CERSI Collaborator: University of California San Francisco: Philip E. Bickler, MD, PhD; Michael Lipnick, MD; Carolyn Hendrickson, MD, MPH; John R. Feiner, MD

FDA Collaborators: Nilsa Loyo-Berrios, PhD; Malvina Eydelman, MD; Allison O’Neill, PhD; Michelle Tarver, MD, PhD; Gene Pennello, PhD; Rebecca Ward, MPH; Rebecca Torguson, MPH; Maritze Ortega, MS; Christine Merenda, MPH, BSN, RN

Project Start Date: May 2, 2022

Regulatory Science Challenge

A pulse oximeter is a device that is usually placed on a fingertip to estimate the amount of oxygen present in the blood. These devices were used often during the COVID-19 pandemic. This study will address some concerns related to the accuracy of pulse oximeter readings among adult (22 years old and older) patients with darker skin pigmentation (color). The project will assess if the pulse oximeter errors are related to skin pigmentation or other factors such as patient’s poor perfusion (flow of blood) or other health conditions.

Evidence from the scientific literature suggests pulse oximeters may be less accurate in patients with darker skin pigmentation, as compared to patients with lighter skin pigmentation. Additional studies are needed to increase our understanding of the relationship between the accuracy of pulse oximeter readings and skin pigmentation. This understanding can be improved using information from skin pigmentation assessments, and comparing blood oxygen level measurements by pulse oximeter (SpO2) with measurements by arterial blood line (SaO2), measured at the same time. This study will also evaluate if there are factors other than skin pigmentation that may impact the accuracy of pulse oximeter readings (e.g., low perfusion, patients’ health conditions).

Understanding how pulse oximeter devices perform in clinical practice is important in protecting public health. Findings from this study will inform health care providers, patients, and the regulatory evaluation and monitoring of pulse oximeters.

Project Description and Goals

The goal of this study is to evaluate pulse oximeter performance in hospitalized adult patients of different skin pigmentation levels. The study will assess the level of error in pulse oximeter readings for oxygen blood saturation levels (SpO2) across skin pigmentation groups. The study will also assess the extent that factors such as low perfusion and/or light transmittance may impact the accuracy of pulse oximeter readings.

The study is designed to capture data on skin pigmentation, SpO2, and SaO2, measurement of patients’ flow of blood to the extremities of the body (peripheral perfusion), type of pulse oximeter, as well as probe and site of probe placement, among different patient populations. Skin pigmentation will be captured through the use of colorimetry tools, and numerical scales for classification (Fitzpatrick and von Luschan skin color scales). Data on other factors that may impact the performance of pulse oximeters such as demographics, body levels of carboxyhemoglobin (compound formed from binding of carbon monoxide and blood hemoglobin), presence of multiple medical conditions, supplemental oxygen therapy, and core temperature will be captured from electronic health records. Patients with SaO2 values in the high 80% to low 90% will be included in the study.

The study is being conducted at the University of California San Francisco (UCSF) Health Medical System. Subjects being treated at the Intensive Care Unit (ICU), operating room, or the emergency room, who have an arterial blood line in place qualify for study participation.