Funded through our latest Health Seed Grant Round, Dr Amal Osman is investigating an urgent and overlooked issue: pulse oximeter accuracy in people with darker skin. 

This grant was possible through the generosity of Flinders Foundation supporters alongside funding from Flinders University. 

Pulse oximeters explained 

Pulse oximeters estimate oxygen levels by shining light through the skin. However, as Dr Osman explains, “The wavelengths of light used were determined by calibrating these devices in mostly people with light skin. The problem with this is that light interacts differently with dark skin.” 

Higher levels of melanin — the pigment responsible for darker skin — absorb and scatter light differently. This can lead to falsely high readings. 

“In most cases, high pulse oximeter readings close to 100% indicate normal oxygen levels. While levels below 90% may require urgent medical attention,” Dr Osman says. “Some people with dark skin have inaccurate higher pulse oximeter readings which falsely indicate the person is healthy when they require lifesaving treatment.” 

These inaccuracies can have serious consequences, affecting hospital admissions, emergency triage, diagnosis and treatment decisions. 

Why this matters for sleep apnoea 

One major focus of Dr Osman’s research is obstructive sleep apnoea — a condition where breathing repeatedly slows or stops during sleep. 

To diagnose sleep apnoea, clinicians monitor overnight oxygen levels. “The threshold for detecting reduced airflow episodes related to sleep apnoea breathing disturbances is an oxygen reduction of at least 3%,” Dr Osman explains. “This means that a pulse oximeter error of just 1% could be a major issue for accurate sleep apnoea diagnosis.” 

Untreated sleep apnoea is not simply about poor sleep. It is linked to serious health conditions including stroke, hypertension and obesity. 

A new and safer way to test accuracy 

Current methods for testing pulse oximeters are complex and involve lowering oxygen levels in healthy volunteers by having them breathe mixed gases — an approach that is difficult to scale and unsuitable for many people with chronic illness. 

Dr Osman and her team have developed an alternative. 

“We developed a new, alternative technique which allows us to experimentally measure changes in breathing during sleep using a test which mimics breathing with sleep apnoea,” she says. “This approach could complement the existing method to test pulse oximeters and allow us to compare responses between different devices to test for device accuracy.” 

Using this method, the team will carefully examine whether oxygen changes during sleep are measured differently depending on skin pigment, and whether this creates a clinically meaningful problem in diagnosing sleep apnoea for some Black patients. 

Driving more equitable healthcare 

Dr Osman describes this as “a 50 year old problem with a common and essential medical technology”, with the goal of this research “to improve how well pulse oximeters work for everyone.” She says “bias with medical technology can lead to harm which could have been avoided by simply supporting more inclusive research.” 

The findings will be shared with hospitals, manufacturers and regulators to strengthen testing standards and ensure safer, more accurate devices. 

Ultimately, this project aims to make sure that a simple fingertip device works equally well for everyone — and that no one misses care because of the colour of their skin.