Desenvolvimento de um algoritmo para caracterização e monitoramento de insônia em pessoas idosas utilizando a plataforma de monitoramento remoto sênior saúde móvel

Abstract

Introduction: Insomnia is defined as the individual report of difficulty sleeping. It is commonly diagnosed in the older people, as this age group tends to have more fragmented sleep and presents more comorbidities that interfere with nocturnal and daytime functioning. Among the methods to evaluate sleep, wearable devices, particularly smartwatches like recent Fitbit models, are gaining prominence. These devices include photoplethysmography-based sensors and motion sensors such as accelerometers and gyroscopes, which estimate not only sleep parameters and stages but also heart rate during wakefulness and sleep. Considering the importance of monitoring technologies with devices that capture physiological signals non-invasively, the Sênior Saúde Móvel (SSM) platform was developed to track these signals and generate health data. Objective: To develop an algorithm for characterizing and monitoring insomnia in elderly individuals using the remote monitoring platform SSM. Methodology: This study focuses on technological development with a cross-sectional, observational, descriptive, analytical, and quantitative approach. The algorithm was developed at the Strategic Health Technology Center, in the Biomedical Computing Laboratory, in Campina Grande, PB, Brazil. Sleep data were collected from participants at the Condomínio Cidade Madura and the Elderly Community Center at the Universidade Aberta à Maioridade. The sample consisted of individuals aged 60 years or older of both sexes. Participants were evaluated and remotely monitored for 7 consecutive days using Fitbit smartwatches. Sleep data were extracted through Grafana, and data analysis was conducted using IBM SPSS Statistics 20 (Statistical Package for the Social Sciences), with a significance level of 5% (p<0.05) for all tests. Results: The algorithm was developed using Python and employed data on "sleep stages," "latency," and "awakenings" to characterize "early awakenings," which occurred on average 1.56 ± 1.08 times per week; "maintenance insomnia," occurring 2.18 ± 1.63 times per week; and "initial insomnia," occurring 0.06 ± 0.32 times per week. Conclusion: The development of the algorithm was successfully carried out with a multidisciplinary team. However, despite the good performance of wearable devices, further investigation is needed into heart rate variability to better clarify awakenings and sleep latency, as well as to validate the algorithm.