Validity of transcutaneous PCO2 in monitoring chronic hypoventilation treated with non-invasive ventilation

Abstract

BackgroundNon-invasive ventilation (NIV) is an efficient treatment for patients with chronic hypercapnic respiratory failure (CRF), but requires regular monitoring to detect both diurnal and nocturnal residual hypercapnia.The present study was designed to determine 1) whether transcutaneous PCO2 (PtcCO2) is a valid tool for monitoring PaCO2 in this group of patients, and 2) if overnight instrumental drift of the PtcCO2 sensor is clinically significant. MethodsSixty-seven patients with CRF on long term NIV were included. Arterial blood gases (ABG) were sampled from the radial artery during PtcCO2 measurement. PtcCO2 was recorded 2 min after ABG sampling. Instrumental drift was tested by measuring a gas of known CO2 concentration after auto-calibration of the sensor in the evening, and on the following morning. FindingsPaCO2 values ranged from 3·97 kPa to 9.0 kPa. Thirty-six (53%) patients were hypercapnic. Correlation between PaCO2 and PtcCO2 was highly significant (r2 = 0.9, p < 0.0001), Bias (d) and SD of bias (s) were 0.23 kPa and 0.28 kPa respectively, with a minor underestimation of PaCO2. Limits of agreement (d ± 2s) were; −0.32; 0.79 kPa. None of the paired values of PaCO2/PtcCO2 had a difference exceeding 1 kPa. The mean drift of PtcCO2 was 0.14 ± 0.54 kPa/8 h (p = 0.04; 95% CI: 0.01–0.27). InterpretationWith the device tested, in stable patients under NIV-treatment for CRF, PtcCO2 accurately reflects PaCO2. PtcCO2 can be used to monitor CO2 overnight during NIV without any clinically significant drift. Trial registration N°NCT01845233.