In stable, optimized COPD patients who have already completed PR, telemonitoring in addition to best care, reduces primary care chest contacts but not hospital or specialist team utilization.

Exhaled volatile organic compounds (VOCs) have shown promise in diagnosing chronic obstructive pulmonary disease (COPD) but studies have been limited by small sample size and potential confounders. An investigation was conducted in order to establish whether combinations of VOCs could identify COPD patients from age and BMI matched controls. Breath samples were collected from 119 stable COPD patients and 63 healthy controls. The samples were collected with a portable apparatus, and then assayed by gas chromatography and mass spectroscopy. Machine learning approaches were applied to the data and the automatically generated models were assessed using classification accuracy and receiver operating characteristic (ROC) curves. Cross-validation of the combinations correctly predicted the diagnosis in 79% of COPD patients and 64% of controls and an optimum area under the ROC curve of 0.82 was obtained. Comparison of current and ex smokers within the COPD group showed that smoking status was likely to affect the classification; with correct prediction of smoking status in 85% of COPD subjects. When current smokers were omitted from the analysis, prediction of COPD was similar at 78% but correct prediction of controls was increased to 74%. Applying different analytical methods to the largest group of subjects so far, suggests VOC analysis holds promise for diagnosing COPD but smoking status needs to be balanced.

Exhaled volatile organic compounds (VOCs) are of interest for their potential to diagnose disease non-invasively. However, most breath VOC studies have analyzed single breath samples from an individual and assumed them to be wholly consistent representative of the person. This provided the motivation for an investigation of the variability of breath profiles when three breath samples are taken over a short time period (two minute intervals between samples) for 118 stable patients with Chronic Obstructive Pulmonary Disease (COPD) and 63 healthy controls and analyzed by gas chromatography and mass spectroscopy (GC/MS). The extent of the variation in VOC levels differed between COPD and healthy subjects and the patterns of variation differed for isoprene versus the bulk of other VOCs. In addition, machine learning approaches were applied to the breath data to establish whether these samples differed in their ability to discriminate COPD from healthy states and whether aggregation of multiple samples, into single data sets, could offer improved discrimination. The three breath samples gave similar classification accuracy to one another when evaluated separately (66.5% to 68.3% subjects classified correctly depending on the breath repetition used). Combining multiple breath samples into single data sets gave better discrimination (73.4% subjects classified correctly). Although accuracy is not sufficient for COPD diagnosis in a clinical setting, enhanced sampling and analysis may improve accuracy further. Variability in samples, and short-term effects of practice or exertion, need to be considered in any breath testing program to improve reliability and optimize discrimination.