Rose GA, Davies RG, Davison GW, et al.

Br J Anaesth. 2018;120(6):1187-1194.

BACKGROUND: Cardiorespiratory fitness can inform patient care, although to what extent natural variation in CRF influences clinical practice remains to be established. We calculated natural variation for cardiopulmonary exercise test (CPET) metrics, which may have implications for fitness stratification.
METHODS: In a two-armed experiment, critical difference comprising analytical imprecision and biological variation was calculated for cardiorespiratory fitness and thus defined the magnitude of change required to claim a clinically meaningful change. This metric was retrospectively applied to 213 patients scheduled for colorectal surgery. These patients underwent CPET and the potential for misclassification of fitness was calculated. We created a model with boundaries inclusive of natural variation [critical difference applied to oxygen uptake at anaerobic threshold (V O2-AT): 11 ml O2 kg(-1) min(-1), peak oxygen uptake (V O2 peak): 16 ml O2 kg(-1) min(-1), and ventilatory equivalent for carbon dioxide at AT (VE/VCO2-AT): 36].
RESULTS: The critical difference for V O2-AT, V O2 peak, and V E/V CO2-AT was 19%, 13%, and 10%, respectively, resulting in false negative and false positive rates of up to 28% and 32% for unfit patients. Our model identified boundaries for unfit and fit patients: AT <9.2 and >/=13.6 ml O2 kg(-1) min(-1), V O2 peak <14.2 and >/=18.3 ml kg(-1) min(-1), V E/V CO2-AT >/=40.1 and <32.7, between which an area of indeterminate-fitness was established. With natural variation considered, up to 60% of patients presented with indeterminate-fitness.
CONCLUSIONS: These findings support a reappraisal of current clinical interpretation of cardiorespiratory fitness highlighting the potential for incorrect fitness stratification when natural variation is not accounted for.