To the Editor:—

Tachibana et al.  1carefully designed their study to explore two issues that have not received much attention in the literature: performance of the NICO2monitor (Respironics-Novametrix, Wallingford, CT) during low tidal volume controlled ventilation and during spontaneous/mixed ventilation. For low minute volume controlled ventilation, the authors report that the NICO2monitor underestimates thermodilution cardiac output and shows decreased precision and accuracy. It is our opinion that the problem was not with the low minute volume. Instead, there are indications that the time allowed for the patients to stabilize following the 50% decrease in minute volume may not have been sufficient.

Taskar et al.  2showed that the time constant for carbon dioxide output (and end-tidal carbon dioxide) is much longer following a decrease in ventilation than an increase in ventilation. The time constant for a decrease in ventilation was 35 ± 10.7 min after a 10% decrease in ventilation. In other reports, only half of the final change in end-tidal carbon dioxide was realized 16 min after a decrease in ventilation. 3Changes in end-tidal carbon dioxide and carbon dioxide output are slow because of the time needed for the venous carbon dioxide partial pressure to adjust to the new level needed to allow carbon dioxide excretion to match metabolic carbon dioxide production.

On the basis of these data, it is likely that the venous carbon dioxide levels were still increasing at 15 min after the large decrease in minute volume described by Tachibana et al.  1This is supported by the authors’ report that average carbon dioxide output during the low minute volume mode was still 28% less than in the normal minute volume tests. If we can assume that the patients’ metabolic rates did not change with the ventilator adjustment, then the venous carbon dioxide level was not yet stable at the time of measurement. Tachibana et al.  noted that the NICO2monitor values appeared to be stable during the hypoventilation stage. This does not necessarily imply that the venous carbon dioxide levels were stable, only that they were increasing more slowly. Increasing venous carbon dioxide levels would be expected to cause partial rebreathing cardiac output to read low. This effect would be exacerbated at high cardiac outputs, as the differential end-tidal carbon dioxide signal is smaller.

Also, precision reported for the low minute ventilation test improves considerably (1.27–1.07 l/min), when the single outlier shown in the scatter plot (fig. 1C of the article) is rejected as an artifact (reducing n from 25 to 24). Moreover, not forcing the regression to pass through the origin improves the correlation significantly, from r = 0.34 to 0.72.

It is our opinion that had the patient remained in hypoventilation until the carbon dioxide output reached its preadjustment level, then the NICO2monitor readings would have been as accurate as those observed under normal ventilation. The NICO2monitor has been designed to calculate cardiac output after changes in ventilation that are typical in clinical practice. The severe decrease in ventilation tested by Tachibana et al.  1was outside the typical expected range of change.

Tachibana K, Imanaka H, Takeuchi M, Takauchi Y, Miyano H, Nishimura M: Noninvasive cardiac output measurement using partial carbon dioxide rebreathing is less accurate at settings of reduced minute ventilation and when spontaneous breathing is present. A nesthesiology 2003; 98: 830–7
Taskar V, John J, Larsson A, Wetterberg T, Jonson B: Dynamics of carbon dioxide elimination following ventilator resetting. Chest 1995; 108: 196–202
Lumb AB: Nunn's Applied Respiratory Physiology, 5th edition. Boston, Butterworth Heinemann, 2000, p 239