This website is intended for healthcare professionals only.

Hospital Healthcare Europe
Hospital Healthcare Europe

Press Releases

Take a look at a selection of our recent media coverage:

Systematic review finds that cardiac data from smartwatches are potentially less accurate for darker skinned individuals

19th April 2022

A systematic review suggests that cardiac data obtained via a smartwatch is potentially less accurate for those of a darker skin tone

The cardiac data such as heart rate and rhythm obtained from a smartphone watch might not be as accurate for individuals with a darker skin tone. This was the conclusion of a systematic review by researchers from the University of Toronto, Canada, presented at the American College of Cardiology Conference 2022.

Cardiovascular disease is a leading global health issue and associated with an increasingly large economic burden. Technological innovations have become ingrained into everyday life and consumers are beginning to use consumer-grade software, such as smart wearables with numerous sophisticated sensors, to provide health insights. Today, commercial wearables can be used to collect cardiac data through electrocardiography (ECG) or photoplethysmography (PPG) and PPG has become the most popular technique for heart rate measurement. On the wrist, PPG detects blood flow rates by capturing the light intensity reflected from skin based on LEDs and photodetectors. However, questions have arisen over the accuracy of heart rate monitoring devices based on differences in skin tone. For example, because PPG detects changes in a beam of green light directed at the skin and since darker skin contains more melanin, it absorbs more green light than lighter skin. Moreover, previous research demonstrated that inaccurate PPG heart rate measurements occur up to 15% more frequently in dark skin as compared to light skin. In addition, pulse oximeter technology, which is also employed in smartwatches can be less accurate in darker skin as shown in one study, where Black patients had nearly three times the frequency of occult hypoxaemia, that was not detected by pulse oximetry as in White patients.

With smartwatches being used for health monitoring, for the present study, the team undertook a systematic review to determine the accuracy of cardiac data by wrist-worn wearable devices for participants of varying skin tones. They included studies in which heart rate and rhythm data were stratified according to the participant’s race and/or skin tone, which was measured using the Fitzpatrick score, which ranges from 1 to 6, with higher scores reflecting darker skin.

Cardiac data and skin tone

The literature search identified 10 studies with a total of 469 participants and the frequency-weighted Fitzpatrick score was reported in 6 of these studies, with 293 patients and the overall mean score was 3.5 (i.e., from the range of 1 – 6).

In 40% of studies, the researchers found a significant reduction in accuracy of heart rate measurements with a wearable device in those with darker skin compared to individuals with lighter skin tones and/or the gold standard measurements such as an ECG or a chest strap. Interestingly, one study found that wearable devices recorded significantly fewer data points for people with darker skin tones, despite no discrepancy in heart rate accuracy. A single study assessed ECG changes and noted a significant reduction in the accuracy of the R-R interval measurements in people with darker skin compared to ECG data (r = 0.98, p < 0.05).

Commenting on these findings, the lead author, Daniel Koerber, said “People need to be aware that there are some limitations for people with darker skin tones when using these devices, and the results should be taken with a grain of salt,” He added that “algorithms are often developed in homogeneous white populations, which may lead to results that are not as generalisable as we would like. Ongoing research and development of these devices should emphasise the inclusion of populations of all skin tones so that the developed algorithms can best accommodate for variations in innate skin light absorption.”

Koerber D et al. The effect of skin tone on accuracy of heart rate measurement in wearable devices: A systematic review. J Am Coll Cardiol 2022

Wearable devices can accurately predict clinical laboratory measurements

1st June 2021

Measurement of vital signs such as temperature, heart rate and the electrical properties of skin with a wearable device can be used to predict clinical laboratory results.

The routine measurement of parameters such as body temperature, heart and respiration rate, although non-specific, are of value in the overall assessment of a patient’s general wellbeing. In addition to the assessment of physical signs, further information can be gathered from laboratory analysis of blood or urine. However, all of these evaluations require that the person attends a clinic appointment. In recent years, the development of wearable technology has enabled the measurement of some vital signs such as heart rate and temperature but there has been limited research into the value of this longitudinally collected data. Nevertheless, it is possible that wearable technology has the potential to provide useful data for managing health conditions. But in order to fully utilise the data captured by wearable devices, it is necessary to determine whether these data are able to accurately mirror the results obtained in a clinic. This was the idea behind a study by a team from the Department of Genetics, Stanford University, US, which set out to determine if the vital signs collected from a wearable device (wVS) could be used as a non-invasive proxy measurement of clinical laboratory data using models of the relationship between wVS and clinical labs.

A total of 54 participants with a mean age of 56 years (44% male) were monitored for an average of 3.3 years. The wearable device measured heart rate, skin temperature, accelerometery and electrodermal activity (EDA). This latter measurement of the electrical properties of the skin, can assess skin hydration. Using heart rate, the researchers found that the wVS recordings was identical to the clinic-based measurements (both 71 beats per minute). Using machine learning models, the researchers extended their work and observed a high correlation between changes in wVS data and haematocrit, red blood cell count, haemoglobin and platelet count. For example, a higher body temperature and lower levels of movement (as assessed by accelerometery) tended to indicate illness, which matched up with a higher white blood cell count.

Commenting on their findings, the authors suggested that the continuous data collected by a wVS could be used to detect deviations from normal and thus serve as a means of identifying the need for further clinical laboratory evaluation. Although more work needs to be done, these initial results suggest that wearable devices have the potential to assess clinical parameters that, at present, can only be measured in laboratories.

Dunn J et al. Wearable sensors enable personalised predictions of clinical laboratory measurements. Nat Med 2021