GENERAL CLINICAL INVESTIGATION/ORIGINAL CONTRIBUTION

I P Esophageal Detector Device Versus

Detection of End-Tidal Carbon Dioxide Level

in Emergency Intubation

From the Department of Emergency Medicine, The Johns Hopkins University School of Medicine, Baltimore, Maryland.

Received for publication April 3, 1995. Revision received October 31, 1995. Accepted for publication December 6, 1995.

The Chesapeake Medical Corporation provided one end-tidal CO 2 detector.

Presented in part at the Annual Meeting of the Society for Academic Emergency Medicine, Washington DC, May 1994.

Copyright © by the American College of Emergency Physicians.

William P Bozeman, MD David Hexter, MD, MPH Horace K Liang, MD Gabor D Kelen, MD

Study objectives: To confirm the ability of the esophageal detector device (EBB) to indicate positioning of endotracheal tubes (ETTs) in patients intubated under emergency conditions and to compare the performance of the EDD with that of end- tidal carbon dioxide (ETc02).

Methods: This single-subject study comprising a prospective case series was conducted in the emergency department of an urban university hospital. All adult patients were intubated either in the ED or by paramedics in the field. ETT position was initially evaluated by means of auscultation, then EDD, and, finally, spectrographic qualitative ETco 2 monitoring in each patient. Discrepancies between the EDD and ETco 2 results were resolved by means of direct laryngoscopy.

Results: In 100 intubated patients, both the EDD and ETCo 2 monitoring detected the single esophageal intubation that occurred. Of the remaining 99 tracheal intubations, the EDD correctly indicated tracheal placement in 98 (sensitivity, 99%) and was indeterminate in 1 case because of blockage of the ETT by secretions resulting from pulmonary edema. By comparison, ETco 2 monitoring correctly indicated tracheal placement in 86 cases (sensitivity, 87%) and was incorrect in 13 cases (P<.01). ETco 2 monitoring failed in 2 patients with pulmonary edema and in 11 patients with cardiac arrest. Among the 37 patients in the cardiac arrest group, the EDD correctly indicated ETT placement in 37 patients (sensitivity, 100%). In contrast, ETco 2 monitoring correctly indicated ETT placement in 26 patients (sensitivity, 70%; P<.01 ).

Conclusion: The EDD reliably confirms tracheal intubation in the emergency patient population. The EDD is more accurate than ETcO 2 monitoring in the overall emergency patient popula- tion because of its greater accuracy in cardiac arrest patients.

[Bozeman WP, Hexter D, Liang HK, Kelen GD: Esophageal detec- tor device versus detection of end-tidal carbon dioxide level in emergency intubation. Ann Emerg Med May 1996;27:595-599.]

MAY 1996 27:5 ANNALS OF EMERGENCY MEDICINE 5 9 5

EDD VERSUS ETco z Bozeman et al

INTRODUCTION

Correct placement of an endotracheal tube (ETT) is of paramount importance because unrecognized esophageal intubation rapidly produces iatrogenic mor- bidity and mortality. Current methods of confirming ETT placement include clinical examination, pulse oximetry, and end-tidal carbon dioxide (ETc02) moni- toring. All of these methods have limitations.

The esophageal detector device (EDD) was first described in 1988.1,2 This simple, inexpensive device confirms endotracheal intubation by aspiration of air from a correctly placed ETT. Aspiration of gas or fluid from a misplaced esophageal tube is prevented by col- lapse of the walls of the esophagus around the tip of the ETT. The EDD has been shown to be 98% to 100% accurate 1-6 and to have 100% agreement with ETCO 2 monitoring in the operating suite setting. 6

Several cases of equivocal results have been reported with the EDD. These cases may have involved blockage of an ETT, which prevents aspiration of air from the tra- chea.l,5,6 The incidence of thick secretions, aspirated materials, and foreign bodies that may block ETTs is potentially much higher in emergency patients than in elective surgical patients, in whom the EDD has been tested. We hypothesized that an unacceptably high num- ber of false-negative results caused by ETT blockage may occur in patients requiring emergency intubation. We therefore undertook prospective evaluation of the ability of the EDD to confirm tracheal intubation in the emer- gency population. The results were compared with spec- trographic qualitative ETco 2 detection, an objective method of confirming tracheal intubation.

M A T E R I A L S A N D METHODS

The EDD was constructed in the fashion described by Wee3 A 60-mL catheter-tip syringe was connected to a 15- mm internal diameter adapter (catalog no. 48312; Hudson Oxygen Company) with a length of semirigid tubing. In our version, a 6-mm uncuffed ETT was used. The tube was cut approximately 5 cm from the end to eliminate the "Murphy's eye" side hole (Figure).

Objective confirmation of ETT placement was obtained with a spectrographic qualitative ETco 2 monitor (Mini Cap III, catalog number 74571; MSA Catalyst Research Company). The CO 2 detection threshold was .5%. The ETCO 2 monitor was checked at least weekly during the study period for proper function and battery freshness.

The study protocol specified endotracheal intubation by the oral or nasal route, after which tube placement

was confirmed by means of auscultation of lung fields and the epigastrium while the patient was hyperventilat- ing with 100% oxygen. For clinical reasons, any patient known to have an esophageal intubation was reintubated before entering the study phase. After correct placement of the ETT was confirmed and the ETT was secured, the EDD was tested, after which ETCO 2 detection was per- formed. Discrepancies between the EDD and ETco 2 results without apparent cause were resolved by means of a second laryngoscopy to directly visualize ETT placement.

The EDD test consisted of removing the ventilation device and attaching the EDD. The plunger was gently withdrawn past the 40-mL mark on the syringe and released. Any rebound motion of the plunger caused by aspiration of a vacuum was allowed to occur. After the final position of the plunger was noted, the EDD was removed and the ventilating device reattached. The EDD results were interpreted to indicate tracheal ETT place- ment when an easy aspiration of at least 40 mL of air was possible. Results were interpreted to indicate esophageal ETT placement when tactile aspiration of a vacuum was possible and the EDD plunger rebounded after release to 10 mL or less. Results were interpreted as indeterminate when aspiration of an intermittent vacuum was possible or a total return of 11 mL to 39 mL of air was obtained.

The ETco 2 monitor had two indicators. An increase in the CO 2 level with each ventilation cycle produced an audible tone and a flashing green light on the monitor. This was interpreted to indicate tracheal ETT placement. Lack of any increase in the ETCO 2 level after 30 seconds produced an audible alarm and flashing red lights on the monitor. This was interpreted to indicate esophageal ETT placement.

All adult patients (18 years or older) who required intubation for any reason in the ED or who had been

Figure. The EDD.

596 ANNALS OF EMERGENCY MEDICINE 27:5 MAY 1996

EDD VERSUS ETco 2 Bozeman et al

intubated in the field by paramedics immediately before reaching the ED were eligible for study inclusion.

The EDD testing was performed by resident and attend- ing physicians in the ED of an urban university hospital. This ED handles approximately 50,000 visits per year. Training in the study protocol and use of the EDD and ETCO 2 monitor were completed by all operators before they performed data collection. Direct observation and a written test confirmed retention of the material by the operators and their ability to use the equipment correctly. Study inclusion required the availability of a qualified operator to perform EDD and ETCO 2 testing. Any devia- tion from the study protocol resulted in exclusion of the case from analysis. Data were collected from July 1993 to February 1994. Approval was obtained from the institu- tional review board with a waiver of informed consent. Analysis of the data included standard calculations of sen- sitivity and McNemar's test for paired samples to calculate ){2 statistics and P values. Because indeterminate EDD results are clinically undesirable and a corresponding indeterminate outcome with ETco 2 monitoring is not possible, the indeterminate EDD result was grouped with incorrect results for the purpose of statistical analysis.

RESULTS

The EDD was tested in 113 emergency intubations. Thirteen cases were excluded because of failure to follow the study protocol. Of the 100 intubations analyzed, 79 were performed in the ED and 21 were performed in the prehospital setting. Indications for intubation included cardiac arrest (n=37), trauma (n=18), altered mental sta- tus (n=lS), pulmonary edema (n=8), asthma (n=7), sepsis (n=4), and other conditions (n=l 1).

The mean age was 50.1 years (range, 18 to 94 years). Sixty percent of the patients were men. Twelve percent of the intubations were performed by the nasal route and 88% by the oral route. Most of the 100 intubations were performed by emergency medicine residents (n=60); the remaining intubations were performed by paramedics in the prehospital setting (n=21), anesthesiology residents (n= 16), and others, including internal medicine residents and paramedics training in the ED (n=3).

One of the 100 intubations in the study was an esophageal ETT placement and was not detected on clini- cal examination. Both the EDD and the ETCO 2 detector identified this esophageal ETT placement. This case was excluded from further analysis.

Of the remaining 99 endotracheal intubations, the EDD correctly identified 98 (99%) and was indeterminate

in 1 patient with pulmonary edema. Blockage of the ETT was evident in this patient. The ETCO 2 device correctly identified 86 endotracheal intubations (87%). Of the 13 patients in whom the ETCO 2 detector was incorrect, 2 had pulmonary edema and 11 had cardiac arrest. Overall sen- sitivity of the EDD was 99%, whereas overall sensitivity of ETCO 2 monitoring was 87%, a statistically significant dif- ference (P<. 01) (Table).

In the cardiac arrest group (n=37), the EDD correctly indicated the ETT position in 37 patients (100%), whereas ETCO 2 monitoring was correct in 26 patients (70%, P<.01) (Table). Excluding the cardiac arrest group, no statistically significant difference was observed in the sensitivities of the EDD and the ETCO 2 device. No adverse effects or reactions were noted with use of the EDD.

DISCUSSION

The ability of the EDD to detect esophageal intubations is well established. 1-6 Several equivocal and indeterminate EDD test results, however, have been reported. Intermittent or partial vacuums, attributed to thick secretions blocking the ETT or EDD, were noted on aspiration in these cases, s,5 Studies of an EDD model using a self-filling rub- ber bulb rather than a syringe to aspirate air have indicated several trials with "delayed EDD bulb re-inflation. ''3-6 This may be similar to the vacuums created by blockages that result from secretions.

Our study evaluated the ability of the EDD to confirm tracheal intubation in a group of intubated patients at high theoretical risk of ETT or EDD blockage--namely, emergency patients. In fact, our data indicate that the incidence of false-negative EDD results is significantly

Table. Comparison of the EDD and ETco 2 monitoring: Ab~lRy to confirm endotracheal ~ntubation.

95% Sensitivity Confidence

Type of Intubation Correct (%) Interval P*

All emergency endotracheal intubations (n=99)

ED0 98 99 97-100 <.01 ETco 2 86 87 80-94

Cardiac arrest endotracheal intubations (n=37)

EDD 37 100 - - <.01 ETco z 26 70 68-85

*McNemar's test for paired samples.

MAY 1996 27:5 ANNALS OF EMERGENCY MEDIOINE 5 9 7

EDD VERSUS ETco 2 Bozeman et al

lower than that of qualitative ETCO 2 detection. This over- all effect is the result of poor performance of ETco 2 moni- toring in the cardiac arrest group, with false- negative results in more than one quarter of such patients in our series.

ETCO 2 detectors have limitations in the emergency setting. Other studies have demonstrated failure of colori- metric ETCO 2 detectors to confirm tracheal intubation in more than 25% of correctly intubated Cardiac arrest patients.8,9, lo These failures have been presumed to be caused by a low circulatory flow or a severe mismatch between pulmonary ventilation and perfusion, r The con- sistent presence of ETCO 2 unequivocally confirms endo- tracheal intubation, but in cardiac arrest patients, the lack of detectable ETco 2 does not provide reliable information on ETT position, r

ETCO 2 detectors depend on the metabolic, circulatory, respiratory, and ventilatory systems of the patient. CO must be produced in the tissues, transported by the sys- temic and pulmonary circulatory systems, exchanged across alveoli, and excreted by ventilation before it can be detected by an ETco 2 monitoring device. Anesthesiologists may preoperatively evaluate all of these systems in elective surgical patients. Subsequently, negative readings by an ETCO 2 device in a newly intubated elective surgical patient may be interpreted in the operating theater as a failure of ventilation caused by a misplaced ETT. Any of these mechanisms, however, may be dysfunctional in the emergency patient, and a negative ETCO 2 reading may be caused by failure of any of the above systems.

In contrast to ETcO 2 monitoring, the EDD relies on the single parameter of anatomic differences between the trachea and the esophagus. The cartilaginous rings of the trachea hold it open during aspiration of air from the lungs, whereas the flbromuscular walls of the esophagus collapse when negative pressure is applied. These anatomic differences persist regardless of the metabolic or perfusion state of the patient. The lack of dependence on multiple systems makes the EDD particularly applicable in the emergency setting, where assumptions cannot be made about the condition of a previously unknown, criti- cally ill patient.

Failure of ETco 2 detectors because of blockage of cor- rectly placed ETTs has also been described. Colorimetric ETco 2 detectors have been noted to be blocked by blood, secretions, and debris, particularly in cases of acute pul- monary edema, r Similar failure of a sampling-type ETco 2 monitor has been reported in the operating suite setting, lo We report two additional cases of failure of ETCO 2 moni- toring in patients with pulmonary edema.

In addition to being more reliable than ETco 2 moni- tors in the emergency population, the EDD is less expen- sive, costing less than $2 to manufacture as described herein. The EDD requires neither electronics nor batteries to operate and has an indefinite shelf life.

We chose to construct the EDD in the fashion described by Wee and O'Leary in their original papers 1,2, using a 60-mL clear syringe to aspirate air. A second, equally effective type of EDD uses a self-inflating rubber bulb. 3,4,~ Our reasons for choosing the original type are threefold. First, the use of the rubber bulb model requires up to 30 seconds of apneic observation to confirm nonfifling of the bulb. 6 Emergency intubations often occur in hypoxic patients, and rapid testing is desirable because of a lack of optimal preoxygenation. In our experience, the syringe model was considerably faster. Second, the use of tactile sensation, rather than simple visualization and timing of bulb filling rate, was considered superior for detection of transient or partial vacuums. Third, aspiration of vomitus has been reported with the EDD. 4 The use of a clear syringe enables the tester to examine the aspirated material easily. Observation of vomitus may be difficult in the setting of an ETT secured with tape along its extraoral portion and connect by opaque fittings to a colored rubber bulb.

Limitations of this series of patients include its relatively small size and the lack of a single consistent gold standard. In fact, our test device performed better than the ETCO 2 monitor that was meant to confirm its results; hence the use of a repeated laryngoscopy to directly visualize tube placement became the default gold standard in cases of discrepant results. This study was designed to examine the ability of the EDD to confirm tra- cheal intubations in a group of correctly intubated emer- gency patients because any patients with clinically evident esophageal intubations were reintubated before the testing phase of the study. We believe that the ability of EDDs to detect esophageal intubations has been adequately demonstrated, although a theoretical concern remains about the possible false-negative rate of the device in the emergency population. Our data address this concern. Last, although EDD testing was physically performed in the ED setting, this study did include intubations performed by paramedics in the field. We believe that the external validity of this study is high and that results are applicable to adult patients intubated in both the ED and prehospital settings.

The EDD possesses great potential for use in the emer- gency setting. As its use becomes more common in the ED and prehospital care, larger studies will be warranted.

5 9 8 ANNALS OF EMERGENCY MEDICINE 27:5 MAY 1996

EDD V E R S U S ETco 2

Bozeraan et al

The EDD safely and reliably confirms tracheal intuba- tion in the emergency patient population. The EDD is more accurate than ETco 2 monitoring in confirming endotracheal intubation in cardiac arrest patients. As a result, the EDD is more accurate than spectrographic qualitative ETCO 2 monitoring in the overall emergency patient population.

Failure of a spectrographic qualitative ETco 2 monitor to confirm endotracheal ETT placement is common in patients with cardiac arrest. Although positive ETCO 2 results are reliable in patients with cardiac arrest, negative results by such devices should be confirmed by means of another method, such as the EDD.

On the basis of these data, we support the use of the EDD in emergency intubations. Caution should be exer- cised in interpreting indeterminate or negative results by the EDD in situations of pulmonary edema or other cases of possible blockage of the ETT.

Reprint no. 47/1/71618 Address for reprints:

Gabor D Kelen, MD

Bepartment of Emergency Medicine

The Johns Hopkins Hospital

600 North Wolfe Street

Baltimore, Maryland 21287

410-955-8191

Fax 410-955-0141

REFERENCES 1. Wee MYK: The oesophagaal detector device. Anesthesia 1988; 43:27-29.

2. O'Leary J J, Pollard B J, Ryan MAJ: Method of detecting oesophageal intubation or confirming tracheal intubation. Ann Intensive Care 1988;16:299-301.

3. Williams KN, Nunn JF: The oesophageal detector device. Anesthesia 1989;44:412-41&

4. Oberly D, Stein S, Hess D, et ah An evaluation of the esophageal detector device using a cadaveric model. Am J Emerg Med 1992;10:317-320.

5. Wee MYK: Comments on the esophageal detector device (letter). Anesthesia 1989;44:930-931.

6. Zaleski L, Abello D, 6aid MI: The esophageal detector device: Bees it work? Anesthesiology 1993;79:244-247.

7. Ornate JP, Shipley JB, Racht EM, et ah Multicenter study of a portable, hand size calorimetric ETCO z detection device. Ann Emerg Mad 1992;21:518-523.

8. MacLeed BA, Heller MB, 6erard J, et ah Verification ef endotracheal tube placement with cot- orimetric end-tidal CO 2 detection. Ann Emerg Mad 1991;20:267-270.

9. Menegazzi J J, Hailer MB: Endotracheal tube confirmation with calorimetric CO 2 detectors. Anesth Analg 1990;71:441-442.

10. Sosis MB, Sisarnis J: Pulse oximetry in confirmation of correct tracheal tube placement (let- ter). Anesth Analg 1990;71:309-310.

The authors express their appreciation for the generous loan of ETco 2 monitoring equipment by the Chesapeake Medical Corporation.

MAY 1996 27:5 ANNALS OF EMERGENCY MEDICINE 5 9 9