Protection-adjusted UV dose estimated for body areas: Daily self-reported sun protection modification of wearable UV sensor dose

Digital Phenotyping in Health Using Machine Learning Approaches: Scoping Review

BACKGROUND

Digital phenotyping is the real-time collection of individual-level active and passive data from users in naturalistic and free-living settings via personal digital devices, such as mobile phones and wearable devices. Given the novelty of research in this field, there is heterogeneity in the clinical use cases, types of data collected, modes of data collection, data analysis methods, and outcomes measured.

OBJECTIVE

The primary aim of this scoping review was to map the published research on digital phenotyping and to outline study characteristics, data collection and analysis methods, machine learning approaches, and future implications.

METHODS

We utilized an a priori approach for the literature search and data extraction and charting process, guided by the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-analyses Extension for Scoping Reviews). We identified relevant studies published in 2020, 2021, and 2022 on PubMed and Google Scholar using search terms related to digital phenotyping. The titles, abstracts, and keywords were screened during the first stage of the screening process, and the second stage involved screening the full texts of the shortlisted articles. We extracted and charted the descriptive characteristics of the final studies, which were countries of origin, study design, clinical areas, active and/or passive data collected, modes of data collection, data analysis approaches, and limitations.

RESULTS

A total of 454 articles on PubMed and Google Scholar were identified through search terms associated with digital phenotyping, and 46 articles were deemed eligible for inclusion in this scoping review. Most studies evaluated wearable data and originated from North America. The most dominant study design was observational, followed by randomized trials, and most studies focused on psychiatric disorders, mental health disorders, and neurological diseases. A total of 7 studies used machine learning approaches for data analysis, with random forest, logistic regression, and support vector machines being the most common.

CONCLUSIONS

Our review provides foundational as well as application-oriented approaches toward digital phenotyping in health. Future work should focus on more prospective, longitudinal studies that include larger data sets from diverse populations, address privacy and ethical concerns around data collection from consumer technologies, and build "digital phenotypes" to personalize digital health interventions and treatment plans.

Sun exposure reduction by melanoma survivors with wearable sensor providing real-time UV exposure and daily text messages with structured goal setting

Despite knowledge of subsequent melanoma risk and the benefit of sun protection in risk reduction, melanoma survivors often do not engage in adequate sun protection and continue to sunburn at rates similar to individuals without a history of skin cancer. This novel intensive intervention provided a wearable UV sensor delivering real-time UV exposure with a smartphone and daily text messages. On days 1-10 (period 1), behavioral facilitation and outcome expectancies messages were provided. On day 10, participants reviewed and reflected on their daily UV exposure on the previous 10 days and set goals for improving sun protection. Then on days 11-21 (period 2) self-efficacy and self-regulation messages were provided. Sixty melanoma survivors were randomized (1:1) to receive structured or unstructured goal setting queries on day 10. Controlling for cloudy/rain conditions with less UV due to weather, there was a time effect with a significant decrease in UV exposure from periods 1-2 [period 1-2, F (59) = 22.60, p < 0.0001]. In this short-term study, melanoma survivors managed their daily UV exposure to stay below their maximum tolerated UV dose. ClinicalTrials.gov Protocol Record NCT0334796, date of registration Nov 15, 2017.

Real-Time UV Measurement With a Sun Protection System for Warning Young Adults About Sunburn: Prospective Cohort Study

Background

Melanoma is attributable to predisposing phenotypical factors, such as skin that easily sunburns and unprotected exposure to carcinogenic UV radiation. Reducing the proportion of young adults who get sunburned may reduce the incidence of melanoma, a deadly form of skin cancer. Advances in technology have enabled the delivery of real-time UV light exposure and content-relevant health interventions.

Objective

This study aims to examine the feasibility of young adults performing the following tasks daily: wearing a UV dosimeter, receiving text messages and real-time UV-B doses on their smartphone, and responding to daily web-based surveys about sunburn and sun protection.

Methods

Young adults aged 18-39 years (n=42) were recruited in the United States in June 2020 via social media. Participants received the UV Guard sun protection system, which consisted of a UV dosimeter and a smartphone app. During 3 consecutive periods, intervention intensity increased as follows: real-time UV-B dose; UV-B dose and daily behavioral facilitation text messages; and UV-B dose, goal setting, and daily text messages to support self-efficacy and self-regulation. Data were self-reported through daily web-based surveys for 28 days, and UV-B doses were transmitted to cloud-based storage.

Results

Patients’ median age was 22 years (IQR 20, 29), and all patients had sun-sensitive skin. Sunburns were experienced during the study by fewer subjects (n=18) than those in the preceding 28 days (n=30). In July and August, the face was the most commonly sunburned area among 13 body locations; 52% (22/42) of sunburns occurred before the study and 45% (19/42) occurred during the study. The mean daily UV-B dose decreased during the 3 periods; however, this was not statistically significant. Young adults were most often exercising outdoors from 2 to 6 PM, walking from 10 AM to 6 PM, and relaxing from noon to 2 PM. Sunburn was most often experienced during exercise (odds ratio [OR] 5.65, 95% CI 1.60-6.10) and relaxation (OR 3.69, 95% CI 1.03-4.67) relative to those that did not exercise or relax in each category. The self-reported exit survey indicated that participants felt that they spent less time outdoors this summer compared to the last summer because of the COVID-19 pandemic and work. In addition, 38% (16/42) of the participants changed their use of sun protection based on their app-reported UV exposure, and 48% (20/42) shifted the time they went outside to periods with less-intense UV exposure. A total of 79% (33/42) of the participants were willing to continue using the UV Guard system outside of a research setting.

Conclusions

In this proof-of-concept research, young adults demonstrated that they used the UV Guard system; however, optimization was needed. Although some sun protection behaviors changed, sunburn was not prevented in all participants, especially during outdoor exercise.

Trial Registration

ClinicalTrials.gov NCT03344796; http://clinicaltrials.gov/ct2/show/NCT03344796