A MODEL-BASED EVALUATION OF THE SINGLE-BREATH CO2 VENTILATORY RESPONSE TEST

The accuracy of the single-breath CO2 inhalation test as a method for determining peripheral chemoreflex gain (G(p)) is evaluated through computer simulations using a mathematical model of the closed-loop respiratory control system. Estimates of G(p) (G(p)') are based on 'corrected' changes in end-tidal PCO2, because the uncorrected end-tidal values do not accurately reflect changes in alveolar PCO2. The influence of the central chemoreflex on G(p)' is generally <10% but can become disproportionally more significant as the relative contribution of the peripheral chemoreflex diminishes. G(p)' tends to overestimate G(p) with the inclusion of peripheral chemoreceptor rate sensitivity, but this effect is offset by the time constant for adaptation. The spontaneous variability of breathing can seriously impair the resolution of G(p). Averaging of G(p)' deduced from individual single-breath tests can lead to erroneous estimates of G(p) even when a large number of repetitions are performed. This problem can be minimized by first ensemble averaging the data from individual single-breath tests and, then, computing G(p)' from the resulting mean changes.