Latest News

Emerging Tech Aims to Remove Racial Bias from Blood Measurements

By Paul Nicolaus 

January 17, 2023 | A team of researchers is pursuing new technology to address racial disparities in blood measurements. The group ultimately hopes to develop a portable wearable device, like a monitor or watch, capable of accurately and non-invasively reading blood regardless of skin tone.  

University of Texas at Arlington bioengineering professor George Alexandrakis and Vinoop Daggubati, CEO of Austin-based Shani Biotechnologies, led a clinical study using an investigational device invented by Sanjay Gokhale, a medical scientist. The research measured 16 healthy participants’ hemoglobin and oxygen saturation levels using their device, which quantifies skin melanin and factors this into its readings.  

“We have developed an algorithm to correct for melanin, which is not there in any other pulse oximeter available in the market,” Gokhale told Diagnostics World. “That’s why our device works better in all people of color.”  

The study compared readouts from the investigational device to results gathered using a commercially available pulse-oximeter and the gold standard invasive measurement method. Their findings, published in the British Medical Journal Innovations (doi: 10.1136/bmjinnov-2022-001018), indicated that the results gathered from their newly developed device “were comparable with the invasive, point of care (POC) method.”  

The authors concluded that their technology and device “could potentially overcome current limitations of the pulse-oximetry devices.” The group sees potential for use in various settings, including outpatient clinics, emergency rooms, intensive care units, and homes. They also highlighted its potential use in the developing world, where severe anemia can be a significant cause of hospital admissions and death.  

To date, the device has been tested in preclinical and clinical studies. However, because of the relatively small number of study participants, more extensive studies are needed for development and validation.  

“Our group has addressed the issues around shorter wavelength, scattering of light and the impact of skin melanin,” Daggubati said in a news release. He added that the scientific community ought to keep an open mind about the use of green light to help address this issue of blood measurement racial disparities. 

FDA Advisory Committee Discusses Pulse Oximetry Concerns 

Oxygen saturation in a person’s blood is commonly measured using co-oximetry or pulse-oximetry methods, but some say these currently available options have notable drawbacks. One option is invasive and requires a blood sample, whereas non-invasive methods are potentially hindered by inaccuracies when used with people of color.  

Emerging data has revealed that pulse oximetry tends to overestimate blood oxygen saturation in people with darker skin because existing pulse oximeters do not correct for skin melanin. “The infrared light-based technology that underlies pulse oximetry has an inherent design flaw that disadvantages melanated individuals,” explained Joseph L. Wright, vice president and chief health equity officer at the University of Maryland Medical System.  

“The wavelengths of light that are transmitted through the skin to detect oxygenated hemoglobin in the blood are disrupted, or absorbed, by melanin,” he told Diagnostics World, and can deliver higher pulse oximetry readings than the actual levels of oxygen that are circulating in the blood. “The clinical result is occult hypoxemia in which patients with potentially dangerously low levels of oxygenation may go undetected when assessed based on a pulse oximeter reading alone.”  

FDA issued a public notice about the possible inaccuracies in pulse oximetry measurements and organized an advisory committee meeting held in November to discuss ongoing concerns. Presentations were delivered by various stakeholders, including FDA, clinicians, researchers, professional societies, patients, and industry.  

“The Panel agreed that the available clinical evidence from real-world studies demonstrated a reduction in the accuracy of pulse oximeter devices in individuals with darker skin pigmentation, which may lead to increased health risks in this population,” noted a summary of the Anesthesiology and Respiratory Therapy Devices Panel Meeting.  

Pulse oximeters are often placed on the fingertip and use light beams to estimate the blood oxygen saturation and pulse without drawing a blood sample. Prescription oximeters undergo FDA review, receive 510(k) clearance, and are only available with a prescription whereas over-the-counter oximeters are sold directly to consumers online or in stores. The COVID-19 pandemic has led to increased use of the latter.  

While pulse oximetry can be helpful, existing devices “have limitations and a risk of inaccuracy under certain circumstances,” explains an FDA safety communication. In addition to skin pigmentation, a variety of other factors can impact the accuracy of a reading, including skin thickness or temperature, poor circulation, fingernail polish, or tobacco use. “In many cases, the level of inaccuracy may be small and not clinically meaningful; however, there is a risk that an inaccurate measurement may result in unrecognized low oxygen saturation levels.”  

The safety communication points out that the most up-to-date scientific evidence reveals some accuracy differences in pulse oximeters between light and dark skin pigmentation. The difference is usually slight at saturations higher than 80% but more significant when those saturations are under 80%.  

It also highlighted a paper published in The New England Journal of Medicine (DOI: 10.1056/NEJMc2029240) that showed black patients had almost three times the frequency of low blood oxygen levels that were not picked up by pulse oximetry compared to white patients. Although this study had limitations, FDA indicated that its results “highlight a need to further evaluate and understand the association between skin pigmentation and oximeter accuracy.” 

Kelly Johnson-Arbor, an ER doctor and medical toxicology physician at the National Capital Poison Center, pointed out that this subject is timely. “This is an important topic because we are currently still within the COVID-19 pandemic and are also seeing many patients who have respiratory infections such as RSV,” she told Diagnostics World. “In both COVID-19 and RSV patients, as well as in patients with other conditions such as asthma, medical decisions and treatment recommendations are often based on a patient’s oxygen saturation, among other factors.”  

“If pulse oximeters provide inaccurate readings, physicians may order the wrong treatment or not recognize the true severity of a patient’s respiratory illness,” Johnson-Arbor added. “This could lead to patients being discharged from hospitals or ERs when they actually need to be admitted or could lead to patients being inadequately treated for their conditions.”  

Green Light Technology Could Lead to a More Equitable Device  

Existing pulse oximeters rely on red-infrared light and are based on technology developed over half a century ago, Daggubati of Shani Biotechnologies explained. “Unfortunately, there’s a lot of limitations to these current existing technologies, and especially regarding people with higher melanin skin content, such as African Americans,” he told Diagnostics World.  

For nearly a decade, he and Gokhale have been developing a technology that uses blue-green light instead. The device is placed on the back of a person’s wrist and reflected light is measured as an electrical signal. It can measure pulse oximetry and hemoglobin values in people with higher melanin skin content. He said this can be especially important regarding certain health issues, such as anemia or sickle cell disease.  

Existing pulse oximeters measure oxygen saturation—a relative measure of the percentage of oxygenation, added Alexandrakis, but they do not measure the amount of oxyhemoglobin per 100 mL of blood. He said this absolute measurement is more difficult to gather, and that is where their device is aimed at filling a gap. “The quantitative concentration of oxyhemoglobin is a measurement that we’re after that is not so easy to measure with existing technology.”  

There are a number of reasons why a person may be interested in this sort of information. “It could be that you have a circulation problem, or for an athlete trying to optimize their performance, or for a person doing dialysis and seeing fluctuations of the amount of hemoglobin in their blood,” he said. “Rather than having constant blood sampling, they could do this on the fly with a non-invasive instrument that they’re wearing.” 

The research team’s device works by sampling blood vessels on the underside of the palm, and the idea is that it can better control the depth of the tissue assessed because of the wavelengths it uses.  

Existing pulse-oximeters using red and infrared technologies can penetrate deeper depending on the wavelengths and the power of the light source. “The difference with this sensor is that it uses selected wavelengths based on the known absorbance spectrums of oxyhemoglobin, deoxyhemoglobin, and skin melanin to differentiate between them so that you can quantify this particular parameter—the oxyhemoglobin concentration in blood—very superficially,” Alexandrakis said.  

In other words, the device uses capillaries right underneath the skin while avoiding excessive contamination from larger, deeper vessels. “By selecting those wavelengths, you can factor out effective skin melanin and therefore be more accurate compared to the measurements that are made with standard near-infrared wavelengths,” he added. 

Next Steps, Notable Challenges, and Big Picture Thoughts 

Daggubati indicated that their group has now successfully conducted phase 1 and 2 trials, and the next step is to move on to phase 3 to conduct more extensive studies on patients. “Our UT Arlington study was all healthy volunteers,” Daggubati explained, “so we’re going to now move to the more clinical and sick patients.” Following the phase 3 trial, the ultimate goal is to shrink this technology into a wearable device that could be used in clinical settings or incorporated into smartwatches.  

Alexandrakis pointed out that as they work on moving this technology forward, they are up against the same challenges facing everyone else in the industry, including existing commercial vendors that already have FDA-approved devices. “Nothing works perfectly,” he pointed out. And melanin isn’t the only issue; there are other underlying anatomy and physiology issues to consider.  

He explained that everyone has different optical properties, meaning one person’s skin thickness may be slightly different than the next person’s. Depending on body weight and general body composition, someone may have more or less adipose tissue below their skin. In addition, some people may have underlying conditions that can affect blood flow, so there are many variables.  

“The challenge for any one of these techniques is to make something that works under all possible conditions,” Alexandrakis said, and many more studies are needed to address that challenge. Even if his team can conduct two or three studies in two or three different patient populations, for instance, he believes progress will eventually come in the form of compiling a large database of measurements under very different physiological conditions and then applying machine learning technologies.  

One key question is whether that data will be shared or kept private. Beyond creating the technology itself, the financial dynamics in this space largely depend upon “who owns the intellectual property of those big databases of personal information that people offer to the vendor, whether they realize it or not,” he added. “I think that's where the real power is, in both making these devices good, as well as helping people.” 

Paul Nicolaus is a freelance writer specializing in science, nature, and health. Learn more at

Load more comments