🏅 Best Practice / Service Development Poster Winner
Diagnosis of COPD using N-TidalTM, a novel hand-held lung function test. (ID 469)
TidalSense Limited
Funding: NIHR (i4i grant), Innovate UK, SBRI Healthcare and Pfizer OpenAir.
Abstract
AimSpirometry is the current gold standard for COPD diagnosis, but is technique-dependent, non-specific, potentially aerosol-generating, and requires administration by an ARTP-accredited professional. This results in significant under and misdiagnosis, particularly in primary care. There is therefore a need for a simple, safe, reliable and precise alternative COPD diagnostic test. This study attempted to accurately diagnose COPD by applying machine learning techniques to a 75-second relaxed breathing CO2 breath trace captured using TidalSense’s N-Tidal™ handheld capnometer.
Methods
A logistic regression model was trained on 82 features derived from capnograms of 234 COPD and 560 non-COPD participants from five clinical studies (CBRS, CBRS2, GBRS, ABRS and CARES). Performance was measured in a validation set of 60 COPD, and 145 non-COPD participants.
Results
The classification model yielded AUROC of 0.877, sensitivity 0.780, specificity 0.849 and positive predictive value (PPV) 0.835 (Figure 1). A likely clinical use for this model is to rule in or rule out a diagnosis in patients where the model is most confident. Therefore, performance was measured on the 49% of capnograms where the model was most confident (>80% and <20% probability of COPD respectively). This yielded a sensitivity of 0.89 and specificity 0.95 (Figure 2). Waveform features driving classification correlated with FEV1% predicted, supporting their hypothesised role as markers of airway obstruction in COPD.
Figure 1: ROC curve for COPD all severities classifier
Figure 2: Performances on full test set and very likely/unlikely COPD only
Conclusion
The N-TidalTM device and machine learning classifier could be used as an accurate, point-of-care diagnostic test for COPD, particularly in primary care where a rapid rule-out or rule-in test could replace spirometry. Further real-world primary care implementation studies assessing acceptability of the device to patients and clinicians is planned.
Conflicts of interest: LT, AS, MH, JCC, HB, RHL, GL, AXP are currently employed, or were employed/funded at the time of the research, by TidalSense Limited. GH is funded by the National Institute for Health Research (NIHR) Community Healthcare MedTech and In Vitro Diagnostics Co-operative at Oxford Health NHS Foundation Trust. All authors declare no competing interests. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.
The conference has been instigated and organised by PCRS. We are grateful to sponsors and exhibitors who have contributed funding towards this event in return for exhibition space. Neither sponsors or exhibitors have had any input into the agenda or the selection of speakers with the exception of any sponsored satellite symposia which are clearly indicated. View the full list of sponsors.