Rama A. Salhi MD, MHS, MSc , Valeria S.M. Valbuena MD , Ashraf Fawzy MD, MPH , Sarah M. Seelye PhD , Theodore J. Iwashyna MD, PhD
{"title":"Impact of Racial Differences in Hypoxia Distribution on the Measured Prevalence of Occult Hypoxemia by Pulse Oximeters","authors":"Rama A. Salhi MD, MHS, MSc , Valeria S.M. Valbuena MD , Ashraf Fawzy MD, MPH , Sarah M. Seelye PhD , Theodore J. Iwashyna MD, PhD","doi":"10.1016/j.chstcc.2023.100011","DOIUrl":null,"url":null,"abstract":"<div><h3>Background</h3><p>Racial differences in pulse oximeter accuracy increasingly have become recognized. However, previously published literature has not examined the extent to which underlying racial differences in levels of hypoxemia, such as those arising from differential testing, disease recognition, and treatment, may confound previously observed differences in pulse oximetry measurement.</p></div><div><h3>Research Question</h3><p>The presented study examined the extent to which underlying differences in arterial oxygen saturation (Sa<span>o</span><sub>2</sub>) drive previously observed racial differences in pulse oximetry occult hypoxemia measurement.</p></div><div><h3>Study Design and Methods</h3><p>Analysis was completed as a secondary data analysis of two existing databases. Data were reanalyzed from the previously published Veterans Affairs (VA) Patient Database (2013-2019) and the Extracorporeal Life Support Organization (ELSO) registry (2019-2020). Patients included general ward and critically ill patients. We compared the measured burden of occult hypoxemia (ie, Sa<span>o</span><sub>2</sub> < 88%, with peripheral capillary oxygen saturation ≥ 92%) when standardizing for population-level distributions of Sa<span>o</span><sub>2</sub> vs when standardizing the sensitivity at each Sa<span>o</span><sub>2</sub>.</p></div><div><h3>Results</h3><p>Black patients showed a higher likelihood of occult hypoxemia when compared with White patients in both data sources (Veterans Affairs Patient Database, 18.8% vs 14.9%; Extracorporeal Life Support Organization registry, 14.6% vs 7.0%). The distribution of Sa<span>o</span><sub>2</sub>, to an extent, does change the measured occult hypoxemia rates; however, large racial differences were persistent after standardizing based on underlying Sa<span>o</span><sub>2</sub> distributions.</p></div><div><h3>Interpretation</h3><p>Underlying differences in Sa<span>o</span><sub>2</sub> distributions were observed in the analyzed data. Such differences point to ongoing differentials in care; however, even when accounting for Sa<span>o</span><sub>2</sub> distributions, differential detection of hypoxemia by race persisted in pulse oximeters in contemporary use.</p></div>","PeriodicalId":93934,"journal":{"name":"CHEST critical care","volume":"1 2","pages":"Article 100011"},"PeriodicalIF":0.0000,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"CHEST critical care","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2949788423000114","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Background
Racial differences in pulse oximeter accuracy increasingly have become recognized. However, previously published literature has not examined the extent to which underlying racial differences in levels of hypoxemia, such as those arising from differential testing, disease recognition, and treatment, may confound previously observed differences in pulse oximetry measurement.
Research Question
The presented study examined the extent to which underlying differences in arterial oxygen saturation (Sao2) drive previously observed racial differences in pulse oximetry occult hypoxemia measurement.
Study Design and Methods
Analysis was completed as a secondary data analysis of two existing databases. Data were reanalyzed from the previously published Veterans Affairs (VA) Patient Database (2013-2019) and the Extracorporeal Life Support Organization (ELSO) registry (2019-2020). Patients included general ward and critically ill patients. We compared the measured burden of occult hypoxemia (ie, Sao2 < 88%, with peripheral capillary oxygen saturation ≥ 92%) when standardizing for population-level distributions of Sao2 vs when standardizing the sensitivity at each Sao2.
Results
Black patients showed a higher likelihood of occult hypoxemia when compared with White patients in both data sources (Veterans Affairs Patient Database, 18.8% vs 14.9%; Extracorporeal Life Support Organization registry, 14.6% vs 7.0%). The distribution of Sao2, to an extent, does change the measured occult hypoxemia rates; however, large racial differences were persistent after standardizing based on underlying Sao2 distributions.
Interpretation
Underlying differences in Sao2 distributions were observed in the analyzed data. Such differences point to ongoing differentials in care; however, even when accounting for Sao2 distributions, differential detection of hypoxemia by race persisted in pulse oximeters in contemporary use.