End Tidal Carbon Dioxide Monitoring During
A Predictor of Outcome
by Jinhee Nguyen, RN MSN
In the emergency department (ED), critical management of the patients airway, ventilation and chest compression requires the utmost effort and sophisticated technology to maximize successful outcomes. Approximately $1 billion is spent on emergency department and in-hospital care for cardiac arrest nonsurvivors (Bonnin et al., 1993; Gray, Capone & Most, 1991). A reliable method of assessing the efficacy and usefulness of ongoing cardiopulmonary resuscitation (CPR) would conserve scarce health care resources, avoid futile care, pain and suffering (Sanders, Kern, Otto, Milander & Ewy, 1989; Wayne et al., 1995). End tidal carbon dioxide (ETCO2) monitoring during resuscitation is a promising technology, which may help maximize patient outcomes, while reducing futile and costly interventions.
ETCO2 is the partial pressure or maximal concentration of carbon dioxide (CO2) at the end of an exhaled breath, which is expressed as a percentage of CO2 or mmHg (LaValle & Perry, 1995; Santos, Varon, Pic-Aluas & Combs, 1993; Trillo, von Planta & Kette, 1994). The normal values are 5% to 6% CO2, which is equivalent to 35-45 mmHg (Hollinger & Hoyt, 1995; Trillo et al., 1994). CO2 reflects cardiac output (CO) and pulmonary blood flow as the gas is transported by the venous system to the right side of the heart and then pumped to the lungs by the right ventricles (LaValle & Perry, 1995; Sanders, 1989). When CO2 diffuses out of the lungs into the exhaled air, a device called capnometer measures the partial pressure or maximal concentration of CO2 at the end of exhalation (Sanders, 1989; Santos et al., 1993). During CPR, the amount of CO2 excreted by the lungs is proportional to the amount of pulmonary blood flow (Benumof, 1998; Lambert, Cantineau, Merckx, Bertrand & Duvaldestine, 1992; Nielsen, Fitchet & Saunders, 1990).
The primary interventions of CPR include establishing a patent airway, ventilation, and precordial compressions. These interventions can help maintain blood flow and oxygen delivery to the vital organs with a cardiac output of about 17% to 27% (Falk, Rackow & Weil, 1988; Gudipati, Weil, Bisera, Deshmukh & Rackow, 1988). During CPR, the most reliable continuous monitoring of the efficacy of precordial compressions for blood flow is the measurement of arterial or intracardiac pressure (Callaham & Barton, 1990; Gudipati et al., 1988). Animal studies have reported the correlation of coronary perfusion pressure (CPP) to the likelihood of resuscitation, but simultaneous measurement of arterial pressure and central venous pressure is required to monitor the progress of the resuscitation (Gudipati et al., 1988; Sanders, Ewy, Bragg, Atlas & Kern, 1995). Furthermore, placing catheters to measure such pressures is invasive, time consuming, and impractical in the ED setting, especially when patients pulses are barely palpable (Callaham & Barton, 1990; Gudipati et al., 1988; Sanders et al., 1989). Therefore, ETCO2 monitoring is one promising noninvasive method of evaluating patients, who are most likely to be intubated during CPR (Cantineau et al., 1996; Lambert et al., 1992; Sanders et al., 1985; Steedman & Robertson, 1990).
This paper will discuss the significance of ETCO2 monitoring and assess its value in predicting the outcomes of CPR in ED and prehospital settings. The inclusion of prehospital setting is crucial since the resuscitation from the field is often continued in ED setting.
The current methods of assessing the effectiveness of CPR are difficult and unreliable (Falk et al., 1988; Sanders et al., 1989; Santos et al., 1993). The presence of femoral or carotid pulsations, pupillary signs, and arterial blood gas (ABG) results have not shown to correlate with successful CPR (Sanders et al., 1989; Santos et al., 1993; Steedman et al., 1990). Currently, palpating pulses is one of the options for assessing the adequacy of blood flow generated by cardiac compressions (Sanders et al., 1989). According to Advanced Cardiac Life Support (ACLS) guidelines by American Heart Association (1998), assessing for pulse is indicated before confirming the return of spontaneous circulation (ROSC). However, a palpable pulse may only represent continuity of a fluid filled vessel but not necessarily the blood flow (Gudipati et al., 1988). Therefore, palpable pulse may serve as an unreliable indication of operative systemic flow.
The most direct technique to measure the adequacy of ventilation is the measurement of the partial pressure of carbon dioxide in arterial blood (PaCO2) from an ABG, which can be expensive and painful (Santos et al., 1993). During cardiac arrest, the CO2 delivery to the lungs decreases due to poor pulmonary perfusion (Sanders et al., 1989). A decrease in pulmonary blood flow will cause the accumulation of CO2 in the venous circuit even with adequate ventilation (Steedman et al., 1990). CO2 delivery to the lungs is decreased due to a high ventilation/perfusion (VQ) ratio, leading to arterial alkalemia (Steedman et al., 1990; Weil, Rackow, Trevino, Grundler, Falk, & Griffel, 1986). ABGs can be misleading during CPR due to a phenomenon called venous paradox, which consists of venous hypercarbic acidosis and arterial hypocarbic alkalemia (Santos et al., 1993). In addition, ABGs only provide intermittent data, whereas capnometry allows continuous and instantaneous measurement of ETCO2 (Santos et al., 1993).
The significance of ETCO2 monitoring during CPR was first noted by Kalenda (1978), who observed the resuscitation of three patients and monitored pulmonary perfusion by means of ETCO2. He reported that ETCO2 monitoring was helpful in assessing rescuer exhaustion. With the external cardiac massage, there was a slight improvement of ETCO2, which diminished as the rescuer became tired. Replacing the fatigue rescuer with a fresh rescuer resulted in an improvement in ETCO2. Additionally, Kalenda (1978) noted that the sudden steep increase in ETCO2 was an indication of the return of spontaneous cardiac activity. This study showed a potential significant value of ETCO2 as a guiding tool to monitor the effectiveness of CPR.
Following the observation made by Kalenda (1978), several experimental studies involving animals have demonstrated the correlation between ETCO2 with CO and CPP (Gudipati et al., 1988; Ornato, Garnett, & Glauser, 1990; Sanders et al., 1985). The next series of in-hospital studies attempted to confirm the same findings seen from experimental studies by looking at ETCO2 monitoring during various hemodynamic stages of CPR in the ED. More recent studies have used ETCO2 monitoring in prehospital settings in an attempt to predict the outcome of CPR (Asplin & White, 1995; Cantineau et al., 1996; Levine, Wayne & Miller, 1997; Wayne et al., 1995).
Organization of Review
The process of gathering the related literature was accomplished by accessing the MEDLINE and CINAHL databases (1985-1999) at University of California, San Francisco. The databases contained both medical and nursing journals. The title word end tidal carbon dioxide in combination with two key words, cardiac output and cardiopulmonary resuscitation, identified 18 articles. Majority of the articles were published by medical journals, whereas only two articles were published by nursing journals (Benumof, 1998; LaValle & Perry, 1995). For the purpose of this paper, the articles related to ED and prehospital settings were selected for in-depth discussion.
Discussion of Theory/Pathophysiology
To understand the significant value of ETCO2, one needs to be familiarized with the following: normal physiology of CO2, principle determinants of ETCO2, CO2 gradient with normal VQ relationship, ETCO2 analyzer (capnometer), and limitations of ETCO2 measurements.
ETCO2 represents the partial pressure or maximal concentration of CO2 at the end of exhalation (LaValle & Perry, 1995; Santos et al., 1993; Trillo et al., 1994). CO2 reflects cellular metabolism (Idris et al., 1994; Sanders, 1989). There are four main stages of normal physiology of CO2 : production, transport, buffering and elimination. CO2 is a metabolic byproduct of aerobic cell metabolism (LaValle & Perry, 1995; Sanders, 1989). As the intracellular CO2 increases, CO2 diffuses out into the tissue capillaries and is carried by the venous circulation to the lungs, where it diffuses from pulmonary capillaries into the alveoli (Idris et al., 1994; LaValle & Perry, 1995). The partial pressure of CO2 (PaCO2) of venous blood entering pulmonary capillaries is normally 45 mmHg; the partial alveolar pressure of CO2 (PACO2) is normally 40 mmHg (LaValle & Perry, 1995; Trillo et al., 1994). The pressure difference of 5 mmHg will cause all the required CO2 to diffuse out of pulmonary capillaries into the alveoli (LaValle & Perry, 1995). The second stage is CO2 transport, which is a way of maintaining the CO2 tension of arterial blood at approximately 35-45 mmHg despite high CO2 production (Trillo et al., 1994).
The third stage is where the buffer action of hemoglobin and pulmonary blood flow maintain the normal level of CO2 tension by eliminating the excess CO2 (Trillo et al., 1994). CO2 can either be carried, dissolved or combined with water (H20) to form carbonic acid (H2CO3), which can dissipate to hydrogen ions (H+) and bicarbonate ions (HCO3-): (CO2 + H20 « H2CO3 « H+ + HCO3-) (LaValle & Perry, 1995; Trillo et al., 1995). The hydrogen ions are buffered by hemoglobin, and the bicarbonate ions are transported into the blood (Idris et al., 1994; Trillo et al., 1994). This mechanism accounts for 90% of CO2 transport (LaValle & Perry, 1995; Trillo et al., 1994). The fourth stage involves CO2 elimination by alveolar ventilation under the control of the respiratory center (LaValle & Perry, 1995; Trillo et al., 1994). This process allows the diffusion of CO2 from blood to the alveoli where the partial alveolar pressure of CO2 is lower than the tissue pressure (Trillo et al., 1994).
During normal circulatory condition with equal VQ relationship, PACO2 is closely comparable to PaCO2 and ETCO2; therefore, PaCO2 is equivalent to ETCO2 (Sanders, 1989; Santos et al., 1993). The difference between PaCO2 and ETCO2 is known as the CO2 gradient (LaValle & Perry, 1995). The normal ETCO2 is about 38 mmHg at 760 mmHg of atmosphere with less then 6 mmHg gradient between PaCO2 and ETCO2 (Sanders, 1989; Santos et al., 1993).
The principle determinants of ETCO2 are alveolar ventilation, pulmonary perfusion (cardiac output) and CO2 production (Benumof, 1998; Callaham & Barton, 1990; Domsky, Wilson & Heins, 1995; Steedman & Robertson, 1990). During acutely low cardiac output state as in cardiac arrest, decreased pulmonary blood flow becomes the primary determinant resulting in abrupt decrease of ETCO2 (Domsky et al., 1995; Wayne et al., 1995). Changes in alveolar ventilation can also influence ETCO2 as PACO2 closely approximates PaCO2 and ETCO2 (Santos et al., 1993). If ventilation and chest compressions are constant with the assumption that CO2 production is uniform, then the change in ETCO2 reflects the changes in systemic and pulmonary blood flow (Garnett et al., 1987). Ultimately, ETCO2 can be used as a quantitative index of evaluating adequacy of ventilation and pulmonary blood flow during CPR (Falk et al., 1988).
ETCO2 Monitoring Technologies
One way of measuring ETCO2 is with the infrared capnometer, which contains a source of infrared radiation, a chamber containing the gas sample, and a photodetector (Santos et al., 1993; Trillo et al., 1994). When the expired CO2 passes between the beam of infrared light and photodetector, the absorbence is proportional to the concentration of CO2 in the gas sample (LaValle & Perry, 1995; Santos et al., 1993; Trillo et al., 1994). The gas samples can be analyzed by the mainstream (in-line) or sidestream (diverting) techniques (Santos et al., 1993).
The mainstream capnometer uses the airway connector placed in-line with the patients breathing circuit directly attached to endotracheal tube (Santos et al., 1993).This technique produces instantaneous and accurate gas analysis (less than 500 msec) far superior to the sidestream technique (Callaham & Barton, 1990; Santos et al., 1993). The limitation of the mainstream technique is that it can be applied only for intubated patients or patients with tight fitting nose or face masks (Santos et al., 1993). On the other hand, the sidestream technique is more applicable for spontaneously breathing non-intubated patients. But, it requires constant aspiration of 100 cc to 300 cc of expired air/minute for analysis (Santos et al., 1993).
ETCO2 analyzer adapters do not interfere with ventilation (Santos et al., 1993). However, there are several factors that can alter ETCO2 measurements such as temperature of photo detector, airway pressure, calibration error, contamination of sample chamber by moisture and secretion (LaValle & Perry, 1995; Sanders, 1989). Furthermore, falsely high or low ETCO2 measurements can be resulted by buffer agents (i.e. NaHCO3), recent ingestion of carbonated beverages, antacids, a-agonists (i.e. epinephrine) and mask ventilation with CO2 containing gases such as nitrous oxide (Benumof, 1998; Sanders, 1989; Santos et al., 1993; Trillo et al., 1994).
An alternative to capnometer is a disposable ETCO2 detector, a colorimetric chemical indicator (Santos et al., 1993). This is a qualitative measurement of ETCO2 that changes color in the presence of CO2 (i.e. purple to yellow with presence of CO2). This paper will only address studies, which utilized capnometers to measure ETCO2 during CPR in ED and prehospital setting.
Research Critique: Experimental Studies
Before 1990, numerous experimental animal studies attempted to identify the relationship between ETCO2 and CO. An early experimental study by Sanders et al.(1985) attempted to determine if ETCO2 measurement during CPR could be used as a prognostic indicator of successful resuscitation from cardiac arrest in animal model. Using the time series design, the researchers measured ETCO2 during different phases of resuscitation (Wilson, 1989). This experimental study has qualities of manipulation, control and randomization (Polit & Hungler, 1995). Twelve healthy mongrel dogs (mean weight 11.3 ± 1.9 kg) were anesthetized with pentobarbital and endotracheally intubated. A capnometer was attached to the endotracheal tube (ETT) to record ETCO2 with each respiration. Catheters were inserted through femoral arteries and veins to reach the heart. Postmortem examination confirmed catheter placement. After the dogs were electrically fibrillated with 50 Hz through the pacing catheter, chest compressions and assisted ventilation (5:1 compression/ ventilation ratio) were started using the mechanical resuscitator. Half of the dogs received high pressure chest compressions (80 lb.) and half received low pressure compressions (40 lb.). Epinephrine 0.5 mg was given at 3, 6, 9, and 12 minutes (min) of ventricular fibrillation (VF). After 15 min of chest compressions and assisted ventilation, animals were defibrillated with 80J. Data on ETCO2, aortic pressure, right atrial pressure, and ECG were recorded. The process for calibration for all these instruments was not found in the article. The mean and standard deviation of ETCO2 and coronary perfusion pressure (CPP) at each minute of VF were calculated for resuscitated and nonresuscitated. For the remainder of this paper, the word return of spontaneous circulation (ROSC) will be used in describing resuscitated subjects.
For statistical analysis, two-tailed Student t test was used to detect the significant difference (P<.05) between ETCO2 of animals with ROSC and without ROSC at each minute of VF. According to Wilson (1989), t test for continuous variables is appropriated selection for this experimental study. There was a significant difference between ROSC and without ROSC animals in the overall mean ETCO2 (9.6 ± 3.2 mmHg vs. 3.2 ± 1.1 mmHg). Linear regression analysis was done to determine the relationship of ETCO2 to CPP during CPR. ETCO2 was highly correlated with CPP (r=.91; P<.01). Thus, it was concluded that CPP could be predicted from ETCO2 measured in this sample. This study used healthy dogs in which VF was induced artificially. Therefore, the applicability of these findings to the resuscitation of patients with cardiac and pulmonary diseases is questionable. Also, thoracic cavity is different in dogs; thus, chest compressions may produce different results. Future clinical studies are required for validation of these findings. Lacking routine measurements of arterial blood pressure and tidal volumes as part of the procedure evidenced some of the limitations of this study. It is possible that differences in these parameters may have influenced ETCO2. However, the measurements were not analyzed by the study; thus, the potential relationship is unknown.
Gudipati et al. (1988) also conducted an experimental time series study to investigate the potential usefulness of ETCO2 as a prognosticator of resuscitability. Minnesota mini pigs (N=22; weight 20-35 kg) were anesthetized with ketamine, pentobarbital sodium and pancuronium, a blocking agent. The pigs were intubated and mechanically ventilated with a set rate, tidal volume and mix flow rate. Femoral arteries and veins were used to insert pressure monitoring catheters to reach the heart. All instruments and procedures were thoroughly described in the article, but the description of ETCO2 analyzer calibration was not found in the article. For the ETCO2 measurement, a sidestream analyzer was attached to the ETT. CO was measured by the thermodilution technique. Before initiating CPR for cardiac arrest of ventricular fibrillation (VF), CO, ETCO2, aortic and right atrial pressures were monitored with pulmonary arterial and aortic blood gases. A current of 5 mA was delivered to the right ventricle to induce VF. Cardiac arrest was confirmed by VF on ECG with a decline in aortic systolic pressure to less than 25 mm Hg and decline in pulse pressure to less than 5 mmHg. At that time, the FIO2 was increased to 100% to maintain PaCO2 of 35-45 mmHg. The ventilator and compressor delivered 5:1 compression/ventilation ratio. Thumper, a programmable cardiopulmonary resuscitator, delivered a set rate of compression with equal interval and depth at a regular cycle. At 1, 5, and 9 min during CPR, CO and ETCO2 measurements were obtained. After 12 min of CPR, countershocks (40, 80, 160 and 320 J) were applied to the anterior chest with a defibrillator. If animals were successfully resuscitated, CO measurements were repeated at 6 min after resuscitation.
The second part of the experiment consisted of laparotomy (N=5), creating a window with a midline incision on pigs, who received same anesthesia, mode of ventilation and chest compressions. After 1 min of VF, direct cardiac massage was performed through the window for 5 min. Then, defibrillation with unknown amount of direct-current countershock was attempted. During this procedure, ETCO2 was continuously monitored and cardiac index (CI) was measured at 1 and 5 min after the initiation of direct cardiac massage. For statistical analysis, the difference in CI and ETCO2 between animals with ROSC and without ROSC were analyzed by the Student t test for unpaired measurements. This test was appropriate for testing the difference between the means of two independent groups (i.e. ROSC vs. without ROSC) (Norman & Streiner, 1997). The measurements of ETCO2 in ROSC and without ROSC animals were analyzed by repeated measurement analysis of variance by Wilks method. According to Polit and Hungler (1995), this test is used when there are three or more measures of the same dependent variable for each subject. For example, this type of test is used when multiple measures of the same dependent variables are collected longitudinally at several points in time (Polit & Hungler, 1995). With linear regression analysis, the correlation between ETCO2 and CI was analyzed as one independent variable (CI) was used to predict the dependent variable (ETCO2) (Polit & Hungler, 1995). The results showed a rapid increase in ETCO2 after ROSC. At 1 and 9 min after precordial compressions, the CI averaged 27 ± 1% and 38 ± 4% of prearrest values, respectively; ETCO2 averaged 27 ± 5% and 43 ± 8% of precardiac arrest value. After 6 min of spontaneous circulation, CO increased to 88 ± 5% and ETCO2 increased to 121 ± 80% of precardiac arrest value. There was a significant difference in the average ETCO2 of the animals with ROSC and without ROSC (N=7 with ROSC, 1.7 ± 0.2%; N=5 without ROSC, 0.5 ± 0.1% , P<.001). The correlation between ETCO2 and CO for 22 animals averaged .92 ± .07 (ranged from .75 to .99; p<.001).
For open chest cardiac massage of 5 animals, ETCO2 and CI were measured before cardiac arrest, 1 and 5 min after beginning direct cardiac massage, and after ROSC. The researchers also reported similar proportional changes in ETCO2 and CI in the open chest animals. The correlation coefficients from the relationship between ETCO2 and CI for 5 animals ranged from .91 to .98 (mean .95 ± .014). This study showed that ETCO2 could be a noninvasive measure of pulmonary blood flow reflecting CO. Furthermore, ETCO2 was shown to identify ROSC and helped predict outcomes during CPR. Additionally, the dramatic increase in ETCO2 to a level exceeding the prearrest values provided a definite evidence of ROSC, suggesting that precordial compressions might not need to be interrupted to assess for ROSC. The findings in this study are promising but the generalizability to humans is speculative, requiring clinical trials to confirm its implication in clinical settings.
Research Critique: In-Hospital Studies (ED)
In the late 1980s, several clinical studies were conducted in an attempt to confirm some of the findings from the experimental studies. The study by Garnett et al. (1987) was one of the first prospective quasi-experimental time series studies to determine the usefulness of ETCO2 in hemodynamic monitoring during CPR. The prospective time series design was an appropriate selection, allowing the researchers to use "before" measurements to establish a baseline while comparing posttreatment measurements (Wilson, 1989). According to Polit and Hungler (1995), quasi-experimental design does not contain a control group lacking randomization but has some manipulation of the procedure. In fact, strengths of this design are practicality, feasibility and generalizability (Polit & Hungler, 1995). Overall, the study design selection was appropriate. The inclusion criteria for the convenience sample were orally intubated adults (age ³ 18 years) with atraumatic prehospital cardiac arrest. The prehospital treatment consisted of basic and ACLS according to AHA. Upon ED arrival, CPR was continued using Thumper. A calibrated sidestream capnometer measured ETCO2 continuously. The article indicated that ETCO2 was reported as the percentage of CO2 in exhaled gas, but the normal value of CO2 was not stated in the article. The threat to the reliability and validity of the instrument was controlled by using a calibrated capnometer. ROSC was confirmed by the detection of pulse rate and audible or palpable blood pressure. Of twenty-three patients, whom had ETCO2 recorded during resuscitation, ten patients had ROSC. For data analysis, unpaired Student t-test appropriately compared the means of ETCO2 in the two groups: the group with ROSC and without ROSC (Norman & Streiner, 1997; Polit & Hungler, 1995). Paired Student t test compared two paired observations (ETCO2 values before ROSC and the peak ETCO2 after ROSC). P< 0.05 was considered significant.
Before ROSC, ETCO2 measurements in the two groups were similar (with ROSC, N=10, 1.7 ± 0.6% vs. without ROSC, N=13, 1.8 ± 0.9%). In the ROSC group (N=10), ETCO2 increased so immediately that it was the very first clinical indicator of any changes in the patients status. The ETCO2 of the ROSC group peaked three times greater than before ROSC within 2-5 min (4.6 ± 1.4%). Then, ETCO2 slowly declined to a stable level (3.1 ± 0.9%). Therefore, this study suggested that ETCO2 could be useful in predicting the positive outcome of CPR. The potential limitation of this study was that error could be introduced by the technological difficulties from Thumper or the method of mechanical ventilation. Although the sample size was small, the findings were generalizable to the ED settings, where ETCO2 could be used to guide the treatment during resuscitation after atraumatic cardiac arrest.
Sanders et al.(1989) conducted the first clinical trial to determine if ETCO2 could be used as the prognostic indicator for ROSC. A convenience sample from two hospitals was included in this study. The inclusion criteria were the same as the previous study by Garnett et al. (1987). The exclusion criteria were described as traumatic cardiac arrest, invasive CPR, and ROSC after defibrillation without the need for intubation. Following intubation, a calibrated mainstream ETCO2 analyzer was placed between ETT and resuscitation bag. Manual ventilation presented a threat to the internal validity as hyperventilation could falsely decrease ETCO2 (Callaham & Barton, 1990). ETCO2 measurements obtained five min after NaHCO3 administration were eliminated from the analysis due to the potential for a false increase in ETCO2 measurements. For data analysis, unpaired two tailed Students t test was used to compare the means of continuous variables (i.e. ETCO2). To determine the association of the nominal data (i.e. initial cardiac rhythm) with ROSC group, chi square analysis was used appropriately. Linear logistic regression is a procedure that analyzes the relationship between multiple independent variables (i.e. ETCO2 values at different time intervals) and the categorical dependent variable (ROSC vs. without ROSC) (Polit & Hunger, 1995). This test was appropriately used to evaluate the predictive effect of ETCO2 (independent variable) on patients with ROSC (yes/no-categorical dependent variables).
Of 133 eligible patients, only 35 patients were continuously monitored with capnometry and were eligible for analysis. Nine patients with ROSC had a mean ETCO2 of 15 ± 4 mmHg, which was greater than the mean ETCO2 for 26 patients without ROSC (7 ± 5 mmHg). The mean ETCO2 of ROSC group at 11 to 15 min was 17 ± 4 mmHg, whereas the mean ETCO2 for the group without ROSC was 9 ± 7 mmHg (P=.04). At 16 to 20 min, the difference between two groups further increased to 18 ± 6 mmHg vs. 6 ± 3 mmHg (P=.0004). All patients (N=9) with ROSC had the ETCO2 ³ 10 mmHg, but six patients without ROSC had ETCO2 of 10 mmHg. Sanders et al. (1989) explained that, despite adequate perfusion pressure for resuscitation, severe myocardial or coronary disease can prevent successful resuscitation. Therefore, one can conclude that ETCO2 greater than 10 mmHg do not guarantee a successful resuscitation. In fact, the threshold of 10 mmHg may not be generalizable for all patients, especially those with comorbidities. However, ETCO2 of 10 mmHg as a prognostic threshold value produced a sensitivity of 100% and a specificity of 77%, positive predictive value of 60% and negative predictive value of 100% in this study. For the initial cardiac rhythm, VT/VF were associated with survival (P<.05) and asystole/EMD were associated with nonsurvival (P<.05). As pointed out by the authors, the generalizability could be limited by the potential inconsistency of instrumentation and collecting data on ETCO2 measurements in the two hospitals. Additionally, the chest compressions and ventilation were not quantified, which could effect ETCO2. The overall findings from the study by Sanders et al. (1989) concluded that ETCO2 could be used as a prognostic guide during CPR after atraumatic cardiac arrest.
Callaham and Barton (1990) also investigated the utility of ETCO2 as a predictive measure of initial outcome of resuscitation (i.e. ROSC) using the same study design and sample selection. The eligible patients (N=55) were atraumatic prehospital cardiac arrest victims, who were transported to the ED, where the study was conducted (age, 69.9 ± 12.2; N=35 male, 64%). The prehospital treatment included basic/advanced life support according to AHA. In the ED, patients were resuscitated by the most senior physicians, who were not aware of the possible significance of ETCO2 during this study period. Orally intubated patients were manually ventilated, which introduced a threat to the internal validity for lacking consistent control of the ventilation. Thumper delivered chest compressions (80/min with 50% compression cycle). Callaham and Barton (1990) used a calibrated mainstream capnometer for obtaining ETCO2. The main focus were the initial values of ETCO2, which were obtained both within five min of ED arrival and after one min of ventilation. The frequency of regular interval monitoring and recording of ETCO2 were not defined in the article. For statistical analysis, unpaired two tailed Students t-test compared the initial ETCO2 values of the two groups (with ROSC and without ROSC). Fourteen of 55 patients had ROSC. The mean initial ETCO2 of ROSC group, whose ETCO2 measurements were obtained both within five min of ED arrival and after one min of ventilation, was three times higher than the group without ROSC (19 ± 14 mmHg vs. 5 ± 4 mmHg, P<.0001). An initial ETCO2 of 15 mmHg correctly predicted ROSC with a sensitivity of 70%, specificity of 98%, positive predictive value 91% and negative predictive value of 91%. The threshold of 15 mmHg was deduced by plotting the ETCO2 values on the receiver operating curve. For the entry cardiac rhythm, there was a significant ROSC in patients presenting with asystole (P<.035) and EMD (P < .01) and compared to VF/VT. This finding was inconsistent with Sanders et al.(1989), who reported that asystole/EMD were associated with nonsurvival (P<.05). Callaham and Barton (1990) concluded that the initial predictive value of 15 mmHg was not limited to a particular rhythm. A potential threat of Hawthorne effect (Polit & Hungler, 1995) could have caused the physicians to perform more aggressive resuscitation on patients in this study although the physicians were not aware of the significance of ETCO2. Furthermore, inadequately controlled manual ventilation could have potentially affected ETCO2. The generalizability of the findings from this study is limited to the ED population of atraumatic cardiac arrest without comorbidities such as COPD and ARDS. This study also demonstrated that a sudden rise of ETCO2 often was the first indication of ROSC as seen in the studies from Garnett et al. (1987). Additionally, Callaham and Barton (1990) were able to deduce a predictive relationship between ETCO2 and the positive outcomes from CPR as did Sanders et al. (1989).
Steedman and Robertson (1990) in United Kingdom (UK) assessed the clinical applicability of ETCO2 measurement during CPR in the ED using the same study design and sample selection as the previously described studies. The major difference was that patients received only basic life support in the field. Upon arrival to the ED, the patients were orally intubated and ventilated with 100% oxygen. The calibrated mainstream capnometer monitored ETCO2 continuously and reported the values as the percentage of CO2 in exhaled gas. Steedman and Robertson (1990) controlled the potential threats to the internal validity by using a Thumper and mechanical ventilation. The paired t test was used to compare ETCO2 before and after ROSC. The unpaired t-test compared the means of ETCO2 in patients with and without ROSC.
Of the small sample size of 12 patients (age 69 ± 17; N=8 male), five patients had ROSC. In the ROSC group, ETCO2 gradually increased from 1.8 ± .92% to 3.38 ± 1.06% within one min (p < 0.05). ETCO2 reached its peak value (3.7 ± 1.08%) at two min. A stable level of ETCO2 (3.02 ± 0.52%) was maintained after a slight drop from the peak value at eight min or more of ROSC. Similar phenomenon was also reported by Garnett et al.(1987). ETCO2 measurements were significantly different (P<0.001) between patients with and without ROSC (N=5, 2.63 ± 0.32%; N=7, 1.64 ± 0.89%, respectively). ROSC group in this study showed a two-fold increase in ETCO2. Four patients received a mean dose of 100 mEq NaHCO3 but did not show any statistical difference in ETCO2 (1.5 ± 0.49% to 1.8 ± .54%; P value was not available). One of the limitations of this study was that ACLS was not performed in prehospital, and the impact of this factor was not analyzed. The findings from this study confirmed the similar results from three previous studies, even though there was a small sample size. Therefore, the generalizability is evident. Steedman and Robertson (1990) concluded that ETCO2 monitoring could provide a useful noninvasive method to assess the effectiveness of blood flow produced by precordial compression after atraumatic cardiac arrest.
Research Critique: Prehospital Studies
Emergency medical service personnel rely on ECG rhythm, down time, time elapsed from CPR and the patients medical history as a prognosticator since there are no reliable data (Aspline & White, 1995). Aspline and White (1995) conducted a study to determine the prognostic significance of initial ETCO2 measurement and to explore the role of capnography in the management and outcome of out of hospital cardiac arrest (OHCA). The purpose of this study was to test the hypothesis that initial ETCO2 measurements are higher in patient with ROSC during CPR and are therefore prognostic for ROSC in patients with OHCA. This was a prospective time series design looking at a convenience sample (N=34) of atraumatic OHCA over 13 months. According to the ACLS guidelines, the cardiac arrest victims were intubated endotracheally and continuously monitored with a mainstream analyzer. The ETCO2 analyzer was calibrated according to the manufacturers guideline. Automated ventilation, which was oxygen powered and time cycled, was used to maintain ventilation (10-12 per min with tidal volume of 600-800 ml). Since this was not a blinded study, having the automated ventilation was a good control for the internal validity. Without this control, the participants might aggressively bag the first 1-2 min to achieve a high initial ETCO2. The initial ETCO2 was recorded after 1-2 min of monitoring, and the maximum ETCO2 was recorded. Due to the potential effect of NaHCO3 on ETCO2, no ETCO2 within 5 min of NaHCO3 administration was documented for the analysis. This also was a good control for the threat to internal validity. For data analysis, one-tailed t test was used to compare the initial and maximum ETCO2 between the two groups (with ROSC and without ROSC). P £ .05 was considered significant. The one tailed test was appropriate since the hypothesis was testing the higher value of ETCO2.
ETCO2 measurements were documented only on 34 patients out of 65 patients. The potential problem related to the attrition was not addressed in the article. Of 34 patients, seven patients had ROSC before capnometry monitoring was started; thus, ETCO2 was not measured. The final 27 patients had ETCO2 readings available during CPR. After 1 min, ETCO2 in 14 patients with ROSC was significantly higher than patients without ROSC (23.0 ± 7.4 mmHg vs. 13.2 ± 14.7 mmHg, P=.0002) as seen in the study by Callaham and Barton (1990). During CPR, the maximum value was higher in patients with ROSC than patients without ROSC (30.8 ± 9.5 mmHg vs. 22.7 ± 8.8 mmHg, P=.0154). Only three patients were survived to hospital discharge, but their neurological status was not described in the article. The usefulness of ETCO2 monitoring as a prognosticator for ROSC was promising but questionable, due to the small convenience sample size and small survival patients. Therefore, a larger sample is recommended.
Wayne et al.(1995) conducted a study to evaluate the quantitative measurement of ETCO2 in prehospital setting and to determine whether this value could be used as a marker to predict death. This study was a prospective time series design looking at patients (N=90, 61 males; age 67.6 ± 13.6), who were victims of normothermic and atraumatic pulseless electrical activity (PEA). The reason for selecting PEA as an inclusion criterion was not described in the article. The exclusion criteria were having VF or VT. All samples were intubated and connected to a self-calibrated mainstream ETCO2 analyzer and ventilated by a standard adult bag valve mask. Using manual ventilation introduced a threat to the internal validity. The termination of resuscitation was based on the presence of persistent ETCO2 level of 10 mmHg or less after 20 minutes of ACLS. The researchers explained that an ETCO2 level of 10 mmHg represented less than 1% of CO2 production and was thus incompatible with life. But in reality, due to potential survival, the resuscitation went on until loss of electrical activity or ROSC, resulting in prolonged effort of CPR. For the analysis, Wilcoxin rank sum test was used for ordinal data. This test involves taking the difference between paired scores and ranking the absolute difference (Polit & Hungler, 1995). It is a nonparametric equivalent of the paired t test to compare group mean (i.e. initial, minimum ETCO2). The logistic regression was appropriately used to see the association between ETCO2 and survival to hospital discharge.
Seven out of 16 patients, who had ROSC, were discharged from the hospital. Four of the seven patients were neurologically intact; three had some neurologic impairment but were able to care for themselves. According to the results, the initial ETCO2 for the patients without ROSC was 11.7 ± 6.6 mmHg vs. 10.9 ± 4.9 mmHg with ROSC. After 20 min of ACLS, the average ETCO2 in the group without ROSC was 3.9 ± 2.8 mmHg and 31 ± 5.3 mmHg with ROSC (P<.0001). In 13 patients, the rise in ETCO2 was the first indication of ROSC before palpable pulse or blood pressure. Following this analysis, Wayne et al. (1995) hypothesized that the ETCO2 measurement £ 10 mmHg was a predictive value for death in the field. The study reported that no patients survived with ETCO2 less than 10 mmHg. At 20 min of ACLS, this threshold value of ETCO2 was accurate in predicting death in patients with atraumatic PEA. However, ETCO2 did not discriminate between the long term survivors and those who were deceased in hospital. The limitation of this study was that it contained a small sample size for prospective study. Further study with a large sample in different population would be required. Secondly, epinephrine did not effect the study results although other studies reported its decreasing effect on ETCO2. Thirdly, there was a potential for threat to the validity due to an interpreter variability in maintaining consistent minute ventilation using a bag valve device. Future studies confirming the accuracy of ETCO2 monitoring as a prognosticator of death or survival can help prevent futile efforts of CPR in the field.
Cantineau et al. (1996) conducted a prospective time series study to test the hypothesis that continuous assessment of ETCO2 during prehospital CPR allowed distinction between patients with ROSC and without ROSC with a sensitivity greater than 90%. A convenience sample (N=120: 84 males and 36 females) with atraumatic prehospital arrest were included in the study. Fire fighters started resuscitation with bag-valve-mask ventilations and supplemental oxygen and chest compressions according to AHA. An anesthesiologist placed an ETT in all patients and continued mechanical ventilation with a set rate and tidal volume. The mechanism of chest compressions was not described in the article. The pharmacological interventions were given according to ACLS guidelines. Epinephrine was given to all patients. NaHCO3 was given for CPR greater than 10-15 min. Lidocaine and direct shock were administered for patients with VF. ROSC was defined as sustained blood pressure for at least 30 seconds. There were two parts to the study. The first part of the study determined a threshold value of ETCO2 that would be tested on a large scale. Twenty four patients had the ETCO2 measured with a sidestream analyzer. The ETCO2 analyzer calibration was not described in the article. There were three categories of ETCO2 for data collection and analysis: an initial ETCO2 measurement after 1 min of mechanical ventilation, minimum ETCO2 and maximum ETCO2 measurements after the first 20 min after intubation and before ROSC. The sensitivity and specificity were calculated for each category. With NaHCO3, the ETCO2 measurements just before and 5 min after medication administration were included for the analysis. The second part of the study prospectively evaluated the maximum ETCO2 of 10 mmHg on 96 patients in a quasi-experimental design. Unlike the first part of the study, ETCO2 was measured by a mainstream analyzer. The reason for changing the technique to analyze ETCO2 was not described in the article. Having two different instruments was a threat to the external validity. For statistical analysis, unpaired student t-test was appropriately used for continuous data comparison of the independent group (cutoff ETCO2). Chi square for analyzing the categorical variables was an appropriate method to classify whether the ETCO2 measurement was below or above categorical groups of minimum and maximum groups. Student t-test (paired) appropriately analyzed the difference between the means of the two related groups (i.e. comparing ETCO2 before and after ROSC, ETCO2 before and after NaHCO3).
For the first part of the study, eight of the 24 patients had ROSC. The mean ETCO2 measurements were not significantly different in the ROSC or without ROSC group (16 ± 6.2 mmHg vs. 10.6 ± 8.4 mmHg; P=.12). ROSC was related to the significant increase in ETCO2 from 26.4 ± 5.4 mmHg to 48.7 ± 13.8 mmHg (P<.01). In the second part of the study, thirty patients out of 96 patients had ROSC. The maximum ETCO2 greater than 10 mmHg predicted ROSC with a sensitivity of 100% and specificity of 66% after 20 min of intubation. The maximum ETCO2 less than 10 mmHg was never associated with ROSC despite 40 min of resuscitation after intubation. Since this study contained a large number of patients with asystole compared to patients with VF/PEA, the generalizability of the findings is limited (asystole: N=109 vs. VF/PEA: N=11). However, this study confirmed the critical application of ETCO2 monitoring during resuscitation, which could be used to assess the prognosis of prolonged CPR in the prehospital setting as well as in the ED.
Critical and Original Analysis of Literature
The overall goal of the previously discussed studies was to determine the relationship of ETCO2 with CO and to evaluate ETCO2 as a prognosticator of CPR. In experimental studies, ETCO2 was shown to correlate well with CO, CPP and resuscitation outcome (Gudipati et al., 1988; Sanders et al., 1985). As immediate indicator of successful resuscitation in animal models, this optimistic quality of ETCO2 served as a basis for human application during atraumatic cardiac arrest.
Applying the significant findings derived from experimental studies, the usefulness of ETCO2 monitoring and the predictive value of ETCO2 were later evaluated in clinical settings (Callaham & Barton, 1990; Garnett et al., 1987; Steedman & Robertson, 1990). The same study design and sample selection were used in these studies. Common limitations to these studies were evidenced by using small sample size for the prospective design and inconsistent method of resuscitation. Garnett et al. (1987) and Steedman and Robertson (1990) used Thumper and mechanical ventilator, whereas Sanders et al. (1989) applied manual ventilation and failed to quantify and indicate the method of chest compression. Callaham and Barton (1990) used Thumper, but ventilation was not controlled mechanically. Sanders et al. (1989) and Callaham and Barton (1990) deduced threshold values for predicting a positive outcome from resuscitation (10 mmHg and 15 mmHg, respectively). Although Garnett et al. (1987) did not find any prognostic value for ROSC, the importance of ETCO2 for monitoring blood flow and guiding the treatment of CPR was confirmed as seen in the studies by Garnett et al.(1987) and Steedman and Robertson (1990). Callaham and Barton (1990) explained that the reason why Garnett et al. (1987) did not achieve the prognostic relationship of ETCO2 was related to using the sidestream technique. The sidestream technique is less accurate method of sampling CO2, and the measurement is depended on the location of sampling tube and the amount of supplemental oxygen (Callaham & Barton, 1990; LaValle & Perry, 1995).
The prehospital studies also showed the prognostic value of ETCO2 during CPR and observed similar findings (Aspline & White, 1995; Cantineau et al., 1996; Wayne et al., 1995). The methods of resuscitation involving ventilation and compressions varied in each study. Aspline and White (1995) concluded that the initial ETCO2 could be a prognosticator for ROSC when automated ventilation was used during CPR. Despite the small sample size, these findings were similar to the findings from Callaham and Barton (1990). Similarly, Wayne et al. (1995) deduced a threshold value of less than 10 mmHg as the prognosticator for irreversible death in OHCA. However, the generalizability of this finding was limited due to the inclusion criterion of OHCA with PEA. With a larger sample size (N=120), Cantineau et al. (1996) conducted a two-part study in establishing that the maximum ETCO2 value of 10 mmHg had a higher sensitivity (>90%) to predict ROSC in OHCA of asystole. The generalizability of this finding was limited to patients with asystole due to the small number of samples with VF/PEA in the study. Additional threat to the validity of ETCO2 was introduced by using different types of ETCO2 analyzer. Nonetheless, all three prehospital studies supported the findings from in-hospital studies, demonstrating the potential value of ETCO2 as a prognosticator and as a guiding tool for the effectiveness of CPR.
Integration and Synthesis
Achieving the threshold value for prognostic determination was the most challenging and difficult accomplishment for all the studies discussed. Sanders et al. (1989) reported a mean ETCO2 of 15 mmHg for patients with ROSC and a mean of 7 mmHg for patients without ROSC. The threshold of 10 mmHg predicted successful resuscitation with a sensitivity 100% and a specificity of 77%. Similarly, Callaham and Barton (1990) reported a mean of 19 mmHg for patients with ROSC. Their threshold of 15 mmHg predicted ROSC with a sensitivity of 71% and a specificity of 98%.
In the prehospital studies, the overall mean value of ETCO2 for ROSC was higher than the mean value found by the in-hospital studies. For example, Aspline and White (1995) reported a mean of 23 mmHg for ROSC after 1 min of resuscitation and 13.2 mmHg without ROSC. Cantineau et al. (1996) reported a mean value of 24 mmHg for ROSC and a mean value of 13.8 mmHg without ROSC. The higher values of ETCO2 from prehospital studies may suggest that the early use of capnography during OHCA reflects differences in the timing and the physiologic state of cardiac arrest (Asplin & White, 1995). Since the timing of ETCO2 measurement after cardiac arrest may be a critical factor, establishing a prognostic value may be difficult, especially when the optimal time for ETCO2 measurement is unknown (Aspline and White, 1995). Finding the threshold value can potentially minimize the cost and futile effort for prolonged CPR.
Although ETCO2 monitoring during CPR has a potential ability to predict the outcomes of CPR, there are several limitations. First, ETCO2 measurements may be effected by inconsistent methods of manual ventilation and chest compressions (Steedman & Robertson, 1990). In real life, manual ventilation and chest compressions could cause ETCO2 to fluctuate with the effort of compression and rate of ventilation (Sanders et al., 1989; Steedman & Robertson, 1990). Ideally, the minute ventilation should be controlled during quantitative monitoring of ETCO2 in cardiac arrest since ETCO2 is effected by the alveolar ventilation (Idris et al., 1994).
Second, the effect of the resuscitation medications such as NaHCO3 and epinephrine should be taken into consideration. According to Callaham, Barton and Matthay (1992), ETCO2 could be decreased inconsistently with epinephrine, but the predictive value of ETCO2 was not eliminated during CPR. The exact mechanism of decreasing ETCO2 after epinephrine administration is unknown (Callaham et al., 1992). The effect of higher dose of epinephrine remains questionable and requires further study. Concerning the effect of NaHCO3 on ETCO2, this medication has proven no benefits in cardiac arrest and should be avoided especially when monitoring ETCO2 (Callaham, 1990; Callaham et al., 1992). NaHCO3 is known as a buffer agent, which can transiently increase ETCO2 but the measurements return to perfusion level within 5 min (Trillo et al., 1993).
Third, the studies reported conflicting findings of cardiac rhythms that were associated with ROSC. According to Sanders et al. (1989), the initial cardiac rhythms of VT/VF were associated with ROSC, and asystole/EMD were associated with nonsurvival (no ROSC). However, these findings were inconsistent from Callaham and Barton (1990), who reported that asystole /EMD were significantly associated with the outcome of ROSC. The analysis from prehospital studies did not support these findings. Wayne et al. (1995) only assessed patients with PEA and excluded VF/VT. The rational for the exclusion criterion was not found in the article. In addition, the findings from the study by Cantineau et al. (1996) were applicable to patients with atraumatic cardiac arrest of asystole.
Lastly, atraumatic cardiac arrests can be frequently caused by acute and chronic illness with comorbidities. As discussed previously, when ventilation and perfusion are equal, PaCO2 is equivalent to ETCO2 (LaValle & Perry, 1995). However, conditions such as COPD or ARDS result in a VQ abnormality and high CO2 gradient, which causes PaCO2 to change and make it difficult to assess the accuracy of ETCO2 (LaValle & Perry, 1995; Levin & Pizov, 1997; Santos et al., 1993). The potential implication of ETCO2 monitoring under these circumstances including traumatic cardiac arrest is limited as evidenced by lack of studies looking at the different groups of patients.
Currently, ETCO2 monitoring is widely used for various clinical practices such as verification of endotracheal tube placement, assessment of conscious sedation safety and evaluation of mechanical ventilation (LaValle & Perry, 1995; Sanders, 1989; Santos et al., 1993). ETCO2 monitoring can guide nurses in providing adequate oxygenation and ventilation to unstable patients if capnography is used correctly (LaValle & Perry, 1995). For example, ETCO2 monitoring in conjunction with ABGs are useful for ensuring adequate ventilation for patients with head injuries (Sanders, 1989). A fall in ETCO2 may mean decrease in lung perfusion. If ventilation has not changed, the decrease in ETCO2 may indicate early signs of shock (Sanders, 1989). However, ETCO2 must be interpreted in the context of other information about the patients clinical status.
In the ED, the application of ETCO2 monitoring contributes an additional valuable asset to the clinical practice, especially during resuscitation. One can use the feedback from ETCO2 to change the effort (depth/rate/force) of chest compression during CPR (Lambert et al., 1992). Due to the ability of ETCO2 to detect ROSC, CPR does not need to be interrupted in order to establish whether spontaneous circulation has been restored (Steedman & Robertson, 1990). The use of ETCO2 monitoring in OHCA can potentially increase the rate of ROSC and survival by instituting earlier monitoring that could guide the effectiveness of CPR (Aspline & White, 1995).
In conclusion, the critical role of ETCO2 monitoring during CPR has been demonstrated and confirmed by both experimental and clinical studies. An ETCO2 analyzer is easy to apply and readily available. As a quantitative indicator of the volume of blood flow produced by pericardial compressions, ETCO2 can be a noninvasive method of monitoring the efficacy of ongoing effort and the outcome of CPR (Falk et al., 1988; LaValle & Perry, 1995; Sanders et al., 1989).
Asplin, B. R., & White, R. D. (1995). Prognostic value of end-tidal carbon dioxide pressures during out-of-hospital cardiac arrest. Annals of Emergency Medicine, 25(6), 756-761.
Hollinger, C. B., & Hoyt, J. W. (1995). Capnography and respiratory monitoring. In Ayres, S. M., Grenvik, A., Holbrook, P. R., & Shoemaker W. C. (Eds.), Textbook of Critical Care (pp. 305-310). Philadelphia: W. B. Saunders.
Benumof, J. L. (1998). Interpretation of capnography. Journal of the American Association of Nurse Anesthetists, 66(2), 169-176.
Bonnin, M. J., Pepe, P. E., Kimbal, K. T., & Clark, P. S. Jr. (1993). Distinct criteria for termination of resuscitation in out-of-hospital setting. Journal of American Medical Association, 270(12), 1457-1462.
Cantineau, J. P., Lambert, Y., Merckx, P., Reynaud, P., Porte, F., Bertrand, C., & Duvaldestine, P. (1996). End-tidal carbon dioxide during cardiopulmonary resuscitation in humans presenting mostly with asystole: A predictor of outcome. Critical Care Medicine, 24(5), 791-796.
Callaham, M. (1990). High dose epinephrine and other advances in management of cardiac arrest. Western Journal of Medicine, 152(6),697-703.
Callaham, M. & Barton, C. (1990). Prediction of outcome of cardiopulmonary resuscitation from end-tidal carbon dioxide concentration. Critical Care Medicine, 18(4), 358-362.
Callaham, M., Barton, C., & Matthay, M. (1992). Effect of epinephrine on the ability of end-tidal carbon dioxide readings to predict initial resuscitation from cardiac arrest. Critical Care Medicine, 20 (3), 337-343
Domsky, M., Wilson, R. F., & Heins, J. (1995). Intraoperative end-tidal carbon dioxide values and derived calculations correlated with outcome: Prognosis and capnography. Critical Care Medicine, 23(9), 1497-1503.
Falk, J. L., Rackow, E. C., & Weil, M. H. (1988). End-tidal carbon dioxide concentration during cardiopulmonary resuscitation. The New England Journal of Medicine, 318 (10), 607-611.
Garnett, A. R., Ornato, J. P., Gonzalez, E. R., & Johnson, B. (1987). End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation. Journal of American Medicine Association, 257(4), 512-515.
Gray, W. A., Capone, R. J., Most, A. S. (1991). Unsuccessful medical resuscitation: Are continued efforts in the emergency department justified? New England Journal of Medicine, 325(20), 1393-1398.
Gudipati, C. V., Weil, M. H., Bisera, J., Deshmukh, H. G., & Rackow, E. C. (1988). Expired carbon dioxide: a noninvasive monitor of cardiopulmonary resuscitation. Circulation, 77(1), 234-239.
Higgins, D., Hayes, M., Denman, W., & Wilkinson, D. J. (1990). Effectiveness of using end tidal carbon dioxide concentration to monitor cardiopulmonary resuscitation. British Medical Journal, 300(6724), 581.
Idris, A. H., Staples, E. D., OBrien, D. J., Melker, R. J., Rush, W. J., Del Duca, K. D., & Falk, J. L. (1994). End-tidal carbon dioxide during extremely low cardiac output. Annals of Emergency Medicine, 23(3), 568-572.
Kalenda, Z. (1978). The capnogram as a guide to the efficacy of cardiac massage. Resuscitation, 6, 259-263.
Lambert, Y., Cantineau, J. P. Merckx, P., Bertrand, C., & Duvaldestin, P. (1992). Influence of end-tidal CO2 monitoring on cardiopulmonary resuscitation. Anesthesiology, 77(3A), A1081.
LaValle, T. L., & Perry, A. G. (1995). Capnography: Assessing end-tidal CO2 levels. Dimensions of Critical Care Nursing, 14(2), 70-77.
Levin, P. D. & Pizov, R. (1997). End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest (Letter). The New England Journal of Medicine, 337(23), 1694-1695.
Levine, R. L., Wayne, M. A., & Miller, C. C. (1997). End-tidal carbon dioxide and outcome of out-of-hospital cardiac arrest. The New-England Journal of Medicine, 337(5), 301-306.
Nielsen, M. S. Fitchet, A., & Sanders, D. A. (1990). Monitoring cardiopulmonary resuscitation by end tidal carbon dioxide concentration. British Medical Journal, 300(6730), 1012-1013.
Norman, G. R., & Streiner, D. L. (1997). PDQ Statistics (2nd ed.). St. Louis: Mosby.
Ornato, J. P., Garnett, A. R., & Glauser, F. L. (1990). Relationship between cardiac output and the end-tidal carbon dioxide tension. Annals of Emergency Medicine, 19(10), 1104-1106.
Polit, D. F., & Hungler, B. P. (1995). Nursing research: Principles and methods (5th ed.). Philadelphia: J. B. Lippincott.
Sanders, A. B., Ewy, G. A., Bragg, S., Atlas, M., & Kern, K. B. (1985). Expired PCO2 as a prognostic indicator of successful resuscitation from cardiac arrest. Annals of Emergency Medicine, 14(10), 948-952
Sanders, A. B. (1989). Capnometry in Emergency Medicine. Annals of Emergency Medicine, 18(12), 1287-1290.
Sanders, A. B., Kern, K. B., Otto, C. W., Milander, M. M., & Ewy, G. A.(1989). End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation: A prognostic indicator for survival. Journal of American Medicine Association, 262(10), 1347-1351.
Santos, L. J., Varon, J., Pic-Aluas, L., & Combs, A. H. (1993). Practical uses of end-tidal carbon dioxide monitoring in the emergency department. The Journal of Emergency Medicine, 12(5), 633-644.
Steedman, D. J. & Robertson, C. E. (1990). Measurement of end-tidal carbon dioxide concentration during cardiopulmonary resuscitation. Archives of Emergency Medicine, 7, 129-134.
Trillo, G., von Planta, M., & Kette, F. (1994). ETCO2 monitoring during low flow states: clinical aims and limits. Resuscitation, 27, 1-8.
Wayne, M. A., Levine, R. L., & Miller, C. C. (1995). Use of end-tidal carbon dioxide to predict outcome in prehospital cardiac arrest. Annals of Emergency Medicine, 25(6), 762-766.
Weil, H. M., Rackow, E. C., Trevino, R., Grundler, W., Falk, J. L., & Griffel, M. I. (1986). Difference in acid-base state between venous and arterial blood during cardiopulmonary resuscitation. The New England Journal of Medicine, 301(3), 153-156.
Wilson, H. S. (1989). Research in nursing (2nd ed). California, Addison-Wesley.
"End Tidal Carbon Dioxide Monitoring During CPR: A Predictor of Outcome"
by Jinhee Nguyen, RN MSN [Jinhee.Nguyen@ucsfmedctr.org]
© Jinhee Nguyen, RN MSN May 1999
is presented by Emergency Nursing World ! [http:ENW.org]
Tom Trimble, RN [Tom@ENW.org] ENW Webmaster
ENW name, logos, and layout ©Tom Trimble, RN
Fabulous Places! --
Check frequently for updates
for the Year Ahead!
Meetings & Symposia
|Clinical Research||Behind the Scenes at ENW!||The World of Emergency Nursing|
Thermometry in Acute Care
Latest Research in Resuscitation
Applies to all portions of this
Policies of this Website
Information for Authors
The Emergency Nursing
& Emergency Nursing World ! FAQ
ENW is Listed@
What's This All About?
Tom Trimble's Tale
Library of Resources &
CPEN Review - Putting
It All Together
Bedlam Among the Bedpans
Body Piercing Removal Kit
& Training Program
Emergency Nursing 5-Tier
Emergency Newborn Care
Quick Reference to Triage
The Emergency Nursing "Cool Web-Find!"
Honor Gallery of Previous Winners
An 1895 Look At Nursing
Beatitudes For Leaders
E-Mail Lists & Usenet Groups
Emergency Nursing WebLinks
Em-Nsg-L: The Emergency Nursing List
Night-Shift Survival Tips
Old-Aid -Archaic & Obsolete
University-Level Emergency Nursing Education
Words & Thoughts
|California ENA Website||ENW! Supports ENA|