Can cleaning, shaving, and securing a wrist heart-rate monitor improve inaccurate readings during cardiac rehabilitation?

Details

Wrist-worn photoplethysmography (PPG) monitors could help track physical activity in cardiac patients, but their heart-rate accuracy can vary, particularly during intensive exercise. Research on how device positioning, sensor cleanliness and body hair affect PPG accuracy has been limited. This study assessed whether technical adjustments could improve exercise heart-rate accuracy in cardiac patients with low PPG accuracy, and whether accuracy differed by exercise type.

In this before-and-after technical optimisation study, 30 patients attending a cardiac rehabilitation programme wore both a PPG-based wrist heart-rate monitor (Fitbit Inspire 2) and an ECG-based reference chest-strap monitor (Polar H10) during training sessions. For the baseline assessment, each patient completed three sessions following standard guidance: the wrist monitor was placed three fingers above the wrist joint and tightened securely but comfortably.

The abstract defines low baseline accuracy as a mean absolute percentage error (MAPE) ≥10% during ≥70% of training time. Patients meeting this definition entered an optimisation phase with three combined adjustments: degreasing the PPG sensor, shaving part of the forearm to reduce interference from hair, and securing the monitor at the wrist with tape and gauze to maintain its position and reduce movement-related artefacts. Accuracy was reassessed after three further training sessions. The researchers also conducted subanalyses for each exercise subtype performed during training.

The 30 patients had a mean age of 58.1 years, and 33.3% were female. At baseline, 20 patients (66.7%) had high PPG accuracy, with accurate readings (MAPE <10%) during an average of 88.4±5.2% of training time. The 10 patients with low baseline accuracy had accurate readings during an average of 54.3±15.7% of training time.

After optimisation, three of these 10 patients (30%) improved above the accuracy threshold, with a reported average increase of 17.4±7.1% in accurate measurements and a post-optimisation mean of 83.5±5.4%. The remaining seven patients (70%) continued to have low accuracy, with a post-optimisation mean of 56.4±17.8%. Those who improved had a higher baseline percentage of accurate training time than those who did not improve—66.0±3.8% versus 49.3±16.3%—but this difference was not statistically significant (p=0.08; Figure 1).

At baseline, exercises involving intense arm movements—rowing and arm biking—had a lower percentage of accurate PPG measurements than other exercises: 56.6% ± 4.3% versus 78.2% ± 8.8% (p=0.03; Figure 2).

The researchers concluded that reliable PPG-based physical activity monitoring is feasible only for a subgroup of cardiac patients. They also concluded that technical adjustments contribute minimally to improving PPG accuracy, with only 30% of patients with low baseline accuracy improving after the intervention.