Diversity Reporting Improved in Pulse Oximetry Testing After FDA Guidance

There was “low but improving uptake” of reporting about the diversity of participants in summary documents for FDA-cleared pulse oximeters after voluntary guidance was issued in 2013, an analysis of public FDA records found.

Reports of race, ethnicity, or skin tone occurred 320 times in 12.5% of 767 FDA summary documents that discussed pulse oximeter performance testing from 1996 to 2024, reported Kadija Ferryman, PhD, of the Berman Institute of Bioethics at Johns Hopkins University in Baltimore, and co-authors in a research letter published in JAMA.

Mentions of these characteristics were included in 4.1% (17 of 412) of documents before the 2013 guidance was issued, 13.8% (12 of 87) of documents around the time the guidance was released, and 25% (67 of 268) of documents following the guidance’s release (P

“The most important finding is that although there were more mentions of skin color descriptors in performance testing after the FDA’s guidance, a majority of the public clearance documents for pulse oximeters did not include any mention of testing in diverse individuals,” Ferryman told MedPage Today in an email.

“Clinicians who work in hospital settings often do not get to choose which pulse oximeter device they use with their patients,” Ferryman said. “Because this research is based on the public record, it suggests that even if clinicians wanted to do their own research on the performance of pulse oximeters across diverse populations, the majority of FDA-cleared device records do not include any information about testing in different skin tones.”

Pulse oximeter readings in patients with darker skin tones tend to overestimate oxygen saturation, a long-standing issue described in multiple studies and discussed by an FDA advisory committee.

“Clinical data have demonstrated that many hospital pulse oximeters routinely read falsely higher in individuals with darker skin color than in those with lighter skin color — that is, they display a racially and ethnically disparate statistical bias,” noted Carmel Shachar, JD, MPH, of Harvard Law School in Boston, and co-authors of an accompanying viewpoint article.

“Precluding a simple adjustment, pulse oximeter readings are also more variable in individuals with darker skin tones than in white individuals — that is, they are less precise. These errors combine to produce failures in detecting life-threatening low blood oxygen levels among individuals with darker skin tones,” they wrote.

“The ultimate goal should be to have pulse oximeters that work effectively for the full range of skin tones,” Shachar and colleagues added. “No single actor can at present solve this problem alone.”

For their analysis, Ferryman and her team researched the FDA’s database to find unique device documents based on certain product codes for oximeters starting in 1978. These summary documents are a standard part of the agency’s 510k pathway, and include key data used in the FDA’s review and decision to permit the marketing of a device.

The researchers collected summaries that used any of 17 search terms related to race and ethnicity or skin color, and then manually confirmed the terms were related to study participants. Of 1,024 clearances of oximeters identified, 767 summary documents with relevant information were retrievable from the FDA’s website.

To gauge the impact of the FDA’s guidance, they separated results into three time periods: pre-guidance from 1996-2012, a “washout period” around the time of the guidance from 2013 to 2015, and from 2016 to 2024.

The study’s greatest strength was its ability to analyze skin tone descriptors in performance testing of pulse oximeters cleared by the FDA from all publicly available documents, Ferryman observed.

Limitations included that only publicly available documents were used. “Pulse oximeter device makers do submit longer applications to the FDA that are not available in full publicly, so it is possible that those non-public documents contain some information about skin tone not included in the documents we examined,” Ferryman said.

The researchers also highlighted the “seeming confusion about concepts of skin tone, race and ethnicity, and national origin” in FDA documents, noting that more explicit direction about what to report may be helpful.

Newer FDA guidance on pulse oximeter testing that’s under development may correct some of these problems, but no single change in guidance “is likely to be sufficient to fully correct the problems of development, marketing, and dissemination of fully equitable pulse oximeters,” they wrote.