Edinburgh Napier University

  Introduction There is a need of innovative models of care for patients with severe COPD and frequent AECOPD, and Telehealth (TH) is part of these programs. But current systems are limited by the parameters feasibly monitored in a domestic setting and lack of a reliable method of predicting exacerbations. Evidence from hospital based studies show that breathlessness increases during exacerbations. If respiratory rate (RR) could be reliably monitored remotely it may provide a significant advance in predicting and identifying COPD exacerbations and monitoring recovery. The aim of this study is to validate five non-invasive RR monitors (M1 to M5) in patients with COPD in a laboratory setting against a gold standard measurement of RR.

Methods and results Five RR monitors identified in the literature were selected for validation against RR measured with a gold standard method (Oxycon mobile, Carefusion) in 23 patients with COPD (13 males, age 70 ± 8.3 years, FEV1 58.3 ± 17.1%pred) during a 52 min protocol of a total of 19 activities of daily living (i.e sitting, standing, walking at different speeds, climbing stairs, lifting objects and sweeping the floor). Patients wore simultaneously the five monitors and the Oxycon mobile and RR was recorded breath by breath and averaged by minute. One minute of each activity was selected for analysis using Bland and Altman plots. Bias and limit of agreement (LoA) was established for each monitor (Figure 1). Bias and LoA for the five monitors were the following (M1 2.15 (-17.9 to 22.2), M2 3.1 (-8.7 to 14.9), M3 2.2 (-12.12 to 16.6), M4 -2.5 (-11.7 to 6.8) and M5 -1.9 (-10.8 to 6.9)). Patients were compliant with the use of the five monitors.

Conclusions Monitoring RR is feasible and non-intrusive in patients with COPD. We have identified two monitors (M4 and M5) with the lowest bias and the narrower LoA. These monitors will be further investigated in a home setting.