scholarly journals Fitbit data show poor correlation with measures of activity and sleep among hospitalized general medicine patients

2020 ◽  
Author(s):  
Robert Clark Wu ◽  
Vikas Patel ◽  
Sabreena Moosa ◽  
Laura Langer ◽  
Thomas E. MacMillan ◽  
...  
Keyword(s):  
Heart Rate ◽  
Sleep Duration ◽  
Pearson Correlation ◽  
Hospital Setting ◽  
Poor Correlation ◽  
Poor Sleep ◽  
Medical Patients ◽  
Activity Data ◽  
Heart Rates ◽  
Patient Reported

Abstract BackgroundWearable devices such as Fitbits may provide important insights about hospitalized patients that include data on low activity and poor sleep. Monitoring this information could spur interventions to improve mobility and sleep which may reduce the adverse effects associated with hospitalization. However, there is a lack of studies assessing the accuracy of wearables in hospitalized medical patients. The purpose of our study was to determine the accuracy of Fitbit heart rate, sleep and physical activity in hospitalized medical patients.MethodsWe conducted a prospective cohort feasibility study enrolling 50 medical inpatients at two hospitals providing them with a Fitbit Charge. Our main measures were Fitbit heart rate, sleep and activity data as well as nurse recorded heart rates, patient reported sleep, and nurse assessments of activity.ResultsOf the 50 patients who consented to the study, 47 patients wore the devices. Comparing pairs of heart rate data from Fitbit and nurse recorded vital signs for the same minute, there were 261 pairs available for comparison. The mean difference was 0.45 bpm (SD: 13.0, Pearson correlation: 0.68 P<0.001) and the 95% limits of agreement were -25 to 26 bpm. The association between the patient-reported sleep score and Fitbit total sleep duration was 0.19 (P=0.24) and between the self-reported hours of sleep and Fitbit total sleep duration was 0.21 (P=0.21). The correlation between nurse-recorded activity and Fitbit daily steps was 0.06 (P=0.52). ConclusionsFitbit heart rates correlated well with nurse-recorded heart rate but did not correlate well with nurse assessments of activity nor with patient self-assessment of sleep. This study highlights limitations of the accuracy of current wearable wrist-worn device algorithms in activity and sleep detection in patients in hospital. The findings call into question the validity of Fitbits for assessment of patient activity and sleep in the hospital setting and suggest that they should not be routinely used without further validation.Trial RegistrationClinicalTrials.gov NCT03646435

scholarly journals Fitbit data show poor correlation with measures of activity and sleep among hospitalized general medicine patients

2020 ◽  
Author(s):  
Robert Clark Wu ◽  
Vikas Patel ◽  
Sabreena Moosa ◽  
Laura Langer ◽  
Thomas E. MacMillan ◽  
...  
Keyword(s):  
Heart Rate ◽  
Sleep Duration ◽  
Pearson Correlation ◽  
Hospital Setting ◽  
Poor Correlation ◽  
Poor Sleep ◽  
Medical Patients ◽  
Activity Data ◽  
Heart Rates ◽  
Patient Reported

Abstract Background: Wearable devices such as Fitbits may provide important insights about hospitalized patients that include data on low activity and poor sleep. Monitoring this information could spur interventions to improve mobility and sleep which may reduce the adverse effects associated with hospitalization. However, there is a lack of studies assessing the accuracy of wearables in hospitalized medical patients. The purpose of our study was to determine the accuracy of Fitbit heart rate, sleep and physical activity in hospitalized medical patients.Methods: We conducted a prospective cohort feasibility study enrolling 50 medical inpatients at two hospitals providing them with a wrist-worn Fitbit Charge. Our main measures were Fitbit heart rate, sleep and activity data as well as nurse recorded heart rates, patient reported sleep, and nurse assessments of activity.Results: Of the 50 patients who consented to the study, 47 patients wore the devices. Comparing pairs of heart rate data from Fitbit and nurse recorded vital signs for the same minute, there were 261 pairs available for comparison. The mean difference was 0.45 bpm (SD: 13.0, Pearson correlation: 0.68 P<0.001) and the 95% limits of agreement were -25 to 26 bpm. The association between the patient-reported sleep score and Fitbit total sleep duration was 0.19 (P=0.24) and between the self-reported hours of sleep and Fitbit total sleep duration was 0.21 (P=0.21). The correlation between nurse-recorded activity and Fitbit daily steps was 0.06 (P=0.52). Conclusions: Fitbit heart rates correlated well with nurse-recorded heart rate but did not correlate well with nurse assessments of activity nor with patient self-assessment of sleep. This study highlights limitations of the accuracy of current wearable wrist-worn device algorithms in activity and sleep detection in patients in hospital. The findings call into question the validity of Fitbits for assessment of patient activity and sleep in the hospital setting and suggest that they should not be routinely used without further validation.Trial Registration: ClinicalTrials.gov NCT03646435

scholarly journals How well are heart rates measured by pulse oximeters and electronic sphygmomanometers? Practice-based evidence from an observational study of acutely ill medical patients during hospital admission

2019 ◽  
Vol 18 (3) ◽  
pp. 144-147
Author(s):  
Mary Rimbi ◽  
◽  
Immaculate Nakitende ◽  
Teopista Namujwiga ◽  
John Kellett ◽  
...  
Keyword(s):  
Heart Rate ◽  
Observational Study ◽  
Hospital Admission ◽  
Pulse Oximeter ◽  
Medical Patients ◽  
Limits Of Agreement ◽  
Heart Rates ◽  
Electrocardiogram Ecg ◽  
Pulse Oximeters

Background: heart rates generated by pulse oximeters and electronic sphygmomanometers in acutely ill patients may not be the same as those recorded by ECG Methods: heart rates recorded by an oximeter and an electronic sphygmomanometer were compared with electrocardiogram (ECG) heart rates measured on acutely ill medical patients. Results: 1010 ECGs were performed on 217 patients while they were in the hospital. The bias between the oximeter and the ECG measured heart rate was -1.37 beats per minute (limits of agreement -22.6 to 19.9 beats per minute), and the bias between the sphygmomanometer and the ECG measured heart rate was -0.14 beats per minute (limits of agreement -22.2 to 21.9 beats per minute). Both devices failed to identify more than half the ECG recordings that awarded 3 NEWS points for heart rate. Conclusion: Heart rates of acutely ill patients are not reliably measured by pulse oximeter or electronic sphygmomanometers.

scholarly journals ADOLESCENTS REPORT SLEEP QUALITY IMPAIRMENTS ACUTELY POST-CONCUSSION

2020 ◽  
Vol 8 (4_suppl3) ◽  
pp. 2325967120S0022
Author(s):  
Julie C. Wilson ◽  
Michael W. Kirkwood ◽  
Aaron J. Provance ◽  
Gregory A. Walker ◽  
Pamela E. Wilson ◽  
...  
Keyword(s):  
Sleep Quality ◽  
Sleep Duration ◽  
Negative Impact ◽  
Independent Effect ◽  
Poor Sleep ◽  
Post Injury ◽  
Time In Bed ◽  
Patient Reported ◽  
Concussion History ◽  
Chi Square Analysis

Background: Recently, the association between concussion and sleep has been explored. Few studies have used validated patient-reported outcome measures to examine how sleep quality differs between adolescents who have recently sustained a concussion compared to healthy adolescents. Hypothesis/Purpose: Our purpose was to evaluate the effect of acute concussion on sleep quality using the Pittsburgh Sleep Quality Index (PSQI). We hypothesized that adolescents with acute concussion symptoms would report worse sleep quality compared to healthy adolescent athletes undergoing a pre-participation evaluation. Methods: Patients seen for initial evaluation at a sports medicine clinic within 14 days of sustaining a concussion were matched with healthy control subjects undergoing a pre-participation evaluation based on age, sex and concussion history at a 3:1 control:concussion subject ratio. Controls with history of concussion were included if fully recovered at their evaluation. Both groups completed the PSQI and Post-Concussion Symptom Inventory (PCSI). Those with concussion reported sleep quality since their injury, while controls reported sleep quality over the past month. The primary outcome was global PSQI score, where higher scores indicate worse sleep quality. We compared sleep quality ratings between groups using independent samples t-test, and proportion of poor sleep ratings using Chi-square analysis. We then constructed a multivariable regression model to assess the effect of acute concussion on global PSQI scores while adjusting for sex, age and concussion history. Results: We evaluated 17 patients with concussion (mean 8±3 days post-injury) who were matched with 51 controls. Demographic variables did not differ significantly between concussion and control groups (Table 1). The concussion group reported higher PCSI symptom rating (Table 1), and worse overall sleep quality (Table 2). Specifically, the concussion group reported longer time to fall asleep, later morning awakening, longer sleep duration and longer time in bed, compared to controls. The concussion group was also less likely to rate their sleep as “very good” compared to controls (Table 2). After adjusting for the independent effect of sex, age, and concussion history, concussion was significantly associated with higher PSQI scores (β=2.01, 95% CI= 0.35, 3.68; p= 0.019). Conclusion: Adolescents assessed within 14 days of concussion reported worse overall sleep quality, longer sleep duration, and decreased sleep efficiency compared to healthy adolescents. These results, obtained utilizing a validated measure, highlight the negative impact of acute concussion on sleep quality and provide additional support that assessment of sleep characteristics is an important component of acute concussion management. [Table: see text][Table: see text]

scholarly journals Bidirectional Associations Between Activity Diversity and Sleep

2020 ◽  
Vol 4 (Supplement_1) ◽  
pp. 652-652
Author(s):  
Soomi Lee ◽  
Susan Charles ◽  
David Almeida
Keyword(s):  
Sleep Duration ◽  
Well Being ◽  
Total Activity ◽  
The United States ◽  
Poor Sleep ◽  
Short Sleep Duration ◽  
Sleep Health ◽  
Activity Data ◽  
Bidirectional Relationship ◽  
Bidirectional Associations

Abstract Activity diversity is important for psychological well-being and cognitive functioning. Yet, little is known about the relationship between activity diversity and sleep. This study examined how overall and nightly sleep health are associated with activity diversity. Participants (N=1841) from the Midlife in the United States Study II provided activity data for 8 days. We constructed overall and daily activity diversity scores. A composite score of overall sleep health across 8 dimensions and nightly sleep duration were measured. Analyses adjusted for sociodemographics, total activity time, and positive/negative affect. Participants with poorer sleep health overall had a lower activity diversity. On days following nights with short (&lt;6hrs) or long (&gt;8hrs) sleep duration, participants engaged in fewer-than-usual activities. Conversely, fewer daily activities also predicted long (but not short) sleep duration. Our results suggest cyclical associations between poor sleep health and activity diversity day-to-day, which, may accumulate over time to form a bidirectional relationship.

scholarly journals 1033 Sleep and Patient-Reported Outcomes in Persons With Heart Failure With Preserved Ejection Fraction

SLEEP ◽  
2020 ◽  
Vol 43 (Supplement_1) ◽  
pp. A392-A393
Author(s):  
L M Baniak ◽  
C W Atwood ◽  
P J Strollo ◽  
D E Forman ◽  
E R Chasens
Keyword(s):  
Quality Of Life ◽  
Heart Failure ◽  
Sleep Quality ◽  
Sleep Duration ◽  
Patient Reported Outcomes ◽  
Summary Score ◽  
Poor Sleep ◽  
Preserved Ejection Fraction ◽  
Patient Reported

Abstract Introduction Sleep quality has a significant bearing on disease. A better understanding of sleep quality may help identify opportunities to improve patient-reported outcomes (PROs) in persons with heart failure with preserved ejection fraction (HFpEF). We aimed to explore the association between sleep and PROs in patients clinically diagnosed with HFpEF. Methods Cross-sectional study of 22 participants (71.2±7.2 years, 95% male, 86.4% white) with HFpEF, recruited from a heart failure (n=14) and sleep clinic (n=8). Sleep disordered breathing was measured objectively using one-night in-home obstructive sleep apnea (OSA) testing (ApneaLink). Actigraphy (7 days) was used to assess sleep duration, efficiency, and wake after sleep onset (WASO). Subjective sleep measures included the Insomnia Severity Index (ISI), Epworth Sleepiness scale (ESS), and Pittsburgh Sleep Quality Index (PSQI). PROs included functional status (Functional Outcomes Sleep Questionnaire [FOSQ]), depression (PROMIS Depression), fatigue (PROMIS Fatigue), and heart failure specific quality of life (Kansas City Cardiomyopathy Questionnaire [KCCQ]; overall summary score [KCCQ-os] and clinical summary score [KCCQ-cs]). The KCCQ-cs includes physical function and symptom scores to corresponds with NYHA Classification. Results Fifty percent of the participants had poor sleep quality (PSQI &gt;5) and 2 (9.1%) had ISI scores &gt;14. The majority (64%; n=14) had OSA; 10 currently on OSA therapy. Mean oxygen desaturation index (ODI) was 20.8±17.8. Mean actigraphy data indicated poor sleep (sleep duration 302±116 minutes, sleep efficiency 70.0±18.6%, and WASO 52±28 minutes) despite only 5 (22.7%) participants reporting excessive daytime sleepiness (ESS&gt;10). Greater insomnia symptom severity was associated with lower heart failure specific quality of life (KCCQ-os) and functional status, and, greater fatigue and depression (all p-values &lt;.05). FOSQ was negatively associated with PSQI (r= -.710, p= &lt;.001) and positively with sleep efficiency (r=.496, p=.026). The KCCQ-cs was positively associated with sleep duration (r=.496, p=.026) and negatively but not significantly associated with ODI (r= -.453, p=.07). Conclusion Impaired sleep and OSA are highly prevalent in patients with HFpEF and both are adversely associated with PROs. Goals to improve sleep is important for effective symptom management and for potential improvements in PROs. Support American Nurses Foundation, Preventative Cardiovascular Nurses Association

scholarly journals MP49: Does reduced cough capacity in minor thoracic trauma leads to more atelectasis development?

CJEM ◽  
2020 ◽  
Vol 22 (S1) ◽  
pp. S60-S60
Author(s):  
M. Emond ◽  
S. Hegg ◽  
E. Thériault
Keyword(s):  
Emergency Department ◽  
Odds Ratio ◽  
Thoracic Trauma ◽  
Smoking Status ◽  
Pearson Correlation ◽  
Poor Correlation ◽  
Outpatient Basis ◽  
Initial Visit ◽  
Significant Difference ◽  
Patient Reported

Introduction: Minor thoracic trauma (MTI) accounts for approximately 15% of all injuries treated in the emergency department (ED). Many of which are minor and will be handle on an outpatient basis. MTI and rib fractures especially cause non-negligible pain. The pain experienced by patients can lead to reduce pulmonary function, decrease mucous clearance and decrease cough capacity leading in infectious problems and atelectasis. To our knowledge, there is no study of atelectasis development caused by reduced cough capacity in the setting of MTI. Objective: Evaluate if a variation in cough capacity leads to atelectasis development. Evaluate if there was a difference in cough capacity perception between the nurse, the physician and the patient himself. Methods: A prospective observational cohort study (2006-2012) in 4 ED recruited patients with a chief complaint of MTI, ≥ 16 years old, discharged home from the ED. Exclusion criteria: 1) a confirmed hemothorax, pneumothorax, fail chest, lung contusion or any other important thoracic or abdominal internal injury at the initial visit or unable to attend follow-up visits. Patients were assessed at 7- and 14- days. For each patient, age, sex, mechanism of injury, dyspnea, COPD/asthma and smoking status were collected. Chest x-ray was done at each visit; pulmonary complications were assessed by a blind radiologist. Cough capacity was assessed on a scale of 0 to 10 by a nurse, physician and patient himself at 0, 7- and 14- days. Pain was scored on a scale of 0 to 10. Chi -squared and odds ratio (IC: 95%. p ≤ 0.05) were assessed to determine if the cough capacity variation leads to atelectasis development. A Pearson correlation test was assessed the correlation in cough capacity among participants. Results: 1474 patients were recruited. Initial visit: 9% had atelectasis, 7 days: 7% and 4.6% at 14 days. 1105 patients were retained for analysis after exclusion of missing data. The median initial pain score was 7-8 for all patient categories. At 7 days, the odds ratio of atelectasis development were (score (0-3) 1.18 (0.42-3.34); score (4-7) 1.20 (0.48-3.03); p<=0.05). The Pearson correlation of cough capacity assessment, in patients without atelectasis were (0.53 nurse vs. patient; 0.37 physician vs. patient; 0.51 nurse vs. physician p<=0.05). As for the cough capacity perception correlation in patients with atelectasis were (0.62 nurse vs. patient; 0.40 physician vs. patient; 0.51 nurse vs. physician; p<=0.05). Conclusion: There is no statistically significant difference in atelectasis development depending on cough capacity and there is poor correlation regarding the perception of cough capacity except for the nurse. It would be interesting to develop a patient reported outcome measure questionnaire which targets minor thoracic trauma as it is a common emergency department complaint and it could help us improve medical management and patient quality of life

scholarly journals Evaluating the accuracy and utility of wearable technology to continuously monitor patients in a hospital setting: a systematic review (Preprint)

2019 ◽  
Author(s):  
Vikas Patel ◽  
Ani-Orchanian Cheff ◽  
Robert Wu
Keyword(s):  
Physical Activity ◽  
Systematic Review ◽  
Heart Rate ◽  
Hospital Setting ◽  
Wearable Technology ◽  
Wearable Devices ◽  
Hospitalized Patients ◽  
Poor Sleep ◽  
Inpatient Setting ◽  
Physical Activity Monitoring

BACKGROUND The term ‘post-hospital syndrome’ has been used to describe the condition in which elderly patients are transiently frail after hospitalization and have a high chance of readmission. Since low activity and poor sleep contribute to ‘post-hospital syndrome’, continuous inpatient monitoring of these important parameters using affordable wearables may help and reduce this syndrome. While there have been systematic reviews of wearables for physical activity monitoring in the hospital setting, there is limited data on use of wearables measuring other parameters in hospitalized patients. OBJECTIVE This systematic review aimed to evaluate the utility and accuracy of wearable devices in their ability to monitor inpatients. METHODS This review incorporated a comprehensive search of seven databases and included articles which met the following inclusion criteria: inpatients above age 18, device studied in the articles had to be wearable technology and have at least one sensor, articles had to describe an element of continuous monitoring (greater than 24 hours) and monitoring had to include more than just physical activity. There were no restrictions on publication period, but only English language studies were included. From each study we extracted basic demographic information along with characteristics of the intervention. RESULTS From 2,012 articles that were screened, 15 articles met the selection criteria. All articles included were observational in design. Nine different commercial wearables, with various body locations, were examined in this review. The devices collectively measured 7 different health parameters across all studies. Only 6 studies validated their results against a reference device or standard. Of those that did validate results, many found that certain variables were inaccurate with wide limits of agreement. Heart rate and sleep had the most evidence for being valid in the hospital. Overall, wearable devices were found to be a feasible alternative for inpatient monitoring as 13 of the 15 studies had a mean participation completion rate greater than 80%. CONCLUSIONS Overall, assessment of studies in this review suggested that wearable devices showed promise in monitoring the heart rate and sleep of patients in hospital. The results demonstrate that many devices have not been validated in the inpatient setting, and amongst those that do, some wearable measurements were not found to be valid. Further research is needed to validate the wearable health variables in hospitalized patients and eventually determine whether these devices improve health outcomes.

Relationship between Invasive and Non-Invasive Hemodynamic Measures in Experimental Pulmonary Hypertension

2020 ◽  
Vol 16 (1) ◽  
pp. 47-53
Author(s):  
Vicente Benavides-Córdoba ◽  
Mauricio Palacios Gómez
Keyword(s):  
Heart Rate ◽  
Pulmonary Hypertension ◽  
Ejection Fraction ◽  
Right Ventricle ◽  
Correlation Coefficient ◽  
Pearson Correlation ◽  
Systolic Pressure ◽  
Control Group ◽  
Pearson Correlation Coefficient ◽  
Non Invasive

Introduction: Animal models have been used to understand the pathophysiology of pulmonary hypertension, to describe the mechanisms of action and to evaluate promising active ingredients. The monocrotaline-induced pulmonary hypertension model is the most used animal model. In this model, invasive and non-invasive hemodynamic variables that resemble human measurements have been used. Aim: To define if non-invasive variables can predict hemodynamic measures in the monocrotaline-induced pulmonary hypertension model. Materials and Methods: Twenty 6-week old male Wistar rats weighing between 250-300g from the bioterium of the Universidad del Valle (Cali - Colombia) were used in order to establish that the relationships between invasive and non-invasive variables are sustained in different conditions (healthy, hypertrophy and treated). The animals were organized into three groups, a control group who was given 0.9% saline solution subcutaneously (sc), a group with pulmonary hypertension induced with a single subcutaneous dose of Monocrotaline 30 mg/kg, and a group with pulmonary hypertension with 30 mg/kg of monocrotaline treated with Sildenafil. Right ventricle ejection fraction, heart rate, right ventricle systolic pressure and the extent of hypertrophy were measured. The functional relation between any two variables was evaluated by the Pearson correlation coefficient. Results: It was found that all correlations were statistically significant (p <0.01). The strongest correlation was the inverse one between the RVEF and the Fulton index (r = -0.82). The Fulton index also had a strong correlation with the RVSP (r = 0.79). The Pearson correlation coefficient between the RVEF and the RVSP was -0.81, meaning that the higher the systolic pressure in the right ventricle, the lower the ejection fraction value. Heart rate was significantly correlated to the other three variables studied, although with relatively low correlation. Conclusion: The correlations obtained in this study indicate that the parameters evaluated in the research related to experimental pulmonary hypertension correlate adequately and that the measurements that are currently made are adequate and consistent with each other, that is, they have good predictive capacity.

scholarly journals Design and evaluation of a novel approach to invisible electrocardiography (ECG) in sanitary facilities using polymeric electrodes

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Aline dos Santos Silva ◽  
Hugo Almeida ◽  
Hugo Plácido da Silva ◽  
António Oliveira
Keyword(s):  
Heart Rate ◽  
Pearson Correlation ◽  
Voluntary Action ◽  
Signal Loss ◽  
Standard Equipment ◽  
Self Monitoring ◽  
Battery Lifetime ◽  
Ecg Signals ◽  
Toilet Seat ◽  
Novel Approach

AbstractMultiple wearable devices for cardiovascular self-monitoring have been proposed over the years, with growing evidence showing their effectiveness in the detection of pathologies that would otherwise be unnoticed through standard routine exams. In particular, Electrocardiography (ECG) has been an important tool for such purpose. However, wearables have known limitations, chief among which are the need for a voluntary action so that the ECG trace can be taken, battery lifetime, and abandonment. To effectively address these, novel solutions are needed, which has recently paved the way for “invisible” (aka “off-the-person”) sensing approaches. In this article we describe the design and experimental evaluation of a system for invisible ECG monitoring at home. For this purpose, a new sensor design was proposed, novel materials have been explored, and a proof-of-concept data collection system was created in the form of a toilet seat, enabling ECG measurements as an extension of the regular use of sanitary facilities, without requiring body-worn devices. In order to evaluate the proposed approach, measurements were performed using our system and a gold standard equipment, involving 10 healthy subjects. For the acquisition of the ECG signals on the toilet seat, polymeric electrodes with different textures were produced and tested. According to the results obtained, some of the textures did not allow the acquisition of signals in all users. However, a pyramidal texture showed the best results in relation to heart rate and ECG waveform morphology. For a texture that has shown 0% signal loss, the mean heart rate difference between the reference and experimental device was − 1.778 ± 4.654 Beats per minute (BPM); in terms of ECG waveform, the best cases present a Pearson correlation coefficient above 0.99.

scholarly journals 651 Longitudinal, unobtrusive, and ecologically valid sleep metric estimation from a smart bed to predict the pathology of COVID-19

SLEEP ◽  
2021 ◽  
Vol 44 (Supplement_2) ◽  
pp. A255-A255
Author(s):  
Dmytro Guzenko ◽  
Gary Garcia ◽  
Farzad Siyahjani ◽  
Kevin Monette ◽  
Susan DeFranco ◽  
...  
Keyword(s):  
Heart Rate ◽  
Sleep Quality ◽  
Predictive Model ◽  
Sleep Duration ◽  
Wearable Sensors ◽  
Test Results ◽  
Cloud Data ◽  
Cardiorespiratory Function ◽  
Group Data ◽  
Rich Information

Abstract Introduction Pathophysiologic responses to viral respiratory challenges such as SARS-CoV-2 may affect sleep duration, quality and concomitant cardiorespiratory function. Unobtrusive and ecologically valid methods to monitor longitudinal sleep metrics may therefore have practical value for surveillance and monitoring of infectious illnesses. We leveraged sleep metrics from Sleep Number 360 smart bed users to build a COVID-19 predictive model. Methods An IRB approved survey was presented to opting-in users from August to November 2020. COVID-19 test results were reported by 2003/6878 respondents (116 positive; 1887 negative). From the positive group, data from 82 responders (44.7±11.3 yrs.) who reported the date of symptom onset were used. From the negative group, data from 1519 responders (48.4±12.9 yrs.) who reported testing dates were used. Sleep duration, sleep quality, restful sleep duration, time to fall asleep, respiration rate, heart rate, and motion level were obtained from ballistocardiography signals stored in the cloud. Data from January to October 2020 were considered. The predictive model consists of two levels: 1) the daily probability of staying healthy calculated by logistic regression and 2) a continuous density Hidden Markov Model to refine the daily prediction considering the past decision history. Results With respect to their baseline, significant increases in sleep duration, average breathing rate, average heart rate and decrease in sleep quality were associated with symptom exacerbation in COVID-19 positive respondents. In COVID-19 negative respondents, no significant sleep or cardiorespiratory metrics were observed. Evaluation of the predictive model resulted in cross-validated area under the receiving-operator curve (AUC) estimate of 0.84±0.09 which is similar to values reported for wearable-sensors. Considering additional days to confirm prediction improved the AUC estimate to 0.93±0.05. Conclusion The results obtained on the smart bed user population suggest that unobtrusive sleep metrics may offer rich information to predict and track the development of symptoms in individuals infected with COVID-19. Support (if any):

Sign in / Sign up

Export Citation Format

Share Document