End-tidal carbon dioxide monitoring during procedural sedation

To prospectively determine whether end-tidal carbon dioxide (ETCO2) monitors can detect respiratory depression (RD) and the level of sedation in emergency department (ED) patients undergoing procedural sedation (PS).

This was a prospective observational study conducted in an urban county hospital of adult patients undergoing PS. Patients were monitored for vital signs, depth of sedation per the physician by the Observer's Assessment of Alertness/Sedation scale (OAA/S), pulse oximetry, and nasal-sample ETCO2 during PS. Respiratory depression was defined as an oxygen saturation <90%, an ETCO2 >50 mm Hg, or an absent ETCO2 waveform at any time during the procedure. The physician also determined whether protective airway reflexes were lost during the procedure and assisted ventilation was required, or whether there were any other complications. Rates of RD were compared with the physician assessment of airway loss and between agents using chi-square statistics. Spearman's rho analysis was used to determine whether there was a correlation between ETCO2 and the OAA/S score.

Seventy-four patients were enrolled in the study. Forty (54.1%) received methohexital, 21 (28.4%) received propofol, ten (13.5%) received fentanyl and midazolam, and three (4.1%) received etomidate. Respiratory depression was seen in 33 (44.6%) patients, including 47.5% of patients receiving methohexital, 19% receiving propofol (p = 0.008), 80% receiving fentanyl and midazolam, and 66.6% receiving etomidate. No correlation between OAA/S and ETCO2 was detected. Eleven (14.9%) patients required assisted ventilation at some point during the procedure, all of whom met the criteria for RD. Pulse oximetry detected 11 of the 33 patients with RD. Post-hoc analysis revealed that all patients with RD had an ETCO2 >50 mm Hg, an absent waveform, or an absolute change from baseline in ETCO2 >10 mm Hg.

There was no correlation between ETCO2 and the OAA/S score. Using the criteria of an ETCO2 >50 mm Hg, an absolute change >10 mm Hg, or an absent waveform may detect subclinical RD not detected by pulse oximetry alone. The ETCO2 may add to the safety of PS by quickly detecting hypoventilation during PS in the ED.