Pulse oximetry is a non-invasive test that measures the oxygen saturation level in a patient’s blood, and it has become an important tool for monitoring many patients, including those with Covid-19. But new research links erroneous readings from pulse oximeters to racial disparities in health outcomes, potentially leading to higher death rates and complications such as organ dysfunction in patients with darker skin.
It is well known that non-white intensive care patients (ICU) receive less accurate readings of their oxygen levels using pulse oximeters – the common devices that clip onto patients’ fingers. Now, an article co-authored by MIT scientists reveals that inaccurate pulse oximeter readings could lead to critically ill color patients receiving less supplemental oxygen during ICU stays.
The newspaper, “Assessment of Racial and Ethnic Differences in Oxygen Supplementation in Intensive Care Patients,” published in JAMA Internal Medicinefocused on whether there were differences in supplemental oxygen administration between patients of different races and ethnicities that were associated with pulse oximeter performance differences.
The findings showed that inaccurate measurements of Asian, Black and Hispanic patients resulted in them receiving less supplemental oxygen than Caucasian patients. These results provide insight into how health technologies such as the pulse oximeter contribute to racial and ethnic inequalities in health care, the researchers said.
The study’s senior author, Leo Anthony Celi, clinical research director and principal investigator at the MIT Laboratory for Computational Physiology, and principal investigator at the MIT Institute for Medical Engineering and Science (IMES), says the challenge is that technology in healthcare is routine. designed around the majority of the population.
“Medical devices are typically developed in wealthy countries with white, fit individuals as test subjects,” he explains. “Drugs are evaluated through clinical trials that disproportionately enroll white individuals. Genomics data is largely from individuals of European descent.”
“It is therefore not surprising that we observe differences in outcomes across demographics, with worse outcomes in those who were not involved in health care design,” Celi added.
Although pulse oximeters are widely used for their ease of use, the most accurate way to measure blood oxygen saturation (SaO2) levels is by taking a sample of the patient’s arterial blood. False readings from normal pulse oximetry (SpO2) can lead to hidden hypoxemia. Elevated bilirubin in the bloodstream and the use of certain medications in the ICU called vasopressors can also interfere with pulse oximetry readings.
More than 3,000 participants were included in the study, including 2,667 white, 207 black, 112 Hispanic and 83 Asian – using data from the Medical Information Mart for Intensive Care Unit version 4 or MIMIC-IV dataset. This dataset consists of more than 50,000 patients admitted to the ICU at Beth Israel Deaconess Medical Center, and includes both pulse oximeter readings and oxygen saturation levels detected in blood samples. MIMIC-IV also includes supplemental oxygen delivery rates.
When the researchers SpO . compared2 levels taken by pulse oximeter to oxygen saturation of blood samples, they found that black, Hispanic and Asian patients had higher SpO2 than white patients for a given blood oxygen saturation level measured in blood samples. The turnaround time for arterial blood gas analysis can take from a few minutes to an hour. As a result, clinicians typically make decisions based on pulse oximetry measurements, without being aware of sub-optimal performance in certain patient demographics.
Eric Gottlieb, the study’s lead author, a nephrologist, a lecturer at MIT, and a Harvard Medical School fellow at Brigham and Women’s Hospital, called for more research to better understand “how disparities in pulse oximeter performance lead.” to worse outcomes; potential differences in ventilation management, fluid resuscitation, triage decisions, and other aspects of care need to be explored, then we need to redesign and properly evaluate these devices to ensure they perform equally well for all patients.”
Celi emphasizes that understanding biases that exist in real-world data is critical to better developing algorithms and artificial intelligence to help clinicians make decisions. “Before investing more money in developing artificial intelligence for healthcare using electronic health records, we need to identify all causes of inequalities in outcomes, including those resulting from the use of sub-optimally designed technology,” he says. “Otherwise, we risk perpetuating and amplifying health inequalities with AI.”
Celi described the project and research as a testament to the value of data sharing, which is at the heart of the MIMIC project. “No team has the expertise and perspective to understand all the biases in real-world data to prevent AI from perpetuating health inequalities,” he says. “The database we analyzed for this project has more than 30,000 recognized users, comprising teams of data scientists, clinicians and social scientists.”
The many researchers working on this topic together form a community that shares and performs quality checks on code and queries, promotes reproducibility of the results, and crowdsources the collection of the data, Celi says. “There is harm if health data is not shared,” he says. “Restricting data access means limiting the perspectives from which data is analyzed and interpreted. We’ve seen countless examples of wrong model specifications and flawed assumptions that have led to models that ultimately harm patients.”