Monday, Feb 28, 2022
Technology Must Be Intentionally Designed to Appreciate Our Differences
I have participated in several group training discussions focused on reducing racism and bias in healthcare. At some point in the training everyone is asked to turn to their neighbor and share how their childhood experiences and family’s viewpoint shaped their perspective on racism. After a few minutes of discussion participants are asked to share a little about their conversation. I know from my own experience that people are very reticent to dig deep to share their real story and want to present the best version of themselves to the group, so I haven’t found the discussion particularly insightful.
Never in those trainings has anyone said that they came from a racist family or that during childhood they wouldn’t sit next to another child who didn’t look like them. In fact, the comments most often heard are, “my family treated everyone the same” and “I don’t see color, I see people”. At first, I didn’t realize the dismissive nature of those comments or how they serve to create inequality.
The light bulb moment that changed the way I think occurred when a wise training facilitator advised that to not see color, and to treat everyone the same is not the goal, nor is it even the right thing to do. The right thing is to appreciate each other for our differences and meet each person where they are to ensure each person has what they need to thrive.
This is a start toward building equity in healthcare. When we don’t attend to important differences between us, and we don’t intentionally create structures which meet each person where they are, we perpetuate systems and policies which are built on a foundation of inequality.
Breakthroughs that bring hope can happen in the darkest of times. Our response to COVID has illuminated the pervasive inequities in healthcare in an unprecedented way, because it came on as an isolated event with enormous data analyzed in real time. This gave us pause to ask, ‘Why?’ Why are Black patients 5x more likely than other races to test positive for COVID? Why is the COVID mortality rate more than 2.5 times higher for Black people than white people? Why are low-income and minority children more than twice as likely than white children to test positive for COVID?
COVID-19 brought focused attention to differences in health outcomes between different populations of people. We must now look beyond the pandemic to the rest of healthcare to understand the harmful impact of implicit bias, racism and disparities brought on through the many social determinants of health. Healthcare must work to eliminate “systematic differences in health outcomes which arise from the social conditions in which people are born, grow, live, work and age” (Health inequities and their causes, 2018)
To eliminate inequalities in healthcare we must attend to those very things that make us different and ensure care considers those differences. Until healthcare has equity at the center, some people will be left out.
Our experience fighting COVID has demonstrated that safe, high-quality, patient-centered care is limited by inequities in how care is delivered. We are fortunate to have technologies focused on improving care delivery. But we must pay close attention to how we design, test, develop, and deploy these new technologies.
Take the pulse oximeter, one of healthcare’s most used medical devices. In the few times that I have had to go to the emergency room, such as when I had an anaphylactic allergic response to a dye used during my MRI or when I tore the ACL on my right knee, one of the first steps in my care was the nurse placing the pulse oximeter on my finger to gauge the oxygenation of my blood. Healthy blood oxygenation should fall between 95-100 mm Hg. Anytime blood oxygenation falls below about 95 mm Hg is cause for concern and if it falls below about 90 mm Hg it may be time to receive life-saving oxygen at the hospital.
Pulse oximeters use infrared and red light shined through your finger to measure the oxygenation of your blood. Well oxygenated blood is bright red, while blood depleted of oxygen is dark purple-red in color. Bright red blood absorbs infrared light and allows more red light to pass through than deoxygenated blood. This information coupled with your pulse determines the oxygenation of your blood.
Simple, right? But not so fast, because to “see” your blood the light must pass through your skin. But pulse oximeters are calibrated for light skin, not darker skin, and as a result they may overestimate saturation levels by several critical points. A 2020 study from University of Michigan Hospital found that oxygenation levels were significantly overestimated by the pulse oximeter when compared with the more accurate arterial oxygen saturation in 11.7% of Black and 3.6% of white patients. This device is an important tool for use at home to help COVID patients know when they have reached a level of illness that requires them to go to the hospital. It is also used by clinicians to help assess when it is time to be placed on supplemental oxygen. What happens when the information is not accurate? What decisions are made about patient care based on that information?
This isn’t the first time this problem has been identified. In fact, there are several technologies that depend on color sensing which can contribute to racial bias, including the health monitoring device you may be wearing on your wrist. When we see everyone as “the same” and we aren’t mindful in our development and deployment of medical technologies, we risk perpetuating racism in healthcare.
It is essential that we employ diverse teams of people with different backgrounds including race, gender, age, and culture to research, design and implement new technologies and data informed solutions and proactively ensure the development of unbiased algorithms in AI and ML which appreciate each other’s differences to improve care for all.
Silver linings can happen in the darkest of times. What we’ve learned from COVID has opened the door we use to hide structural inequities in healthcare. We need to not only open the door wider, but we must step through the door to fix what is devastating to our communities, our patients, and our workforce.
Health inequities and their causes. (2018). Retrieved from World Health Organization : https://www.who.int/news-room/facts-in-pictures/detail/health-inequities-and-their-causes#:~:text=Health%20inequities%20are%20systematic%20differences%20in%20health%20outcomes,people%20are%20born%2C%20grow%2C%20live%2C%20work%20and%20age.
Michael W. Sjoding, M. R. (2020). Racial Bias in Pulse Oximetry Measurement. New England Journal of Medicine, 383:2477-2478. Retrieved from https://www.nejm.org/doi/full/10.1056/NEJMc2029240
Oxygenation and Ventilation. (2021, December 16). Retrieved from NIH COVID-19 Treatment Guidelines: https://www.covid19treatmentguidelines.nih.gov/management/critical-care/oxygenation-and-ventilation/
Yash Mantri, J. V. (2021). Impact of skin tone on photoacoustic oximetry and tools to minimize bias. Biomedical Optics Express, 13(2), 875-887. Retrieved from https://opg.optica.org/boe/fulltext.cfm?uri=boe-13-2-875&id=468620