A pulse oximeter looks like one of the simplest devices a screening programme buys: clip it on a finger, read a number. But pulse oximeter accuracy is not uniform across every patient, and the gap has become one of the most closely scrutinised issues in medical device regulation. For any clinic or population-health programme specifying screening equipment, understanding why that gap exists — and what to ask a manufacturer before signing a purchase order — matters as much as comparing price per unit.
Why Pulse Oximeter Accuracy Varies by Skin Tone
Pulse oximeters estimate blood oxygen saturation by shining light through tissue and measuring how much is absorbed. Research reviewed by regulators has consistently found that this method can overestimate oxygen saturation in patients with darker skin pigmentation, meaning a reading can look reassuring when a patient is actually hypoxaemic. For a screening programme, that is not an abstract statistical footnote — it is a plausible cause of a genuinely low reading being missed, and a delay in the follow-up or referral that should have happened.
What UK and International Standards Are Changing
This is not a fringe concern; it has reached the centre of device regulation on both sides of the Atlantic. An independent review commissioned by the UK's Department of Health and Social Care reported in March 2024, and the government backed its recommendations. Among them: that the MHRA and UK-approved bodies should strengthen approval standards for new pulse oximeters so that manufacturers must supply clinical data demonstrating accuracy across the full range of skin tones, with greater population representativeness in the studies used for testing and calibration. The review also called for manufacturers to update instructions for use so that clinicians and patients are told plainly about any differential accuracy across skin pigmentation, and for the MHRA to issue clearer guidance requiring manufacturers to make that performance data transparent.
The international technical standard for pulse oximeter equipment, ISO 80601-2-61, is itself being revised to address the same issue, with melanin identified as the chromophore driving the pigmentation-related bias; the update is expected to bring in colorimetry measures and revised requirements for how diverse a clinical desaturation study's participants need to be. In the US, the FDA's existing premarket guidance had asked studies to include a minimum of two darkly pigmented subjects or fifteen percent of the study group, whichever was larger — a bar regulators have acknowledged manufacturers did not consistently meet. Draft guidance published in January 2025 goes considerably further: raising the minimum number of test subjects from ten to one hundred and fifty, recommending both a subjective ten-level skin tone scale and an objective, numerical measurement of pigmentation, and proposing a labelling statement for devices that can actually demonstrate comparable performance across skin tones.
What to Ask Manufacturers Before Procuring
None of this means every pulse oximeter on the market today is unsuitable for screening use. It does mean that "CE marked" or "MHRA registered" is a floor, not a guarantee of even performance across a patient population — and that procurement conversations need to go a step further than they may have five years ago. A programme specifying blood-pressure monitors, pulse oximeters, and metabolic-screening tools should be asking suppliers directly:
- What is the demographic and skin-tone composition of the clinical studies behind this device's accuracy claims, and can that data be supplied in writing?
- Does the manufacturer report any known bias in accuracy at lower oxygen saturation levels, where the clinical consequences of an inaccurate reading are greatest?
- What do the instructions for use say about limitations related to skin pigmentation, and is that information something frontline staff will actually see?
- Is the device being actively updated in line with the revised ISO 80601-2-61 requirements as they are finalised, or is it running on an older accuracy dossier?
- Has the manufacturer published, or committed to publish, real-world performance data alongside the original laboratory studies?
The same instinct is worth applying more broadly across a screening equipment budget. Blood-pressure monitors have their own accuracy standards and cuff-size requirements, and metabolic-screening tools such as glucometers depend on regular calibration to stay trustworthy — pulse oximeters are simply the device where the pigmentation evidence gap has been most publicly documented and most directly tied to a specific patient-safety concern. Asking a supplier for the evidence behind an accuracy claim, rather than accepting a certification mark at face value, is a habit that serves a programme well across the whole screening equipment list, not only for oximetry.
Building This Into Everyday Practice
Procurement is only the first half of the answer. Regulators have also been clear that clinicians and screening staff should be trained to treat a single pulse oximeter reading as one data point rather than an absolute threshold — watching the trend over several readings, cross-checking against other clinical signs, and knowing when a reading that looks borderline-normal still warrants a closer look. For programmes running screening at scale across a diverse population, that combination — better procurement questions up front, and realistic expectations about the limits of any single device — is what turns a good-looking spec sheet into a genuinely reliable screening tool.
Conclusion
Pulse oximeter accuracy is under more regulatory scrutiny now than at any point since the device became a routine part of clinical screening, and the standards a screening programme should hold suppliers to are visibly tightening. Building manufacturer evidence checks into procurement, rather than treating a certification mark as the end of due diligence, is the practical way clinics and population-health programmes can keep pace with where the standards are heading.