Classification of Sleep Apnea Severity by Electrocardiogram Monitoring Using a Novel Wearable Device.

Sleep apnea (SA) is a prevalent disorder diagnosed by polysomnography (PSG) based on the number of apnea-hypopnea events per hour of sleep (apnea-hypopnea index, AHI). PSG is expensive and technically complex; therefore, its use is rather limited to the initial diagnostic phase and simpler devices are required for long-term follow-up. The validity of single-parameter wearable devices for the assessment of sleep apnea severity is still debated. In this context, a wearable electrocardiogram (ECG) acquisition system (ECG belt) was developed and its suitability for the classification of sleep apnea severity was investigated using heart rate variability analysis with or without data pre-filtering. Several classification algorithms were compared and support vector machine was preferred due to its simplicity and overall performance. Whole-night ECG signals from 241 patients with a suspicion of sleep apnea were recorded using both the ECG belt and patched ECG during PSG recordings. 65% of patients had an obstructive sleep apnea and the median AHI was 21 [IQR: 7-40] h - 1 . The classification accuracy obtained from the ECG belt (accuracy: 72%, sensitivity: 70%, specificity: 74%) was comparable to the patched ECG (accuracy: 74%, sensitivity: 88%, specificity: 61%). The highest classification accuracy was obtained for the discrimination between individuals with no or mild SA vs. moderate to severe SA. In conclusion, the ECG belt provided signals comparable to patched ECG and could be used for the assessment of sleep apnea severity, especially during follow-up.

Telehealthcare for Chronic Obstructive Pulmonary Disease in Switzerland Is Feasible and Appreciated by Patients.

BACKGROUND: Earlier detection of acute exacerbations (AE) of chronic obstructive pulmonary disease (COPD) could reduce emergency admissions and hospitalisations. Studies investigating COPD management programs supported by telehealthcare (THC) have shown conflicting results. OBJECTIVES: To test the feasibility, safety and acceptance of THC for COPD. METHODS: Patients daily filled out an online questionnaire focused on the detection of AECOPD. The THC platform is integrated in a comprehensive electronic patient data repository, which has to be available for all patients in Switzerland by law by 2017. The study team called the patient by phone in case of suspected AECOPD. RESULTS: Of 339 screened patients, 14% were included. Main reasons for exclusion were missing technical equipment and unwillingness to participate in a study (50%). Data completeness was 88%; 94% completed the study. The current THC approach triggered 230 telephone calls, which led to the verification of 60 AECOPD in 22 patients. Three AECOPD were not detected. Sensitivity, specificity, positive and negative predictive value of the questionnaire-based THC approach in detecting AECOPD was 95, 98, 26 and 99.9%, respectively. Overall patient satisfaction in respect to their health condition improved significantly (VAS 8-8.7; p = 0.002). CONCLUSIONS: Adding THC to state-of-the-art COPD management is feasible in a selected subgroup of patients. We estimate that up to 50% of COPD patients could be eligible for a THC strategy. Patient compliance, acceptance and satisfaction were very high. With the proposed approach, we missed only very few AECOPD events. However, a telephone-based verification of THC alerts was required. Overall, in this proof-of-concept study, we experienced a positive effort-to-benefit ratio.

Smartphone-based cough monitoring as a near real-time digital pneumonia biomarker.

Background: Cough represents a cardinal symptom of acute respiratory tract infections. Generally associated with disease activity, cough holds biomarker potential and might be harnessed for prognosis and personalised treatment decisions. Here, we tested the suitability of cough as a digital biomarker for disease activity in coronavirus disease 2019 (COVID-19) and other lower respiratory tract infections. Methods: We conducted a single-centre, exploratory, observational cohort study on automated cough detection in patients hospitalised for COVID-19 (n=32) and non-COVID-19 pneumonia (n=14) between April and November 2020 at the Cantonal Hospital St Gallen, Switzerland. Cough detection was achieved using smartphone-based audio recordings coupled to an ensemble of convolutional neural networks. Cough levels were correlated to established markers of inflammation and oxygenation. Measurements and main results: Cough frequency was highest upon hospital admission and declined steadily with recovery. There was a characteristic pattern of daily cough fluctuations, with little activity during the night and two coughing peaks during the day. Hourly cough counts were strongly correlated with clinical markers of disease activity and laboratory markers of inflammation, suggesting cough as a surrogate of disease in acute respiratory tract infections. No apparent differences in cough evolution were observed between COVID-19 and non-COVID-19 pneumonia. Conclusions: Automated, quantitative, smartphone-based detection of cough is feasible in hospitalised patients and correlates with disease activity in lower respiratory tract infections. Our approach allows for near real-time telemonitoring of individuals in aerosol isolation. Larger trials are warranted to decipher the use of cough as a digital biomarker for prognosis and tailored treatment in lower respiratory tract infections.

Longitudinal change of COPD assessment test (CAT) in a telehealthcare cohort is associated with exacerbation risk.

BACKGROUND: There are only scarce data regarding the evolution of the chronic obstructive pulmonary disease (COPD) assessment test (CAT) over time. Our aim was to investigate the evolution of the CAT in a telehealthcare (THC) cohort and to evaluate its potential to predict exacerbations. PATIENTS AND METHODS: The CAT was measured weekly over up to 1 year in 40 COPD patients undergoing a THC intervention. The evolution of the CAT was analyzed using linear regression. The association between this evolution and the occurrence of exacerbations was evaluated using the Andersen-Gill formulation of the Cox proportional hazards model for the analysis of recurrent time-to-event data with time-varying predictors. RESULTS: The median CAT at inclusion was 17 (interquartile range 13-22) points. During the study, 25% of patients had a significant negative slope (median -7 points per year [ppy]), 38% were stable (median +0 ppy) and 38% had a significant positive slope (median +6 ppy). The median slope of the CAT in the overall cohort was +1 (interquartile range -3 to +6) ppy. A significant positive association was found between the change in CAT scores and the risk of exacerbations (hazard ratio =1.08, 95% CI: 1.03-1.13; p<0.001). There was an 8% increase of the risk of exacerbation per unit increase in CAT. We detected a significant learning effect in filling out the CAT in 18.4% of patients with a median learning phase of five filled questionnaires. CONCLUSION: Sixty-three percent of the COPD patients monitored by THC experienced a stable or improved CAT during 1-year follow-up. We found a significant positive association between the evolution of the CAT over time and the risk of exacerbations. In about one-fifth of patients, there was a significant learning effect in filling out the CAT, before reliable results could be obtained. The evolution of the CAT could help to assess the risk for future exacerbations.