Nocturnal Scratch as a Digital Biomarker

Nocturnal scratch is the amount of scratching a person does while asleep, reported as duration, number of events, or proportion of the night. It exists as a digital measure because of a clean mismatch between what matters clinically and what can be reported.

Itch is consistently the symptom people with atopic dermatitis rank as most burdensome, and it is worst at night. But scratching during sleep is by definition not observed by the person doing it, so morning recall is an estimate of something they were unconscious for. Bed partner reports and video are impractical at scale.

Wrist accelerometry solves that directly, and it is one of the clearest cases in this library where a digital measure captures something no questionnaire realistically can, rather than merely capturing it more often.

Accelerometers on one or both wrists record movement through the night, and an algorithm identifies the characteristic rhythmic pattern of scratching and separates it from other sleep movement. Early work established that wrist activity monitors could quantify nocturnal scratching in atopic dermatitis, and later approaches applied machine learning to improve discrimination.

Validation uses infrared video annotated by trained observers as the reference standard, which is why studies are typically small and conducted in controlled overnight settings.

Because the same recording contains sleep information, contemporary work has developed nighttime scratch and sleep measures together from a single wrist device, which is efficient: the disease disrupts sleep through itch, so measuring both from one sensor describes the mechanism and the consequence at once.

The main use is as an objective endpoint in atopic dermatitis and chronic pruritus trials, where it complements clinician assessed severity and patient reported itch. Because it is objective and continuous, it can detect change earlier and with less noise than a weekly recalled itch score.

The second use is mechanistic: pairing scratch with sleep measures from the same night shows whether reduced itch actually translated into better sleep, which is one of the outcomes patients most want and which severity scores do not capture.

Measures are reported alongside the Eczema Area and Severity Index, which is the clinician assessed counterpart, and alongside the Dermatology Life Quality Index, which captures the wider impact on daily life.

Algorithms differ between research groups and are not standardised, so scratch counts from different studies are not directly comparable. Detection has to distinguish scratching from other rhythmic sleep movement, and performance in unrestricted home settings is less well characterised than the controlled validation studies suggest.

Wrist placement misses scratching done with other body parts, and bilateral recording is more accurate but doubles the wear burden.

Reference standards are also limited: video annotation is labour intensive and involves observer judgement about when scratching starts and stops. And while the measure is objective, the amount of change that matters to a person has not been firmly established, so a reduction in scratch minutes still needs a patient reported anchor to interpret.

• Mahadevan N, et al. Development of digital measures for nighttime scratch and sleep using wrist-worn wearable devices. NPJ Digit Med. 2021. pubmed.ncbi.nlm.nih.gov
• Moreau A, et al. Detection of nocturnal scratching movements in patients with atopic dermatitis using accelerometers and recurrent neural networks. IEEE J Biomed Health Inform. 2018. pubmed.ncbi.nlm.nih.gov
• Ebata T, et al. Use of a wrist activity monitor for the measurement of nocturnal scratching in patients with atopic dermatitis. Br J Dermatol. 2001. pubmed.ncbi.nlm.nih.gov
• Ikoma A, et al. Measurement of nocturnal scratching in patients with pruritus using a smartwatch: initial clinical studies. Acta Derm Venereol. 2019. pubmed.ncbi.nlm.nih.gov