Performance Failure of a Colorimetric End-Tidal CO2 Detector

In the course of obtaining an emergency airway a colorimetric end-tidal CO2 detector erroneously failed to display a color change despite placement within the trachea.

The case involved a morbidly obese 47-year-old female with a history of bronchiectasis, recurrent hemoptysis, schizophrenia, and remote burns. Intubation was required in the emergency department (ED) because of recurrent, large-volume hemoptysis, coupled with episodes of severe desaturation (<80%). The therapeutic plan involved admission to the intensive care unit and emergent bronchoscopy.

The patient was sedated on ED arrival for severe agitation using 5 milligrams of intravenous haloperidol. Vital signs prior to intubation included a pulse rate of 104, a pulse oximetry reading of 95% on a non-rebreathing mask, blood pressure of 118/72, and a respiratory rate of 12. For intubation, the patient was given 20 mg of etomidate and 20 mg of vecuronium in rapid sequence fashion. With onset of induction the patient was positioned into an ear-to-sternal notch position, and gently ventilated using an oral airway and face mask, with low volume and low pressure.1 Pulse oximetry rose to 100% prior to laryngoscopy and during the 90-second delay between drug delivery and laryngoscope insertion. Using a Macintosh #4 blade the initial percentage of glottic opening score was 5% but after bimanual laryngoscopy the laryngoscopist reported 75% of the glottic opening was visible.2 An 8.0 mm ID tracheal tube was inserted into the trachea under direct vision. The stylet was shaped in a straight-to-cuff 35 degree fashion and the tip of the tube was visualized coming over the posterior cartilages.3 Depth of insertion was 21 cm at the teeth. Upon removal of the stylet and cuff inflation breath sounds were ascultated in both axillae, and absent over the stomach. Pulse oximetry remained at 100% with a regular waveform. A Vital Signs (Totowa, NJ) C-CO2 end-tidal colorimetric indicator was opened by the respiratory therapist, the pull-tab removed from the device, and connected between the tube and the resuscitation bag. No color change occurred despite continuous 100% saturation and well-heard breath sounds as noted. A second colorimetric indicator was then used and color change occurred. Post-intubation chest-xray and CT scanning confirmed intra-tracheal location and proper depth.

An examination of the first colorimetric indicator showed that a small piece of the plastic pull-tab remained hidden within the device covering the colorimetric indicator. Further examination of the pull-tab on other units revealed that the pull-tab is very easily torn along the straight edge of the unit where the tab exits. Furthermore, although the device packaging and insert warns that failure to fully remove the pull-tab will cause the device to fail, the pull-tab has no markings that show whether the tab has been fully removed. Upon opening the package, 18 millimeters of the pull-tab is visible (Figure). The full length of the pull-tab is 58 millimeters. This long length is counterintuitive given the 28 millimeter size of the overall unit, but the tab is actually folded over itself within the unit (at 41 mm). Based on the size and location of the aperture over the colorimetric window, we believe that the device may fail to work unless at least 50 millimeters of the tab is removed.

View Large Image Download to PowerPoint

Figure. Vital Signs C-CO2 Colorimetric Indicator, showing pull-tab (white) in place when initially opened, and full length of pull-tab after complete removal. Length of tab protruding from unit is 18 mm; fold is at 41 mm (marked); full length of tab is 58 mm. The point at which tab covers colorimetric window is 50 mm (marked).

Although we have no other clear-cut cases of device malfunction, we believe there have been several possible cases in which this may have occurred. We are in the midst of year-long study of laryngoscopy with and without the use of cricoid pressure in emergency airways. Several cases in our study have been documented in which the colorimetric result did not match the operator's perception of tube placement within the trachea. Tracheal tubes were then removed because of the end-tidal detector reading, and on several occasions these cases then involved repeat laryngoscopy efforts and prolonged mask ventilation episodes. In emergency airways repetitive laryngoscopy correlates with a marked increase in the risks of severe hypoxemia, regurgitation, and cardiac arrest.4

We encourage all clinicians to slowly and carefully remove the pull-tab when using this device, and to inspect it to be certain the entire plastic tab has been removed before relying on its performance. The problem we've presented in this report could be easily corrected with a plastic tab that cannot be torn, and by printed markings on the distal tab that clearly shows the tab has been entirely removed. We have contacted the manufacturer about this malfunction and in the interim are switching out our colorimetric devices used in our ED.