The Journal of Emergency Medicine, Vol. 36, No. 2, pp. 116 120, 2009 Copyright 2009 Elsevier Inc.

. Printed in the USA. All rights reserved 0736-4679/09 $see front matter

doi:10.1016/j.jemermed.2007.10.022

Original Contributions

THE PRESENCE OF AN EMERGENCY AIRWAY RESPONSE TEAM AND ITS EFFECTS ON IN-HOSPITAL CODE BLUE
Sean O. Henderson,
MD,*

Christian D. McClung, MD, MPHIL and Stuart P. Swadron, MD*

(CANTAB),*

Chanida Sintuu,

MD,*

*Department of Emergency Medicine and Department of Preventive Medicine, Keck/USC School of Medicine, Los Angeles, California USC Medical Center, 1200 N. State Street Reprint Address: Sean O. Henderson, MD, Department of Emergency Medicine, LAC Room 1011, Los Angeles, CA 90033

e AbstractThe survival rate from in-hospital cardiac arrest due to pulseless electrical activity (PEA)/asystole in our institution was higher than expected (70%). It was the impression of the Emergency Department-led Code Blue Team (CBT) that many of these patients were actually suffering respiratory arrests before their cardiac events. To address this, the facility developed an early intervention team focused on early airway interventionthe Emergency Airway Response Team (EART). The objective of this study was to assess the effect of early intervention in patients during the pre-Code Blue period, specically with regard to airway stabilization. Our hypothesis was that there would be fewer CBT calls (cardiac arrests) due to PEA and asystole and that the survival from these events would decrease. This was a retrospective review of all cardiac arrests responded to by the CBT and EART for a period of 2 years. Charts were reviewed for the initial presenting rhythm (as dened by the Utstein Format) and event survival for the 12-month period immediately before and immediately after the establishment of the EART (Time Periods 1 and 2, respectively). The total number of CBT calls decreased by 15%, return of spontaneous circulation from any rhythm decreased by 9%, and survival to discharge decreased by 8% (p non-signicant). The number of CBT calls specically for asystole/PEA decreased by 8%. Deaths in hospital were signicantly associated with Period 2 (odds ratio 1.84; 95% condence interval 1.033.28) after adjusting for age, gender, and presenting rhythms. The total number of CBT calls decreased slightly with the creation of the Emergency Airway Response Team. Return of spontaneous circu-

lation and survival to hospital discharge after cardiac arrest due to asystole/PEA were signicantly decreased, suggesting early intervention may have benet. 2009 Elsevier Inc. e Keywordsrespiratory arrest; resuscitation; cardiac arrest

INTRODUCTION In an effort to strengthen the chain of survival in our institution, we formalized an emergency physician-led Code Blue Response Team (CBRT) in 2000 (1). Although survival increased in every presenting rhythm category, the biggest change occurred in the pulseless electrical activity (PEA)/asystole group, with a 30% absolute increase in return of spontaneous circulation (ROSC). With rates of ROSC after PEA/asystole arrest so much higher than recently published series, we were led to believe that some of these cases were not typical PEA/asystole arrest events (2,3). It was our hypothesis that a proportion of patients who were successfully resuscitated from PEA/asystolic arrest were actually undergoing a primary respiratory event prior that led to the arrest. In fact, multiple authors have documented the presence of pulmonary symptoms before the onset of in-hospital cardiac arrest (4,5). We further hypothesized that early intervention in these patients would prevent their deterioration to cardiac arrest and

RECEIVED: 16 November 2006; FINAL ACCEPTED: 8 October 2007

SUBMISSION RECEIVED:

5 October 2007;

116

Emergency Airway Response Team

117

thus indirectly decrease the survival rate of the remaining group with PEA/asystolic arrest to rates more consistent with previous series. Based on these data and this hypothesis, we established an Early Airway Intervention Team (EART) to assess and provide urgent airway intervention for patients in the precode time period. We report here on the effect of the EART on our in-hospital cardiac arrest rate and survival.

and classies the initial rhythm as either PEA/asystole, bradycardia/perfusing rhythm, or ventricular brillation/ tachycardia (VF/VT) (6). These data are abstracted and entered into a spreadsheet for quality improvement purposes. False arrests and patients who are designated as do not resuscitate/do not intubate are also included in the database but were excluded from our analysis.

METHODS This study was approved by the Institutional Review Board of the Health Sciences campus of the University of Southern California. It was a before and after study that involved a retrospective review of all CBT and EART activations in our facility, a Level One Trauma Center with an annual census of 141,000 adult patients per year. The average number of inpatient beds during the period under study was 750. The CBT in place at this facility has been described previously (1). Briey, before July 1997, cardiopulmonary events were covered on an individual basis by the patients primary physician and service (e.g., internal medicine post-graduate year 2 or 3, or general surgery post-graduate year 3, 4, or 5). Nursing responders were drawn from available oor staff. No formalized training was required of the team members, and individuals were summoned to the event location by overhead paging. Beginning on July 15, 1997, a CBT with an emergency physician team leader (post-graduate year 4 or attending staff level) and critical care nurses was established. The team also included an assigned scribe who completed a standardized data collection/ quality improvement tool organized in the Utstein format. Communication with the CBRT was via a phone extension and dedicated pagers for team members. Daily checks of the system and team member availability occurred at 9:00 a.m. each day. Due to the conguration of the facility, the CBT was not responsible for pediatric arrests, the Emergency Department (ED), or the intensive care units. On July 1, 2004, the institution formalized an EART, consisting of the team leader of the CBT and a respiratory therapist. In the event that a physician or nurse determined that a patient needed an assessment for an airway intervention, a dedicated telephone number was accessed and the EART members were notied. Criteria for activation of the EART were kept deliberately vague to allow staff the freedom to call for help whenever they believed they needed it. Patients were seen, evaluated, and treated at the discretion of the EART physician. Documentation for both the EART and CBT is accomplished on a standardized form that follows the Utstein guidelines for the reporting of in-hospital cardiac arrest

Statistical Analysis Data from the year before the formation of the EART (Period 1) were compared to that after the formation of the EART (Period 2). Analysis was divided between Period 1 and Period 2 and excludes the airway cases identied during Period 2. However, the patients with airway as the reason for calling the code are listed for demographic comparison. There were 7 patients who were called as an airway code and then later called as a Code Blue. Data analysis was performed using STATA 8.0 software (StataCorp LP, College Station, TX). Descriptive analysis using the chi-squared test was used to evaluate demographic variables. The primary outcome in this study is survival to hospital discharge; whereas return of spontaneous circulation was a secondary outcome. Logistic regression analysis was used to study associations with primary and secondary outcomes. Dummy variables for rhythm were used to control for this variable with the most prevalent rhythm, bradycardia /perfusing, used for baseline comparisons. The multiple logistic regression predicts death in hospital and nonreturn of spontaneous circulation so that the observed odds ratios would be easier to interpret.

RESULTS There were 363 total events documented during the 2-year study period. Thirty-ve of these were false arrests and do not resuscitate/do not intubate events that were excluded from further analysis. Two further events were removed due to the amount of missing data. The nal cohort, therefore, included 273 patients with 344 unique events (a maximum of four Code Blues were observed on a single patient). Table 1 lists the demographic description of our study population and nal outcome. The majority of included patients were over 50 years old, male, and Hispanic. There were no signicant differences between the study groups when comparing proportions using a chi-squared statistic. Airway cases did have improved rates of ROSC and survival to hospital discharge when compared to either Period 1 or Period 2 groups (p 0.001).

118
Table 1. Demographic Prole Pre-Airway Code Team 2003 (n 157) Variable Age 1829 years 3049 years 5069 years 70 years Total Gender Male Female Total Ethnicity Asian Black Hispanic White Other/unknown Total Initial Rhythm Asystole/PEA Bradycardia/perfusion VF/VT Total ROSC Yes No Total Survival to discharge Yes No Total n 14 51 56 32 153 110 43 153 12 29 74 32 6 153 66 79 12 157 109 48 157 49 101 150 % 9 33 37 21 72 28 8 19 48 21 4 42 50 8 69 31 33 67 Post-Airway Code Team 2004 (n 133) n 10 39 60 23 132 94 38 134 20 26 60 25 1 134 45 77 11 133 80 53 133 33 98 131 % 8 30 45 18 NS 71 29 NS 15 20 45 19 1 NS 33 59 8 NS 60 40 NS 25 75 NS 54 0 54 33 21 54 p-Value*

S. O. Henderson et al.

Airway Cases 2004 (n 54) n 1 19 28 6 54 36 18 54 4 11 26 13 0 54 N/A N/A N/A % 2 35 52 11 NS 67 33 NS 7 20 48 24 0 NS N/A N/A N/A 100 0 0.001 61 39 0.001 p-Value

* p-Value comparison to pre-airway group, (chi-squared test); NS non-signicant. p-Value comparison of airway cases to pre-EART cases. PEA pulseless electrical activity; VF/VT ventricular brillation/tachycardia; ROSC

return of spontaneous circulation.

Table 2 lists the odds ratios of a univariate analysis with the outcome of death in hospital for all events in the 2-year study period. Increasing age was associated with death in hospital. When comparing arrest rhythms to a baseline of bradycardia/perfusing, patients with asystole/ PEA rhythms had signicant associations with death in hospital, nearly sixfold. In addition, patients with VF/VT also had signicant associations with death in hospital. Table 3 lists the odds ratios for both outcomes of death in hospital and non-return of spontaneous circulation by Period 1 and Period 2 groups (excluding airway cases). Overall, the Period 2 group was associated with death in the hospital and non-ROSC when controlling for the initial arrest rhythms, age, and gender.

DISCUSSION We noted in our original article that our event survival (ROSC) rate of 58% was better than that reported by

others previously, especially in those patients with a presenting rhythm of PEA/asystole, in whom it was 48% (1). Huang et al. reported in-hospital survival from asystole of 17% and from PEA of 19%, whereas the large National Registry of Cardiopulmonary Resuscitation described an overall ROSC rate of 44% and a 35% rate of ROSC for asystole (2,3). Our survival-to-discharge rate was also higher than that reported by other authors. Cohn et al. reported a survival-to-discharge rate of 21% during the same time period that our survival to discharge from cardiac arrest events was as high as 36% (7). The National Registry of Cardiopulmonary Resuscitation reported a nationwide average of 17% for survival to hospital discharge (3). Once a signicant number of arrests due to respiratory compromise were removed, our survival rates for cardiac arrest, particularly in those with a presenting rhythm of asystole/PEA, and those decreased are more comparable to others. Survival to discharge also became more similar, at 25%. These changes point to the importance of the

Emergency Airway Response Team

Table 2. Univariate Analysis of Associations with Outcome of Death in Hospital 95% Odds Condence Ratio Interval 1.00 2.22 2.83 3.23 1.39 1.00 1.59 1.00 0.59 0.55 0.72 0.23 1.00 5.80 4.97 N/A 0.875.67 1.137.13 1.129.29 1.031.87 N/A 0.902.77 N/A 0.251.36 0.271.10 0.361.45 0.041.11 N/A 2.9911.26 1.4117.53 ventricular

119

Outcome

Control Variable

Death in hospital Age 1829 years 3049 years 5069 years 70 years Trend Gender Female Male Ethnicity Hispanic Asian Black White Other/unknown Initial rhythm Bradycardia/perfusing Asystole/PEA VF/VT

PEA pulseless electrical activity; VF/VT brillation/tachycardia.

early recognition of deteriorating patients and prompt intervention and resuscitation (5). It has been previously reported that early intervention in a deteriorating patient is associated with a decreased risk of cardiac arrest and avoidable death (8). If identied early enough, action may be taken to prevent further physiological deterioration and improve outcome. Recently, in-hospital evaluation and resuscitation teams, labeled as Medical Response teams, Emergency Teams, or Patient-at-risk teams, have been proposed (9,10). These teams are created with the belief that early identication and intervention in patients destined to become unstable will result in fewer adverse outcomes

and improved outcomes, and usually consist of medical or surgical staff physicians and nurses (9,10). At many institutions, the additional workload and time spent outside of the ED may preclude the involvement of ED personnel in such teams. At our institution, however, it has proven feasible, and has beneted not only the patients involved, but also the trainees in our Emergency Medicine residency program. Participation in both the CBRT and EART has increased the exposure of the Emergency Medicine residents to cardiac and airway arrest events, particularly in settings that are less controlled and equipped for such events (e.g., general hospital wards). We believe that this is an important aspect of their training, as emergency physicians in many community hospitals often will be called upon to manage such events throughout the hospital while on duty. The responsibility of serving as captain of the CBRT and EART lies with a senior resident (post-graduate year 3 or 4) on duty in the ED. Because we have several residents working in the clinical areas of our department at all times, we assign the team leader responsibility to a resident who is working in a lower acuity area, one where patients have already been screened by a triage process. Although team activations pull one resident from their patient care responsibilities in the ED for brief periods, we have enough physician personnel remaining in the department, both resident and attending staff, to maintain adequate coverage at all times. The program has been extremely well received by the residents, who upon graduation have consistently rated their experience directing resuscitations as one of the strongest elements of their training. Limitations One limitation to this study is the possible misclassication of the patients rhythm, specically the misdiag-

Table 3. Multiple Logistic Regression of Death in Hospital, and Non-Return of Spontaneous Circulation, Controlling for Arrest Rhythm, Age, and Gender Outcome Death in hospital Control Variable Post-airway code team Rhythm (vs. bradycardia/perfusing) Asystole/PEA VF/VT Age (each additional year) Gender (male vs. female) Post-airway code team Rhythm (vs. bradycardia/perfusing) Asystole/PEA VF/VT Age (each additional year) Gender (male vs. female) ventricular brillation/tachycardia; ROSC Odds Ratio 1.84 6.23 5.94 1.02 1.92 1.79 3.74 2.04 1.01 1.59 95% Condence Interval 1.033.28 3.1412.4 1.6321.72 1.001.03 1.043.57 1.063.03 2.166.48 0.785.38 0.991.02 0.882.89

Non-ROSC

PEA

pulseless electrical activity; VF/VT

return of spontaneous circulation.

120

S. O. Henderson et al.

nosis of PEA/asystole. The diagnosis of PEA was initially made by the rst responder and later conrmed by the CBT. Misclassication of these patients (i.e., mistaking a severely hypotensive patient as PEA instead of as perfusing rhythm) would not necessarily have altered our ndings, as the process of rhythm identication was the same during both study periods. The observed drop in survival could be the result of chance with a relatively small sample size. However, the statistical analysis with an alpha level of 0.05 suggests that this is unlikely. Given the magnitude of the overall effect, odds ratio of 1.84 (95% condence interval 1.03 3.28), the drop in survival to hospital discharge was likely real. One way to strengthen the study would have been to randomize the intervention by day or week; however, ethical considerations precluded us from using this type of design. We felt that a before and after study was thus the most appropriate way to examine the effect of an early intervention team on patient outcomes. Finally, it should be reiterated that there were no standard criteria for EART activation, although this would not have been expected to affect the outcome of the Code Blue events we are reporting here.

not decrease signicantly, the overall odds ratio of ROSC and survival to hospital discharge did. We believe that this decrease in survival is due, in large part, to the removal of preventable hypoxia-induced events from the cardiac arrest pool.

REFERENCES
1. Henderson SO, Ballesteros D. Evaluation of a hospital-wide resuscitation team: does it increase survival for in-hospital cardiopulmonary arrest? Resuscitation 2001;48:111 6. 2. Huang CH, Chen WJ, Ma MHM, Chang WT, Lai CL, Lee YT. Ractors inuencing the outcomes after hospital resuscitation in Taiwan. Resuscitation 2002;53:26570. 3. Peberdy MA, Kaye W, Ornato JP, et al. Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation 2003;58:297308. 4. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest 1990;98: 1388 92. 5. Franklin C, Mathew J. Developing strategies to prevent inhospital cardiac arrest: analyzing responses of physicians and nurses in hours before the event. Crit Care Med 1994;22:244 7. 6. Cummins RO, Chamberlain D, Hazinski MF, et al. Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: the in-hospital Utstein style. American Heart Association. Ann Emerg Med 1997;29:650 79. 7. Cohn AC, Wilson WM, Yan B, et al. Analysis of clinical outcomes following in-hospital adult cardiac arrest. Intern Med J 2004;34: 398 402. 8. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modied Early Warning Score in medical admissions. QJM 2001; 94:521 6. 9. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ 2002;324:38790. 10. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia 1999;54:853 60.

CONCLUSION Cardiac arrests occurring in the in-hospital population may be preceded by warning signs that, if heeded, may lead to earlier intervention and lower rates of morbidity and mortality. With the advent of an Emergency Medicine-led, early intervention response team at our facility, although the number of cardiac arrest calls did