B54-38 Feasibility of Bedside Skin Tone Assessments to Identify Patients at Risk for Occult Hypoxemia

Abstract Study Rationale There is increasing concern and evidence that pulse oximetry may inaccurately overestimate oxygen saturation in darker-skin, meaning occult hypoxemia could lead to missed treatment. Relevant studies often rely on skin tone assessments that are unvalidated, expensive, and/or impractical (e.g. color measurements equipment or trained researchers using administered visual scales). Until such measurement bias is corrected, inexpensive, practical bedside skin tone assessments are needed to identify patients at risk for occult hypoxemia in research and clinical care. Methods Prospective, feasibility study of parents and two nurses assessing the skin tone of their child/patient using the Monk Skin Tone Scale (a 10-shade visual scale). Acceptability was revealed by participants’ willingness or refusal to provide such assessments. If the two nurses’ assessments were discordant, a third nurse was asked to make an assessment. Assessments made on patients’ skin were completed with an optimally printed score card (labeled 1-10) that was placed on the dorsal surface of the limb where the pulse oximeter was placed (e.g. right index, left toe). A neutral, 3100-Lumen flashlight was utilized to keep lighting consistent for scoring. Nurses also assessed their own skin tone. Spearman’s rank correlations using Bonferroni’s adjustment were estimated to see how well skin tone assessment correlated between parent and nurses and nurses themselves. A logistic regression was fitted to investigate if nurses with divergent skin tone were more likely to have discordant assessments of their patient’s skin tone. Results Seventeen patients have been assessed to date (2 in the lightest skin tone tertile; 4 in the darkest tertile). Indicating good acceptability of parents and nurses assessing their child’s/patient’s skin tone, 17 parents (out of 30 approached) assessed their children’s skin tone, an no nurse declined to assess skin tone. Skin tone assessments were highly correlated between parent and nurses (rho=0.92, p 0.0001 for both nurses) and between the two nurses (rho=0.87, p 0.0001). Four nurse dyads were discordant in their assessments; each differing by 1 shade. Nurses themselves having divergent skin tones was not associated with nurses’ discordant patient skin tone assessments (p = 0.51). Conclusion This ongoing study of skin tone assessments may suggest that having a parent and/or nurse assess skin tone using an administered visual scale is an acceptable, practical method to help identify patients at risk for occult hypoxemia in research and clinical care. This abstract is funded by: None