The goal of this clinical trial is to learn whether artificial intelligence AI-enhanced remote monitoring (AIRM) is feasible, safe and acceptable for adults with fibrotic interstitial lung disease (fILD). The artificial intelligence (AI) system uses information collected through home monitoring to identify patterns that may suggest a person's lung condition is getting worse. The AI does not make treatment decisions. Its outputs are reviewed by healthcare professionals.
The main questions this study aims to answer are:
Is AIRM feasible and acceptable for people with fILD and their healthcare professionals? Can it be used safely to help identify possible worsening of fILD? How well does the AI system identify clinical deterioration?
Participants will take part for about 12 months. They will:
Measure their lung function and oxygen levels at home each week using connected monitoring devices.
Report their breathlessness and cough symptoms each week using the patientMpower app.
Complete questionnaires about their health and their experience of remote monitoring at the start of the study and every 3 months.
Continue to receive their usual clinical care.
Some participants will also be invited to an interview or focus group to discuss their experience of AI-enhanced remote monitoring. Healthcare professionals involved in the study may also be invited to discuss their experience of using the system.
The AI system is investigational and is being evaluated as part of this study. Healthcare professionals will review relevant AI-generated alerts before taking any clinical action. The AI will support, rather than replace, clinical judgement.
Aged 18 years or older.
Diagnosis of fibrotic interstitial lung disease (fILD) confirmed through multidisciplinary team review at a tertiary ILD centre, based on high-resolution computed tomography (HRCT) and standard clinical assessment.
Forced vital capacity (FVC) >45% predicted.
Able to perform spirometry and use a home spirometer safely and reliably.
Previous participation in the REMILD study.
Estimated life expectancy of less than 6 months, as judged by the treating clinician.
Contraindication to spirometry in accordance with Association for Respiratory Technology and Physiology (ARTP) guidelines.
Current or recent acute exacerbation of ILD, confirmed clinically and/or radiologically, within the previous 4 weeks.
Single group
AIRM v1.0 AI-Enhanced Remote Monitoring
DeviceAI-enhanced remote monitoring (AIRM) delivered through the patientMpower platform for 12 months. Participants will use connected home spirometry and pulse oximetry devices and report symptoms and patient-reported measures. AIRM version 1.0 is an investigational artificial intelligence (AI) algorithm that analyses longitudinal multimodal remote monitoring data to identify patterns associated with potential clinical deterioration in fibrotic interstitial lung disease (fILD). AIRM-generated outputs that may contribute to clinical management will be reviewed by an appropriately qualified clinician before any clinical action is taken. The algorithm will remain fixed during the prospective study and will not be retrained or modified during participant follow-up. AIRM supports, rather than replaces, clinical judgement and usual clinical care.
Recruitment Rate of Eligible Participants
Recruitment rate will be calculated as the number of eligible individuals who provide informed consent and enrol in the study divided by the total number of eligible individuals approached for participation, multiplied by 100. The outcome will be reported as a percentage.
Percentage of Enrolled Participants Retained at 12 Months
Retention rate will be calculated as the number of enrolled participants who remain enrolled in the study through to the end of the 12-month active remote monitoring period divided by the total number of participants enrolled, multiplied by 100. The outcome will be reported as a percentage.
Percentage of Weekly Home Spirometry Measurements Completed
Calculated as the number of scheduled weekly home spirometry measurements completed by participants divided by the total number of scheduled home spirometry measurements, multiplied by 100.
Percentage of Scheduled Pulse Oximetry Measurements Completed
Calculated as the number of scheduled weekly pulse oximetry measurements completed by participants divided by the total number of scheduled pulse oximetry measurements, multiplied by 100.
Percentage of AIRM Alerts That Are False-Positive Alerts
Calculated as the number of AIRM alerts not associated with a clinician-confirmed clinical deterioration event divided by the total number of AIRM alerts, multiplied by 100.
Sensitivity of AIRM for Identifying Clinically Significant Deterioration Events
Calculated as the number of clinician-confirmed clinical deterioration events identified by AIRM divided by the total number of clinician-confirmed clinical deterioration events, multiplied by 100.
Percentage of AIRM Alerts Leading to Clinical Review
Calculated as the number of AIRM-generated alerts that result in clinical review by a clinician divided by the total number of AIRM-generated alerts, multiplied by 100. The outcome will be reported as a percentage.
Percentage of AIRM Alerts Leading to Clinical Intervention
Calculated as the number of AIRM-generated alerts that result in clinical intervention divided by the total number of AIRM-generated alerts, multiplied by 100. The outcome will be reported as a percentage.
Specificity of AIRM for Identifying Clinically Significant Deterioration Events
The specificity of AIRM for identifying clinically significant deterioration will be assessed by comparing AIRM predictions with clinician-confirmed clinical deterioration. Specificity will be calculated as the proportion of events without clinically significant deterioration that are correctly identified by AIRM as not indicating deterioration and will be reported as a percentage.
Area Under the Receiver Operating Characteristic Curve (AUROC) for Artificial Intelligence enhanced Remote Monitoring Prediction of Clinical Deterioration
The discriminative performance of Artificial Intelligence enhanced Remote Monitoring for predicting clinician-confirmed clinical deterioration will be assessed using the area under the receiver operating characteristic curve (AUROC). AUROC values range from 0.5 to 1.0, with higher values indicating greater ability to discriminate between deterioration and non-deterioration.
Click on the lead sponsor to view all of their trials.
King's College London
Lead sponsor
patientMpower Ltd.
Collaborator
This trial is not recruiting at the moment. You can still explore other options:
