To the Editor:-Two multi-center studies evaluate the use of remifentanil in patients breathing spontaneously during monitored anesthesia care and in the management of acute postoperative care. [1,2]Neither addresses the real potential for respiratory depression in a sound manner. The conclusions by Gold et al. and Yarmush et al. that patients receiving remifentanil maintained adequate respiratory function during their studies cannot be supported with the reported data.
Adequacy of ventilation was assessed by monitoring respiratory rate and oxygen saturation as recorded by pulse oximetry. A respiratory rate of < 8 breaths/min for >or= to 1 min, or O2saturation < 94% on oxygen supplementation were the thresholds for defining respiratory depression, yet both of these values correlate poorly with ventilatory depression. [3–5]Arterial blood oxygen saturations have been shown to be well maintained at the time of peak respiratory depression in patients receiving supplemental oxygen. Under these circumstances, oxygen desaturation may be a late sign of respiratory depression. This failure to detect hypoventilation until oxygen desaturation occurs is an example of the false sense of security seen with pulse oximetry. Normal readings of oxygen saturation in the presence of increased inspired oxygen have been shown to give no information about the adequacy of ventilation. [7,8]Supplemental oxygen may mask carbon dioxide retention. The oxygen supplementation in these two studies was also not defined. The only noninvasive parameters that have been shown to correlate well with depressed ventilation are respiratory pattern, (i.e., paradoxical respiration) end-tidal CO2, or, if the depression is severe, level of consciousness. Sedation is a well-known accompaniment of hypercapnia with somnolence and unconsciousness occurring when PaCO2levels reach 80 mmHg.
Variations in respiratory pattern may be subtle and difficult to assess. End-tidal CO2measurements in the extubated patient monitored via nasal cannulae, in our experience, do not always correlate well with arterial PaCO2. When the patient is obtunded, nasal breathing is often reduced, and low recordings of end-tidal CO2are obtained. On the other hand, sedation level correlates with severe respiratory depression.
Although we have found remifentanil to be effective in the management of surgical pain and although it is being used more frequently in our clinical practice, careful assessment of the adequacy of respiration is required when using potent narcotics in spontaneously breathing extubated patients. Respiratory rate and pulse oximetry with oxygen supplementation do not meet this requirement. Both the cited studies should have included more precise measurements of ventilatory depression to be able to draw the conclusion that significant depression did not occur. A sedation scale, end-tidal CO2, respiratory pattern analysis, and, especially, arterial blood gas analysis should have been considered in the study design if the safety of the technique was the goal of the study. The emphasis on pulse oximetry reinforces the false sense of security of this monitoring technique when supplementary oxygen is being administered.
Michael A. E. Ramsay, M.D.
Amy Macaluso, M.D.
H. A. Tillmann Hein, M.D.
Eric Cancemi, M.D.
Department of Anesthesia; Baylor University Medical Center; 3500 Gaston Avenue; Dallas, Texas 75246
(Accepted for publication December 2, 1997.)