Empirical Investigation of Pulse Oximetry Accuracy in Patients with Dark Skin Tones at Nigerian Teaching Hospitals

Published evidence from high-income countries demonstrates significant overestimation of blood oxygen saturation by pulse oximetry in patients with dark skin tones, a bias with potentially serious clinical consequences for Nigerian patients who are predominantly of darker skin pigmentation, yet empirical characterization of this bias in Nigerian clinical populations is absent from the literature. This study empirically investigates pulse oximetry accuracy across skin tone categories in patients at two Nigerian teaching hospitals in Anambra and Lagos states, measuring the magnitude, clinical significance, and patient safety implications of skin tone-related pulse oximetry bias. One hundred and fifty patients undergoing procedures requiring simultaneous pulse oximetry and arterial blood gas analysis were recruited across the two hospitals. Skin tone was objectively quantified using an Individual Typology Angle scale measured by a chromameter. SpO2 readings from three commercially available pulse oximeters at each measurement event were compared with arterial haemoglobin saturation from co-oximetry on co-drawn blood gas samples. Bias analysis stratified by skin tone category was performed following the Clarke-method bias analysis protocol. Results show a statistically significant mean positive bias of 2.1 percent SpO2 in the darkest skin tone tertile compared to 0.4 percent in the lightest tertile, consistent with published North American data. Clinically significant hidden hypoxaemia, defined as arterial saturation below 88 percent with pulse oximeter SpO2 above 92 percent, was four times more frequent in the darkest skin tone group. The study provides the first empirical characterization of this safety-relevant phenomenon in a Nigerian patient population. Keywords: pulse oximetry bias, skin tone, hidden hypoxaemia, patient safety, Nigerian patients.