The diagnosis of either asthma or chronic obstructive pulmonary disease (COPD) is currently based on a combination of symptoms, different lung tests and sometimes a 'trial by treatment'. Both COPD and Asthma tests currently include forced breathing using a test known as spirometry which can be difficult and uncomfortable for people to perform and needs expert interpretation. Asthma can require multiple tests in sequence which can make the process of diagnosis long and inconvenient for patients. In the UK, there have been challenges providing enough testing for COPD and asthma, in part because the tests are challenging to provide in the community to everyone who needs one.
The STARDUST study aims to test and develop new ways of diagnosing asthma and COPD that are quick, accurate and easy for patient and healthcare professionals to perform. People who have been referred for routine lung testing will breathe normally in and out into a handheld device called the N-Tidal Handset for 75 seconds. This device measures how the level of carbon dioxide (CO2) changes in the breath as people breath through it. The information gathered on the N-Tidal Handsets, called breath records or capnograms, will not be used to alter the diagnosis of any participants in the study.
After the capnograms have been collected, the research team will test 'algorithms' that have been developed using artificial intelligence, to see if they can accurately identify the correct diagnosis. Information collected in the study will also be used to make improvements in these algorithms.
If confirmed to be accurate, these algorithms could be used in clinical practice to help healthcare professionals make faster, more accessible and accurate diagnoses, especially in settings like GP clinics in the community where access to specialist tests may be limited.
Age ≥ 18 years
Referred for respiratory diagnostic testing with a clinical suspicion of asthma or COPD
Participant is willing and able to provide informed consent
Referred for respiratory diagnostic testing for a condition other than asthma or COPD, with no mention in the referral of a clinical suspicion of COPD or asthma
Participants who, in the opinion of the chief investigator, or their delegate, are unlikely to comply with the requirements of the study
Inability to give written informed consent
Participants who are acutely unwell
Participants with a suspicion of asthma and/or COPD
Cohort
Prospective
Partial pressure of CO2 during tidal breathing
Diagnostic testPartial pressure of CO2 during tidal breathing for diagnosis of asthma or COPD
FENO
Diagnostic testFractional exhaled nitric oxide
Spirometry
Diagnostic testSpirometry with or without bronchodilation (if indicated)
Diagnostic accuracy of the diagnostic algorithms of N-Tidal Diagnose 1
PPV, NPV, Sensitivity, Specificity compared to clinician diagnosis of COPD
Diagnostic accuracy of prototype diagnostic algorithms
Prototype asthma diagnostic algorithms (index test; PPV, NPV, Sensitivity, Specificity) trained using a combined capnogram dataset from this and previous studies, compared to clinician diagnosis of asthma
Diagnostic accuracy of the N-Tidal Diagnose 1 severity algorithm
Diagnostic accuracy (PPV, NPV, Sensitivity, Specificity) of the N-Tidal Diagnose 1 severity algorithm in those with a confirmed diagnosis of COPD. Diagnostic accuracy (PPV, NPV, Sensitivity, Specificity) of the diagnostic algorithms embedded in NTidal Diagnose 1 v1.0 in relevant subgroups. Diagnostic accuracy (PPV, NPV, Sensitivity, Specificity) of the pre-production asthma diagnostic algorithms trained using a combined capnogram dataset from this and previous studies in relevant prespecified subgroups. Exploratory analysis of novel machine learning diagnostic algorithms for COPD and asthma.
Feasibility and acceptability outcomes
Visual Analogue Score (VAS) following N-Tidal, spirometry and FeNO (ease of use and comfort for participants, ease of use only for administering healthcare professionals) in a subset of participants. Percentage of participants with a clinical suspicion of COPD with adequate N-Tidal Handset breath record capture, sufficient to generate a diagnostic output using N-Tidal Diagnose 1 v1.0 diagnostic algorithms. Percentage of participants with a clinical suspicion of asthma with adequate N-Tidal Handset breath record capture, sufficient to generate a diagnostic output using prototype asthma diagnostic algorithms.
Click on the lead sponsor to view all of their trials.
Liverpool Heart and Chest Hospital NHS Foundation Trust
Lead sponsor
TidalSense
Collaborator
This trial is not recruiting at the moment. You can still explore other options:
