FDA & ISO-Compliant SpO2 Testing
Accurate pulse oximeter testing is one of the most precise — and most scrutinized — areas of medical device validation. At Vital Signs Research Group, we conduct rigorous pulse oximeter testing and validation studies in our San Francisco clinical research facility, helping device manufacturers meet FDA 510(k) requirements and ISO 80601-2-61 standards.
Whether you’re developing a fingertip SpO2 sensor, a wearable continuous monitor, or a clinical-grade device for hospital use, we design studies built around your device and your timeline.
Our pulse oximeter validation studies follow the FDA’s guidance for non-significant risk devices and the ISO 80601-2-61 standard for SpO2 accuracy testing. Every study includes a diverse participant pool with a range of skin pigmentations — a requirement the FDA has emphasized in recent updated guidance — ensuring your data is both accurate and inclusive.
We recruit from a pre-screened participant database of hundreds of individuals, allowing us to meet study enrollment requirements quickly and without delay. Our clinical team includes registered nurses and experienced research coordinators who manage every step of the study day with precision and care.
Key elements of our pulse oximeter testing process include SpO2 accuracy testing across a saturation range of 70%–100%, controlled desaturation studies per ISO protocols, data collection paired with co-oximetry reference measurements, participant recruitment meeting FDA pigmentation distribution requirements, and a complete clinical report suitable for your FDA submission package. (confirm with client: typical study turnaround time)
We work with medical device startups, established manufacturers, and engineering teams preparing for their first FDA submission. If you’re not sure where to start, we’ll help you design a custom clinical protocol tailored to your device.
Contact Vital Signs Research Group today to discuss your pulse oximeter validation study — let’s move your device from the bench to clearance, one data point at a time.