Cumulative Epinephrine Dose during Cardiopulmonary Resuscitation and Neurologic Outcome

Methods

From 1 August 1991 to 31 December 1995, data were collected on patients admitted to the Department of Emergency Medicine at the University Hospital of Vienna, Austria, after cardiac arrest that occurred before or during hospitalization. Collection was done according to the recommended guidelines for uniform reporting of data on out-of-hospital cardiac arrest [27]. The procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional or regional) and with the Helsinki Declaration of 1975, as revised in 1983.

Patients

Included patients had witnessed, nontraumatic, normothermic cardiac arrest of presumed cardiac origin. To exclude the influence of the initial rhythm on outcome [28-30], we focused on patients whose initial rhythm was ventricular fibrillation. Cardiac arrest was defined as the absence of both spontaneous respiration and palpable pulses. Patients were excluded if spontaneous circulation returned within 3 minutes after the initial three countershocks and if they had an additional cardiac arrest within 6 months after the first. We also excluded patients with known unfavorable overall performance or cerebral performance (New York Heart Association class 3 or 4 or a cerebral performance category [CPC] of 3 or 4) before cardiac arrest, a factor that is known to affect outcome [31-33]. Return of spontaneous circulation was defined as any, even transient (nonsustained), return of a palpable arterial pulse [27]. Acute care consisted of basic and advanced cardiac life support performed by the Vienna Ambulance Service or in-hospital emergency medical technicians and physicians according to standard protocol [2].

The interval from the time of collapse (the presumed time of cardiac arrest) to basic or advanced life support was defined as the “no-flow duration,” and the interval from the beginning of life support until the return of spontaneous circulation or termination of resuscitative efforts was called the “low-flow duration.” The cumulative epinephrine dose was defined as the overall dose of epinephrine subsequently administered during advanced cardiac life support. One physician was responsible for obtaining data through interviews with the ambulance physicians, paramedics, bystanders, and families and from computerized patient records.

Outcome Measures

Cerebral function, expressed in terms of CPC, was assessed prospectively on arrival at the emergency department and 6 hours, 12 hours, 24 hours, 2 days, 3 days, 1 week, 1 month, and 6 months after return of spontaneous circulation [27, 34]. The CPC category is based on the Glasgow outcome performance categories [35]. The performance categories are defined as follows: CPC 1, conscious and alert with normal function or only slight disability; CPC 2, conscious and alert with moderate disability; CPC 3, conscious with severe disability; CPC 4, comatose or in a persistent vegetative state; and CPC 5, certifiably brain dead or dead by traditional criteria. The best CPC score achieved within 6 months was used for calculation. A CPC score of 1 or 2 represents favorable functional neurologic recovery because patients with these scores have sufficient cerebral function for independent activities of daily living. A CPC score of 3, 4, or 5 reflects unfavorable functional neurologic recovery. The investigator assessing the CPC score by examining the patients according to a neuropsychiatric protocol was blinded to the resuscitation data.

Statistical Analysis

Data are given as the median and interquartile range (the difference between the 25th and 75th percentiles) unless otherwise specified. We used the Kruskal-Wallis test and the Mann-Whitney U test to compare continuous variables and used the chi-square test or the Fisher exact test to compare proportions. We calculated Spearman rank correlation coefficients to assess the relation between the cumulative epinephrine dose, the no-flow and low-flow durations, and the sum of these durations. Logistic regression analysis was used to assess the association between neurologic recovery and cumulative epinephrine dose. To control for potential cofounders, we included the following variables in the logistic regression model: age, sex, body mass index, bystander-administered basic life support (yes or no), site of cardiac arrest (in the hospital or outside of the hospital), no-flow duration, and low-flow duration. The logistic regression model was tested for collinearity of the independent variables by using standard procedures. No substantial collinearity between the independent variables was found. Logistic regression coefficients were converted to odds ratios with 95% CIs. All data were computed by using SPS for Windows, release 6.0 (SPS, Inc., Chicago, Illinois). A P value less than 0.05 was considered statistically significant.

Results

Within the 65-month study period, 698 patients who were having or had had cardiac arrest were admitted to the emergency department. Of these, 178 fulfilled the inclusion criteria and were enrolled. The median age of the 178 patients was 61 years (range, 50 to 71 years), and 148 patients (83%) were male. In 160 patients (90%), cardiac arrest occurred outside of the hospital. The estimated median no-flow duration was 2 minutes (interquartile range, <1 to 10 minutes), and the median low-flow duration was 16 minutes (interquartile range, 7 to 34 minutes). The median no-flow duration can be attributed to the 57 patients (32%) who underwent bystander-administered basic life support and to the 32 patients (18%) whose cardiac arrest was witnessed by emergency medical service personnel. The median no-flow duration in the patients who had bystander-administered basic life support was 1 minute (interquartile range, <1 to 2.5 minutes). This duration can be explained by the fact that in 35 patients, basic life support was initiated immediately after arrest. If arrest is witnessed by emergency medical service personnel, the no-flow duration can be expected to be 0 minutes.

In all 32 patients whose cardiac arrest was witnessed by emergency medical service personnel, spontaneous circulation could be restored. Twenty-four of these patients (75%) showed favorable functional neurologic recovery. Of 119 patients whose cardiac arrest was not witnessed by emergency medical service personnel, 39 (33%) had favorable functional neurologic recovery. The median cumulative epinephrine dose administered intravenously during cardiopulmonary resuscitation was 4 mg (interquartile range, 1 to 6 mg). The median epinephrine dose administered every fifth minute was 1.2 mg (interquartile range, 0.6 to 2.1 mg).

Patients with and without Return of Spontaneous Circulation

In 151 patients (84%), spontaneous circulation was restored. Compared with patients in whom spontaneous circulation was not restored, patients with restoration of spontaneous circulation had a similar no-flow duration (2 minutes [interquartile range, <1 to 9 minutes] and 1 minute [interquartile range, <1 to 10 minutes]; P > 0.2) and a shorter low-flow duration (13 minutes [interquartile range, 5 to 24 minutes] and 80 minutes [interquartile range, 58 to 119 minutes]; P < 0.001) and received a lower cumulative dose of epinephrine (3 mg [interquartile range, 1 to 6 mg] and 8 mg [interquartile range, 4 to 18 mg]; P < 0.001). Restoration of spontaneous circulation was possible with increasing cumulative doses of epinephrine, but good functional neurologic recovery was less likely (Figure 1 and Figure 2).

Patients with Favorable and Unfavorable Functional Neurologic Recovery after Return of Spontaneous Circulation

Of 151 patients with return of spontaneous circulation, neurologic recovery was favorable in 63 (42%). Within a 6-month observation period, 80 patients died (median duration of survival, 4 days [interquartile range, 0.5 to 14 days; range, 0.4 hours to 111 days]). Of the 80 patients who died within the 6-month observation period, 77 had unfavorable functional neurologic recovery (median duration of survival, 4 days [interquartile range, 0.5 to 13 days]). Durations of survival in the 3 patients with favorable functional neurologic recovery were 19, 30, and 44 days, respectively. Six months after restoration of spontaneous circulation, 60 patients were alive with favorable neurologic recovery and 11 patients were alive with unfavorable neurologic recovery.

Table 1 shows patient characteristics according to outcome category. Compared with patients who had poor functional neurologic recovery, patients with favorable recovery had a shorter no-flow duration, had a shorter low-flow duration, and received a lower cumulative epinephrine dose. Patients with favorable recovery were also more likely to have had their cardiac arrest in the hospital, had a higher body mass index, and were younger. Patients with bystander-administered basic life support had a significantly shorter no-flow duration than patients without this intervention (median, 1 minute [interquartile range, 0 to 2 minutes] compared with 7 minutes [interquartile range, 0 to 10 minutes]; P < 0.001). For all patients whose spontaneous circulation returned, the correlations between the cumulative epinephrine dose and the no-flow and low-flow durations were r = 0.40 (P < 0.001) and r = 0.57 (P < 0.001), respectively. The correlation between the cumulative epinephrine dose and the sum of the no-flow and low-flow durations was r = 0.65 (P < 0.001).

Stratified Analysis

Because the low-flow duration is presumably the time of the cardiopulmonary resuscitation attempt at which epinephrine was administered, our study sample was stratified into four subgroups according to low-flow duration (Table 2). In all four subgroups, patients with unfavorable neurologic recovery received a significantly higher cumulative dose of epinephrine than patients with favorable neurologic recovery.

Figure 3 shows the frequencies of good and bad neurologic outcome after simultaneous stratification of the sample by no-flow duration, low-flow duration, and cumulative epinephrine dose.

Multivariate Analysis

Multivariate logistic regression analysis with age, sex, body mass index, bystander-administered basic life support (yes or no), site of cardiac arrest (in or outside of the hospital), no-flow duration, and low-flow duration as independent variables showed that an increasing cumulative dose of epinephrine was significantly and independently associated with increased risk for unfavorable neurologic recovery (odds ratio, 1.22 [95% CI, 1.01 to 1.46]; P = 0.03). Other variables associated with unfavorable neurologic recovery were age, no-flow duration, low-flow duration, and cardiac arrest occurring outside of the hospital. Sex, bystander-administered basic life support, and body mass index were not independently associated with neurologic outcome (Table 3).

Discussion

Our findings indicate that an increasing cumulative dose of epinephrine administered during advanced cardiac life support is independently associated with unfavorable neurologic function after ventricular fibrillation cardiac arrest.

According to “Standards and Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiac Care” [2], the epinephrine dose recommended for patients in cardiac arrest is 1 mg given intravenously every 3 to 5 minutes. This dose has been classified as “a therapeutic option that is usually indicated, always acceptable, and considered useful and effective” [2]. This recommendation is based on laboratory studies [3-5] done more than 30 years ago. The possibility of using a higher epinephrine dose is classified as “a therapeutic option which is not well established by evidence but may be helpful and probably is not harmful” [2]. This recommendation is based on both animal [37-41] and human [13-17] studies that focused primarily on the effect of higher epinephrine doses on myocardial blood flow and the rate of spontaneous circulation achieved. In humans with cardiac arrest, however, high epinephrine doses did not improve outcome compared with standard doses [6-8, 42].

The standard epinephrine dose has been a subject of debate [6-8, 42], but the limit of subsequent doses of epinephrine every 3 to 5 minutes without assessment of a limit has not been questioned to date. We therefore examined whether there is an association between the cumulative epinephrine dose administered during cardiopulmonary resuscitation and neurologic outcome.

In our patients, no upper limit was found for the cumulative epinephrine dose to restore spontaneous circulation, but none of the patients who received a cumulative dose of more than 13 mg achieved good functional neurologic recovery. Of 44 patients whose cumulative epinephrine dose was more than 6 mg, 36 achieved return of spontaneous circulation; only 2 patients, however, left the hospital without severe functional neurologic impairment. It is obvious that the restoration of spontaneous circulation without consideration of neurologic recovery should not be a goal of resuscitation. However, our data can only cautiously address the question of whether there should be a limit for subsequent epinephrine doses in advanced cardiac life support.

We noted an adverse effect of epinephrine that was independent of the no-flow and low-flow durations. These durations of hypoperfusion are intuitively the major determinant of neurologic outcome. Our findings could be explained by the direct detrimental effects of epinephrine during cardiac arrest and in the postresuscitation period. Possible explanations include increases in myocardial oxygen consumption [20-22] and increases in pulmonary arteriovenous admixture [23]. Furthermore, epinephrine has been shown to increase the severity of myocardial dysfunction after resuscitation [24] and to create a dose-dependent impairment of systemic hemodynamic transport and use of oxygen; this is reflected by a lower cardiac index and increased lactic acid levels [19, 25]. Unfavorable neurologic outcome may be the result of these impairments.

It can be argued that restoration of spontaneous circulation should be the primary goal because without a working heart, the brain cannot be perfused. The counterargument is that patients may ultimately have beating hearts but nonfunctioning brains; the subsequent deaths of these patients are usually preceded by extensive, expensive medical care. These factors underline the significance of an epinephrine dose limit, which may serve as a marker for terminating resuscitative efforts because doses that exceed a certain limit were associated with poor neurologic outcome (Figure 2).

Other independent predictors of neurologic outcome were the no-flow and low-flow durations. The no-flow duration is a well-known determinant of outcome [43-46]. Because emergency medical services witnessed cardiac arrests in 18% of patients and because basic life support was administered to 32% of patients, a short median no-flow duration was seen (Table 1). Nevertheless, bystander initiation of cardiopulmonary resuscitation was not found to be significantly related to outcome, in contrast to earlier reports [46-49]. The possible poor quality of bystander-administered cardiopulmonary resuscitation may explain these observations. On the other hand, our data show that basic life support given by bystanders significantly reduced the no-flow duration and therefore may indirectly improve outcome. Our findings on age as an independent predictor for neurologic outcome contrast with results of a previously published study, which showed that increasing age was not a prognostic factor for the quality of neurologic outcome [45].

The retrospective design of our study has several limitations. Difficulties in conducting clinical studies in patients who have had cardiac arrest (especially difficulties in acquiring comparable groups) are numerous and well recognized [27]. By using an internationally recognized protocol [27], we eliminated some handicaps. The calculation of the no-flow duration is an estimate and cannot be quantitated accurately. However, because only one person was responsible for collecting patient data according to published guidelines [27], we suspect that our findings are accurate. Personal interviews of the witnesses probably minimized imprecision about the exact time of recognition of collapse and emergency medical attention. Another important issue is whether administration of epinephrine is only a surrogate for some other causal determinant of unfavorable neurologic recovery that was not included in our analyses. Further investigations of prospectively collected data are needed to confirm our findings.

In patients with witnessed, nontraumatic, normothermic ventricular fibrillation cardiac arrest, the cumulative epinephrine dose administered during resuscitation was associated with neurologic outcome after restoration of spontaneous circulation. Survival with favorable functional neurologic recovery seems to be unlikely after administration of high doses of epinephrine. Further investigation should attempt to better define limits for epinephrine doses during cardiopulmonary resuscitation.