Interobserver variability in determination of anaerobic threshold by cardiopulmonary exercise testing

Abstract

Listen

Translate article

Anaerobic threshold (AT) is the point during exercise when anaerobic metabolism supplements aerobic with additional CO2 production. The slope of the plot of CO2 production against O2 consumption increases from this threshold. An AT of 11 mlO2.kg−1.min−1 [1] discriminates between higher- and lower-risk patients in non-cardiac surgery. Our CPX testing facility is available for pre-operative assessment and research. A metabolic cart and software system (Zan, nSpire Health Ltd, Hertford) are used to measure and report multiple variables during incremental exercise on a stationary bicycle. We audited inter-observer variability by re-measuring anonymised saved data of 30 studies. Nine assessors independently reported AT by visual inspection using the ‘V-slope method’ of AT determination whereby the inflection point on a plot of CO2 production against O2 consumption is selected using a digital cursor system. The point was confirmed on plots of ventilatory equivalents for CO2 and O2. Data were analysed using spss for Windows V16.0 (SPSS Inc, Chicago IL, USA). Univariate analysis of variance showed significant variability between assessors (p = 0.004). The greatest variation was between assessors F and G (mean (95% CI) 0.65 (0.24–1.05) mlO2.kg−1.min−1; p = 0.002). All nine assessors independently assigned all patients to the same risk category i.e. either high or standard risk on the basis of AT (Fig. 1). Anaerobic threshold (AT) in patients assessed by nine clinicians. The horizontal line represents an AT of 11 mlO2.kg−1.min−1. There was statistically significant inter-observer variability in the determination of AT by nine independent assessors but this was not clinically significant. Mr A. Collingwood, Respiratory Physiologist, Plymouth Hospitals NHS Trust and Dr H. Sanders, Department of Statistics, University of Plymouth assisted in the preparation of this abstract.