Self-reported Race/Ethnicity and Occult Hypoxemia: Comment

Abstract

I am writing in regard to the recently published article by Burnett et al., “Self-reported Race/Ethnicity and Intraoperative Occult Hypoxemia: A Retrospective Cohort Study.”1 In this retrospective cohort study, the authors investigate the association between self-reported race/ethnicity and occult hypoxemia and hypothesize a greater prevalence of occult hypoxemia in non-White cohorts. Analyzing 151,070 oxygen saturation measured by pulse oximetry–arterial oxygen saturation readings in 46,253 patients, Burnett et al. find that occult hypoxemia was higher in Black and Hispanic cohorts than in White ones, in both bivariate (odds ratio, 1.44; 95% CI, 1.11 to 1.87) and multivariable analysis (odds ratio, 1.31; 95% CI, 1.03 to 1.68).1In the introduction, the authors state that their exposure “self-reported race/ethnicity” serves “as a surrogate for skin pigmentation.”1 Previous research has addressed the nonequivalence of socially constructed racial and ethnic categories and skin pigmentation.2,3 The use of this proxy can thus lead to incorrect interpretations or conclusions. Interpretation of this study’s data should focus upon race/ethnicity rather than skin pigmentation. Limitations of using race/ethnicity as a surrogate for skin pigmentation were acknowledged in the discussion of the article but were not sufficient given the scientific and social issues raised by the surrogacy.With the available data set, the article could have investigated race/ethnicity as the primary exposure without it serving as a surrogate for skin pigmentation. Explanations for findings could then have explored (but not been limited to) skin pigmentation, varied placement of pulse oximeters by providers, or different use of certain types of nail treatments among patients (or a combination thereof). That is, could there be nonbiologic reasons that Black and Hispanic cohorts are more likely to experience occult hypoxemia than White patients?4 This question is important given that socially constructed categories do not have meaningful biologic distinction.5 If future investigators are interested in studying the association of skin pigmentation and occult hypoxemia in the vein of previous research,6 they might consider performing a prospective study and use measurement tools such as the Fitzpatrick skin phototype classification.7Research on race and ethnicity in our field is important and much needed. I commend the authors for taking on this subject matter.The author declares no competing interests.