Performance of the MP570T pulse oximeter in volunteers participating in the controlled desaturation study: a comparison of seven probes

Clinical evaluation for the pharyngeal oxygen saturation measurements in shocked patients

Background

Monitoring oxygen saturation in shocked patients is a challenging nursing procedure. Shock syndrome alters peripheral tissue perfusion and hinders peripheral capillary oxygen saturation (SpO2) measurements. Our study aimed to find a solution to this problem. The pharynx is expected to be an accurate SpO2 measurement site in shocked patients. We clinically evaluated the pharyngeal SpO2 measurements against the arterial oxygen saturation (SaO2) measurements.

Methods

A prospective cohort research design was used. This study included 168 adult shocked patients. They were admitted to five intensive care units from March to December 2020 in an Egyptian hospital. A wrap oximeter sensor was attached to the posterior surface of an oropharyngeal airway (OPA) by adhesive tape. The optical component of the sensor adhered to the pharyngeal surface after the OPA insertion. Simultaneous pharyngeal peripheral capillary oxygen saturation (SpO2) and arterial oxygen saturation (SaO2) measurements were recorded. The pharyngeal SpO2 was clinically evaluated. Also, variables associated with the SpO2 bias were evaluated for their association with the pharyngeal SpO2 bias.

Results

The pharyngeal SpO2 bias was − 0.44% with − 1.65 to 0.78% limits of agreement. The precision was 0.62, and the accuracy was 0.05. The sensitivity to detect mild and severe hypoxemia was 100%, while specificity to minimize false alarm of hypoxemia was 100% for mild hypoxemia and 99.4% for severe hypoxemia. None of the studied variables were significantly associated with the pharyngeal SpO2 bias.

Conclusion

The pharyngeal SpO2 has a clinically acceptable bias, which is less than 0.5% with high precision, which is less than 2%.

Accuracy of a New Pulse Oximetry in Detection of Arterial Oxygen Saturation and Heart Rate Measurements: The SOMBRERO Study

Early diagnosis and continuous monitoring of respiratory failure (RF) in the course of the most prevalent chronic cardio-vascular (CVD) and respiratory diseases (CRD) are a clinical, unresolved problem because wearable, non-invasive, and user-friendly medical devices, which could grant reliable measures of the oxygen saturation (SpO2) and heart rate (HR) in real-life during daily activities are still lacking. In this study, we investigated the agreement between a new medical wrist-worn device (BrOxy M) and a reference, medical pulseoximeter (Nellcor PM 1000N). Twelve healthy volunteers (aged 20−51 years, 84% males, 33% with black skin, obtaining, during the controlled hypoxia test, the simultaneous registration of 219 data pairs, homogeneously deployed in the levels of Sat.O2 97%, 92%, 87%, 82% [ISO 80601-2-61:2017 standard (paragraph EE.3)]) were included. The paired T test 0 and the Bland-Altman plot were performed to assess bias and accuracy. SpO2 and HR readings by the two devices resulted significantly correlated (r = 0.91 and 0.96, p < 0.001, respectively). Analyses excluded the presence of proportional bias. For SpO2, the mean bias was −0.18% and the accuracy (ARMS) was 2.7%. For HR the mean bias was 0.25 bpm and the ARMS3.7 bpm. The sensitivity to detect SpO2 ≤ 94% was 94.4%. The agreement between BrOxy M and the reference pulse oximeter was “substantial” (for SpO2 cut-off 94% and 90%, k = 0.79 and k = 0.80, respectively). We conclude that BrOxy M demonstrated accuracy, reliability and consistency in measuring SpO2 and HR, being fully comparable with a reference medical pulseoxymeter, with no adverse effects. As a wearable device, Broxy M can measure continually SpO2 and HR in everyday life, helping in detecting and following up CVD and CRD subjects.

Ensuring that biomedical AI benefits diverse populations

Artificial Intelligence (AI) can potentially impact many aspects of human health, from basic research discovery to individual health assessment. It is critical that these advances in technology broadly benefit diverse populations from around the world. This can be challenging because AI algorithms are often developed on non-representative samples and evaluated based on narrow metrics. Here we outline key challenges to biomedical AI in outcome design, data collection and technology evaluation, and use examples from precision health to illustrate how bias and health disparity may arise in each stage. We then suggest both short term approaches-more diverse data collection and AI monitoring-and longer term structural changes in funding, publications, and education to address these challenges.