients are also possibly contributing to inaccurate test results—like not scanning the forehead sufficiently, study authors said.
And it’s not just pulse oximeters and forehead thermometers—skin cancer diagnostic tools and even a heart-rate monitoring technology called remote plethysmography have been shown to be less accurate in people of color.
It’s likely that we’ll learn of even more medical devices that have difficulty performing optimally based on skin color in the future. “Pulse oximeters and temporal thermometers are likely the tip of the iceberg,” said Dr. Bhavani, “and more studies in the future will uncover hidden biases in other medical devices and diagnostics.”
The researchers hope that healthcare providers will look to alternative medical tools to ensure all patients from all backgrounds can get the most effective care as quickly as possible.
“The short-term solution may be to ensure physicians, nurses, and medical technicians use oral thermometers on Black patients, or all patients, to ensure that readings are accurate,” said Roby.
But the core, long-term problem is how manufacturers develop medical devices—namely how they work on different bodies.
Researchers hope these findings pave the way for future studies to explore the potential racial biases in our medical system, not only with diagnostic medical tools and devices but with scoring systems (which track markers of illness), treatment algorithms, and even clinical trials—all of which are likely influenced in some way, shape or form by racial bias, sex and gender, or age; and thus have serious implications regarding safety and effectiveness medical devices and treatments.
“We need to make sure these tools are designed, calibrated and tested on people from diverse groups. This is absolutely essential,” said Dr. Linos. “These tools need to help everyone, not just a subset of the population.”
Thanks for your feedback!