Remote Monitoring of High-Risk Patients With Chronic Cardiopulmonary Diseases

Overview

In this project, Institute of Bioengineering \& Bioimaging (IBB), A\*STAR would like to collaborate with Massachusetts General Hospital to aggregate patient data and to further develop its software algorithm using machine learning and statistical models for predicting exacerbations and deterioration on 60 patients with cardiopulmonary diseases.

Exacerbations of chronic cardiopulmonary diseases are a major cause of morbidity and mortality worldwide. There are an estimated 23 million patients with heart failure worldwide, and the prevalence of heart failure in the United Sates is projected to rise over the next four decades with an estimated 772,000 new heart failure cases projected in the year 2040. Exacerbations of chronic respiratory disease can accelerate lung function decline and reduce survival. They may also lead to significant rise to the cost of healthcare. Chronic Obstructive Pulmonary Disease (COPD) exacerbations are an important cause of readmissions with a 30-day readmission rate of approximately 20% and subsequent expenditure of US $15 Billion in annual health care spending. Cystic fibrosis (CF), a genetic disorder that affects airways clearance and secretions, has a 30-day readmission rate of approximately 11%. Due to the high cost of hospital stays and emergency department visits, and especially in the setting of the COVID-19 pandemic, more cost-effective "out-of-hospital" management models have become increasingly appealing. Such models not only provide cost benefits to patients, payers, and hospitals, but also increase the ability to provide care to people at home. Respiratory variables have shown to be one of the most sensitive indicators for COPD exacerbation, and a significant correlation between respiratory rate and COPD symptoms has been observed. When combined with pulse rate and oxygen saturation, these variables provide a useful method of identifying exacerbations. Current analytical models are designed to trigger alarms, which are generally based on traditional threshold-type driven analytics. Such methods are not able to identify and recognize trends due to limited access to advanced analytics (e.g., machine learning methods). The device proposed for use in this study will measure respiratory rate, Inspiratory: Expiratory (I:E) ratio, respiratory depth, heart rate, SpO2, SpO2 variability, patient movement, and the investigators will use machine learning and data modeling to analyze their trends over time. The wearable biometric platform (termed 'Respiratory Sensor') developed by Institute of Bioengineering \& Bioimaging (IBB), A\*STAR is to be worn on the chest area via a medical grade adhesive patch. The Respiratory Sensor includes a non-invasive sensor combining an accelerometer and light-based methods to sense chest wall expansion or breathing. The Respiratory Sensor allows the possibility of collecting additional respiratory information (respiratory rate, relative tidal depth and duty cycle) as predictors for exacerbation of chronic cardiopulmonary diseases and perhaps improving advanced analytical models that can provide better sensitivity and specificity compared to traditional models (e.g. clinical diaries). Ultimately, this may allow early prediction of outpatient exacerbations to allow early intervention and reduced re-admissions (via remote interventions). In this project, Institute of Bioengineering \& Bioimaging (IBB), A\*STAR would like to collaborate with Massachusetts General Hospital to aggregate patient data and to further develop its software algorithm using machine learning (e.g. random forest models and long-short term memory models etc.) and statistical models (e.g. regression models and survival analysis with univariate and multivariate analysis) based on respiratory features and hemodynamics for predicting exacerbations and deterioration on 60 patients with cardiopulmonary diseases. The end-points of this collaboration includes the following: 1. To validate hypothesis of using respiratory-based biomarkers in models (disease agnostic and disease specific) to predict exacerbations - benchmarking to be done versus follow-up questionnaires and phone calls 2. To validate level of compliance, drop-out rate and if additional measures are required to get patients to follow-on

Conditions

Interventions

Eligibility

Locations (1)

Outcomes

Central Contacts