Journal Logo

Empirical Investigations

In Situ Simulation Comparing In-Hospital First Responder Sudden Cardiac Arrest Resuscitation Using Semiautomated Defibrillators and Automated External Defibrillators

Kobayashi, Leo MD; Dunbar-Viveiros, Jennifer A. RN, MS; Sheahan, Bethany A. RN, MS, CCRN; Rezendes, Megan H. RN, MSN, AOCN; Devine, Jeffrey RN, NREMT-P; Cooper, Mary R. MD, JD; Martin, Peggy B. MEd; Jay, Gregory D. MD, PhD

Author Information
Simulation in Healthcare: The Journal of the Society for Simulation in Healthcare: April 2010 - Volume 5 - Issue 2 - p 82-90
doi: 10.1097/SIH.0b013e3181ccd75c
  • Free

Abstract

One third of in-hospital sudden cardiac arrest (SCA) cases are reported as not being appropriately defibrillated in the recommended time, ie, within 3 minutes of arrest.1 Multifaceted approaches are needed to promote the knowledge, skills, cultural, and systems factors essential for optimal SCA resuscitation. A strategy to simplify the in-hospital SCA response model and empower its first responders is suggested by the improvements in out-of-hospital SCA survival achieved with automated external defibrillators (AEDs).2–6

Existing hospital SCA response structures generally use manual and semiautomated defibrillators (SDs) that are either fully operator dependent or equipped with optional rhythm analytic guidance. Although these devices allow for sophisticated monitoring, arrhythmia interpretation, and electrical therapies, their complexity often impedes rapid defibrillation when SCA first responders are noncritical care personnel. The adoption and success of AEDs in public venues may indicate viability of an alternative in-hospital SCA response model7–9 in which AEDs serve as a first defibrillation device that supplements SDs.

Investigators versed in acute care and medical simulation designed and conducted the controlled nonblinded Arrhythmia Simulation/Cardiac Event Nursing Training-AED phase (ASCENT-AED) study examining defibrillation performance through simulated medical scenarios. Study objectives were to determine (1) the feasibility of using simulation to objectively evaluate specific aspects of in-hospital SCA resuscitation; (2) baseline hospital registered nurse (RN) performance during medical/surgical floor SCA events with existing protocols and equipment; and (3) comparative performance of a response model incorporating AEDs. The primary study hypothesis was that AED availability on a hospital floor would lead to faster and increased defibrillations with fewer errors and adverse events by first-responding RNs during simulated SCAs. The study protocol was formulated to generate recommendations to institutional leadership as to whether an AED-supplemented resuscitation system would be equivalent or better than the existing SCA response model for achieving rapid, indicated, and safe defibrillation therapy.

METHODS

Setting and Sample

The ASCENT-AED project was conducted at Rhode Island Hospital, an academic 719-bed regional referral hospital and Level 1 trauma center in Providence, RI. Program development and implementation were organized by personnel from the Rhode Island Hospital Medical Simulation Center, Lifespan Risk Management, the Center for Practice Excellence (nursing education), and the Department of Emergency Medicine at the Alpert Medical School of Brown University. Project approval was obtained from Lifespan Institutional Review Board and Rhode Island Hospital Cardiac Arrest Committee. See Figure 1 for study design.

Figure 1.
Figure 1.:
ASCENT-AED study design. The controlled nonblinded study enrolled subjects into either control group [standard, semiautomated defibrillator (SD)-only response model] or experimental group (AED-supplemented response model) for observation of their clinical performance during two arrhythmia resuscitation scenarios.

The ASCENT-AED team developed two on-site intermediate- fidelity simulation scenarios for a Laerdal Resusci-Anne Simulator manikin (Laerdal, Wappingers Falls, NY) for nursing first responders. Scenarios featured two clinical in-hospital patient presentations: respiratory arrest and sinus bradycardia with pulse in an elderly dehydrated patient, and ventricular fibrillation (VF) arrest in an elderly patient with underlying cardiac disease. These patient conditions were selected to evaluate SCA first responder adherence to American Heart Association resuscitation algorithms based on (1) patient responsiveness/level of consciousness, (2) presence/absence of signs of circulation, and (3) presence/absence of arrhythmia requiring defibrillation.

An accompanying observation tool was developed for objective measurement of the defibrillation aspects of in-hospital first responder actions for select cardiac arrhythmias (Fig. 2). The quality of certain resuscitative activities [eg, advanced airway management, cardiopulmonary resuscitation (CPR), and team communications] was not specifically assessed by the study protocol. Literature review and expert panel consultation were completed to construct the tool for optimal recording of clinical performance and timeline data.

Figure 2.
Figure 2.:
ASCENT-AED study tool.

Because of practical concerns, including the need to ensure bedside relevance and stakeholder acceptance of study (face) validity, the experiment protocol specified in situ simulation for study environment realism (eg, accuracy of floor dimensions, code cart location relative to simulated patient, and hallway obstructions). One medical/surgical unit was chosen as the control floor, whereas an equivalent unit on an adjacent level served as the experimental floor. The units featured identical floorplans and room layouts, patient acuity and census levels, nursing assignment patterns, and ancillary support staffing. During ASCENT-AED sessions, a study equipment code cart was placed at each unit's central nursing station. In addition to a bag-valve-mask (BVM) assembly, the study cart housed either a live SD (Zoll M Series, Zoll, Chelmsford MA) when on the control floor, or both live SD and live AED (Zoll AED Plus, Zoll, Chelmsford, MA) when on the experimental floor. The study did not use nonfunctional “trainer” AEDs or defibrillators to avoid the risk of study devices being used during an actual resuscitation.

Subjects were recruited from on-duty hospital clinical nursing staff who are expected to respond to SCA resuscitations on medical floor units (excluding intensive care units and operating rooms) as part of their clinical responsibilities. Only nursing personnel were enrolled in the study, in accordance with current quality management data identifying nurses as the most common in-hospital SCA first responders (data not shown). Nurses at the study institution are authorized to defibrillate on SD and AED devices in the absence of a physician; therefore, the scenarios situated the subjects as independent and solitary first responders.

Data Collection

Demographic information and clinical experience data, including data on prior use of manual/SDs and AEDs as the primary operator on actual patients, were collected from the subjects. A scripted introduction-oriented subjects to the study setting, manikin features and capabilities, and defibrillator device(s) available. The only equipment instructions given to subjects immediately before scenarios concerned the AED “on” switch, and the method of applying modified defibrillation pads to the study manikin. Subjects did not receive any resuscitation training, because medical/surgical floor nurses are expected to have SD and AED training as part of their employment requirements.

All subjects participated in the two simulation scenarios in randomized order for a total of 100 data points. Each simulated patient was presented to subjects as newly arrived from the emergency department, admitted to the study floor, and requiring basic nursing care while awaiting physician orders. Immediately after the scripted introduction, undisclosed simulated SCA was started for calibrated measurements of subsequent response times. Primary study metrics focused on the following critical SCA resuscitation interventions: recognition of unresponsiveness, recognition of apnea, recognition of presence or absence of pulse, call for help; retrieval of resuscitation cart and defibrillation device(s), BVM ventilation, chest compressions; application of defibrillation pads, activation of SD (device and analytic function) or AED, broadcast warning of electrical therapy, and defibrillation. Subjects were encouraged to verbalize their findings and decision-making processes. (A limited number of data points were based on reasonable interpretations of subject actions, eg, if recognition of apnea and/or pulselessness were not explicitly called out, the subject's call to activate a cardiac arrest team was used as a substitute metric.) Scenarios were terminated on completion of all expected resuscitative actions or when the subject reported completion of all actions felt to be necessary. Each subject received a $50 gift certificate incentive for study participation.

Failure criteria recorded included failure to start BVM ventilation; initiating chest compressions (if simulated patient pulse present) or withholding chest compressions (if pulseless); and applying defibrillation (if sinus bradycardia scenario) or withholding defibrillation (if VF scenario). The percentages of simulated patients who did not suffer severe adverse patient care events (ie, those who did not experience any failure criteria) were used to derive composite outcome measures of the study groups' overall performances. This aggregative process to determine control and experimental group performance endpoints was selected as a method analogous to survival analysis in that all critical elements of resuscitation would be required for a patient to successfully recover from SCA. For example, a study group that does not ventilate seven of their 25 bradycardic simulated patients and performs inappropriate chest compressions on eight of the bradycardic patients (of whom three are also not ventilated) would be scored as managing 13 (52%) of their patients appropriately.

Three critical care-experienced RN study investigators (J.A.D.-V., M.H.R., and B.A.S.) completed training on the study protocol and tool during two on-site pilot sessions with the manikin and nonstudy RNs. These sessions were conducted to troubleshoot the study tool and familiarize investigators with the simulation scenario content, flow, and experience; the learning curve for the study protocol was minimal due to investigators' clinical background and the study's format and structure. Nonblinded observations during actual sessions were completed in real time by collaborative two-person study observer teams; this format ensured strict compliance with study protocol and provided opportunities for investigator crosschecks. A portable videorecorder recorded all sessions on DV tape for offline review of dataset accuracy and completeness with StudioCode™ v2.5.45 software (StudioCode Business Group, Camarillo, CA) by the primary investigator (L.K.) on conclusion of all study sessions.

Statistical Analysis

Demographic information and clinical experience were compared between study groups. Sequence of study scenario and study date was analyzed for their effect on primary study outcome measures. Within-group and between-group differences in SCA resuscitation performance were analyzed with Fisher exact test and Mann-Whitney tests for nominal and time data, respectively.

RESULTS

Fifty nurses were enrolled and consented at the study institution through shift announcements and on-shift recruitment from January to March 2009. Twenty-five of 32 available nurses on the control floor were recruited and 25 of 30 nurses on the experimental unit. There were no significant differences in demographics, nursing experience, or prior performance of defibrillation between study groups (Table 1).

Table 1
Table 1:
Study Subject Demographics and Experience

VF Scenario

The study groups performed comparably during the simulated VF scenario in recognizing unresponsiveness, apnea and pulselessness, and retrieving resuscitation cart (Table 2). The control group was slower to call for help (mean time to completion 25 ± 17 seconds) compared with experimental subjects [mean time to completion 18 ± 11 seconds (P < 0.05)]. Chest compressions were initiated by 44.0% of control group at an average of 77 ± 32 seconds; experimental group subjects initiated compressions on 95.8% of their patients (P < 0.001) at 115 ± 37 seconds (P < 0.01). Experimental subjects initiating compressions after CPR prompt by AED (“AED-prompted” subgroup) did so at 128 ± 29 seconds (P < 0.001 relative to control group's time to compressions), whereas subjects compressing before AED prompt (“unprompted” subgroup) started at 76 ± 35 seconds (not significant [NS] relative to control group).

Table 2
Table 2:
Study Results

BVM ventilation was performed equivalently by control (84.0%, with mean time to initiation of 101 ± 40 seconds) and experimental subjects [76.0% (NS), with mean time to initiation of 104 ± 30 seconds (NS)]. The groups were equal in completion and timeliness for turning defibrillation device on [control group with 92.0% and mean time to completion of 101 ± 52 seconds versus experimental group with 100% (NS) and 115 ± 73 seconds (NS)] and application of defibrillation pads [control group with 92.0% and mean time to completion of 110 ± 53 seconds versus experimental group with 100% (NS) and 121 ± 68 seconds (NS)].

Eighty-eight percent of patients treated by the control group were appropriately defibrillated at an average of 155 ± 59 seconds, with 32.0% of all control subjects using semiautomated rhythm analysis for defibrillations at 148 ± 52 seconds (defibrillations without analysis occurred at 161 ± 60 seconds). AED-supplemented response group subjects defibrillated 100% (NS) of their patients at a mean time of 154 ± 72 seconds (NS) with 100% AED rhythm analysis (P < 0.001); no subjects used SD instead of AED. Many subjects in both groups failed to call out warnings before defibrillation [63.6% control versus 68.0% experimental (NS)]. Control and experimental groups scored composite outcome measures of 28% and 72% (P < 0.01), respectively, for adequacy of overall VF resuscitation performance.

Sinus Bradycardia Scenario

The bradycardia scenario elicited between-group differences in initiation of inappropriate chest compressions (Table 2): 28.0% of control group performed compressions at 69 ± 59 seconds versus 69.6% (P = 0.01) of experimental group at 38 ± 49 seconds (NS). Twenty-four percent of control subjects used semiautomated rhythm analysis at 149 ± 79 seconds; 80.0% of experimental group used AED rhythm analysis (P < 0.001) at 142 ± 54 seconds (NS). Two control subjects operating without semiautomated rhythm analysis were observed to manually charge their defibrillators, with defibrillation of one bradycardic simulated patient; no electrical therapy was applied by the AED-supplemented response group (NS). Sixty-four percent of control group and 28% (P = 0.01) of experimental group subjects avoided improper bradycardia management overall.

DISCUSSION

Sudden cardiac arrest resuscitation may be compromised by a variety of factors. Patient characteristics, insufficient (maintenance) training of responding personnel, teamwork deficiencies, inadequacies in equipment supply chains, and embedded limitations in institutional clinical systems are key impediments to optimal SCA response.1,10–12 The current investigation examined the integrated deployment of a user-enabling technology to improve one select area of SCA management. The study successfully expanded on prior work by Kaye, Mancini et al.13,14 through controlled comparison of SD and AED response models, in situ simulation for enhanced realism, and use of two distinct scenarios to examine responder rhythm differentiation for appropriate electrical therapy.

Feasibility

The study investigators used in situ simulation to objectively evaluate the existing institutional SCA response system and an alternative model using AED devices. Research methodology was optimized through use of a simulation platform providing the necessary patient physiologic findings, compatibility with resuscitative interventions, and environmental cues necessary to recreate the initial resuscitation phase of a SCA patient by a solitary provider in a hospital setting. Use of intermediate-fidelity patient simulators with portable audiovisual recording equipment allowed rapid study setup and cleanup in a live clinical environment, thereby reducing interruptions in clinical patient care and departmental flow.

Current Institutional Model of In-Hospital SCA Response

As evidenced by the in-simulation performance of a representative cohort of medical/surgical floor nurses using SDs, existing SCA response systems suffer from several limitations with respect to standard process-of-care markers such as those used by the National Registry of Cardiopulmonary Resuscitation (www.nrcpr.org).15 Subjects displayed varying compliance with Basic life support actions, with 56% failing to initiate chest compressions and 16% never performing BVM ventilation during the pulseless arrest scenario. Arrhythmia-specific management was similarly problematic: two thirds of the SD response model group did not activate rhythm analysis functions and 12% failed to defibrillate. These omissions are of particular interest considering that at least 92% of the same control subjects retrieved, connected, and turned on the SD. As for patient and provider safety, one bradycardic patient received a manual-operation defibrillation, and approximately two thirds of defibrillations were completed without operator warning.

Accounting for errors of both omission (eg, nonperformance of indicated defibrillation) and commission (eg, inappropriate chest compressions on patient with pulse) that would be expected to cause serious patient harm, the SD response group performed the initial resuscitation of 28% of VF patients and 64% of sinus bradycardia patients without severe adverse events (Table 2, Figs. 3 and 4).

Figure 3.
Figure 3.:
Comparison of control and experimental response models [ventricular fibrillation (VF) scenario]. The paired images represent control (top) and experimental (bottom) study groups' performance during VF simulation scenario. Completion of resuscitative actions by each group is represented on the vertical axis, with higher positions indicating better group performance. The horizontal axis reflects the passage of time, starting at the left and moving to the right over the duration of simulation. Each data point circle indicates the mean percent completion and mean time for each action by study group; adjacent line extensions show standard deviations (numerical values, including ranges, are printed for each data element). Clustering of data elements in the upper left indicates better and faster group performance. Significant response superiority for either group is highlighted with ovals. Composite outcome measures are derived from the bag-valve-mask ventilation, chest compression, and defibrillation failure criteria as described in the text.
Figure 4.
Figure 4.:
Comparison of control and experimental response models (sinus bradycardia scenario). The paired images represent control (top) and experimental (bottom) study groups' performance during sinus bradycardia simulation scenario (see Fig. 3 description for graph details).

AED-Supplemented SCA Response Model

Simulated SCA resuscitations by the age- and experience-matched experimental cohort did not suffer from AED- related delays in defibrillation. Basic life support performance was mostly similar, with a few exceptions. Although more chest compressions were administered by the experimental group during VF scenario, the subjects who required AED prompting started 40 seconds later than control or nonprompted experimental subjects. More troubling is the finding that 70% of subjects in the AED-supplemented response group administered unnecessary external cardiac compressions to unresponsive simulated patients with a strong, palpable manikin extremity pulse [with 9 (52.9%) of the 17 subjects compressing improperly only and after AED voice prompt]. This is consistent with other published reports regarding the concordance of AED operator performance with device algorithm design and human-machine interface.16–18

When examined for composite endpoints, the experimental group appropriately managed 72% of VF patients and 28% of bradycardic patients (Table 2, Figs. 3 and 4). Relative to the overall performance of the SD response model, AED deployment contributed significantly to improved VF SCA treatment with higher rates of chest compression. However, in exchange, AEDs (as configured for the study) indiscriminately prompted for chest compressions regardless of cardiac rhythm to promote circulatory support for clinical states necessitating or following defibrillation. Because the experimental group's inferior overall performance during bradycardia scenario was primarily due to nonindicated chest compressions, study-based modifications to the implementation of the AED-supplemented response model were focused on reducing inappropriate chest compressions induced by the AED. Safeguards involving modified AED device signage, amendment of recorded prompt to “If patient has no pulse, start chest compressions,” and training initiatives on appropriate AED activation are being implemented in collaboration with the Biomedical Engineering department, AED manufacturer, and institutional nurse educators. (Notably, this systems feedback generated by the study represents the inherent value of medical simulation for usability studies and real-world debugging of medical devices and related technologies.) Furthermore, study data regarding insufficient ventilatory support and inappropriate defibrillation are being applied by study investigators to advocate for adjustments in institutional resuscitation training.

Study Implications for In-Hospital SCA Resuscitation Response

Informal surveys during ACLS recertification sessions at the study institution revealed that one third of nurses are hesitant to perform defibrillation independently or without a physician present. It follows that the control group's rate of defibrillation during simulations—88% with standard-issue defibrillators—may conceal a lower rate of RN-operated defibrillation in actual SCA settings. Additionally, only 22% of study subjects reported prior experience as primary operators during an actual patient defibrillation, despite average nursing experience of 12 years. In light of infrequent clinical exposure and provider disinclination toward rapid, independent defibrillation, it is likely that the required medical decision-making and human factors-related challenges of the traditional SCA response model impede expeditious and effective medical care. With most patients in SCA discovered by nursing personnel at the study institution, the capacity, willingness, and expectation of these “first-arrivers” to become first responders represent nontrivial “disconnect” areas in the SCA resuscitation mechanism. AED deployment may serve to specifically target these points of weakness, thereby transforming noncritical care staff members into defibrillation-adept independent providers with on-site arrhythmia interpretation backup.

Within this context, the study evaluated potential institutional implementation of AEDs as a simplified, safe, cost-efficient, low-maintenance, user-enabling alternative with minimal sustainment training requirements.2,19–25 Representing a sophisticated link in the cardiac “chain of survival,” AEDs have demonstrated their value in the layperson, emergency medical services/public health sector, and in-hospital settings. However, several studies have revealed a need for tight coordination of the entire resuscitation system to allow for AED efficacy.5,26–34 Consequently, investigators used in situ medical simulation to determine whether the institution's existing resuscitation infrastructure was compatible and amenable to improvement with AEDs.

Presentation of study findings to the institutional Cardiac Arrest Committee along with discussion of published AED literature resulted in the unanimous decision to implement a phased introduction of AEDs to all noncritical care hospital settings. (Phasing was determined by logistic constraints; additionally, AEDs have already been deployed in all nonpatient care areas.) Education programs using the study manikin and AED trainers have been mobilized, whereas monitoring, maintenance, and follow-up mechanisms have been put in place for continuous reassessment of AED functionality and unanticipated problems. Further study is ongoing to ascertain whether the new response model translates to improved bedside resuscitation through empowerment and supportive technologies for in-hospital first responders.

Limitations

Translation of the study's findings to improvements in actual bedside care has not been demonstrated at this time. Study design precluded blinding of observers to defibrillation device and subject group assignment. The study focus on defibrillation and first responder behaviors precluded detailed evaluation of associated SCA resuscitation activities (eg, CPR technique) and nontechnical elements (eg, teamwork and communication). Only one model of AED was tested due to existing equipment supply infrastructure. Although the presession briefing on manikin-specific application methods for defibrillation pads would be expected to have affected both groups' performances equally, the review of AED “on” switch for experimental group subjects may have resulted in a predisposition toward SCA resuscitative actions. The finding of infrequent predefibrillation warning by responders remains unresolved in that it is unclear whether this was due to the limited realism of the study's single-responder scenario.

CONCLUSION

In situ medical simulation of in-hospital resuscitation events experimentally compared an AED-supplemented first responder resuscitation model with the existing response system. Objective findings, both anticipated and unexpected, were used to guide institutional AED deployment, device modifications, and educational programs for improvement of SCA response.

ACKNOWLEDGMENTS

The authors acknowledge Anna C. Cousins and Whit Hill for their assistance in manuscript preparation.

REFERENCES

1. Chan PS, Krumholz HM, Nichol G, Nallamothu BK. Delayed time to defibrillation after in-hospital cardiac arrest. N Engl J Med 2008;358:9–17.
2. Kellerman AL, Hackman BB, Somes G, Kreth TK, Nail L, Dobyns P. Impact of first-responder defibrillation in an urban emergency medical services system. JAMA 1993;270:1078–1013.
3. van Alem AP, Vrenken RH, de Vos R, Tijssen JG, Koster RW. Use of automated external defibrillators by first responders in out of hospital cardiac arrest: prospective controlled trial. BMJ 2003;327:1312.
4. Hallstrom AP, Ornato JP, Weisfeldt M, et al. PAD Investigators. N Engl J Med 2004;351:637–646.
5. Sanna T, La Torre G, de Waure C, et al. Cardiopulmonary resuscitation alone vs. cardiopulmonary resuscitation plus automated external defibrillator use by non-healthcare professionals: a meta-analysis on 1583 cases of out-of-hospital cardiac arrest. Resuscitation 2008;76:226–232.
6. Nichol G, Huszti E, Birnbaum A, et al. PAD Investigators. Cost-effectiveness of lay responder defibrillation for out-of-hospital cardiac arrest. Ann Emerg Med 2009;54:236–238.
7. Kaye W, Mancini ME. Teaching adult resuscitation in the United States-time for a rethink. Resuscitation 1998;37:177–187.
8. 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;58:297–308.
9. Hazinski MF, Idris AH, Kerber RE, et al. Lay rescuer automated external defibrillator (“public access defibrillation”) programs: lessons learned from an international multicenter trial: advisory statement from the American Heart Association Emergency Cardiovascular Committee; the Council on Cardiopulmonary, Perioperative, and Critical Care; and the Council on Clinical Cardiology. Circulation 2005;111:3336–3340.
10. Aufderheide TP, Lurie KG. Death by hyperventilation: a common and life-threatening problem during cardiopulmonary resuscitation. Crit Care Med 2004;32(suppl 9):S345–S351.
11. Marsch SC, Muller C, Marquardt K, Conrad G, Tschan F, Hunziker PR. Human factors affect the quality of cardiopulmonary resuscitation in simulated cardiac arrests. Resuscitation 2004;60:51–56.
12. Starr LM. Rethinking SCA at work. Optimizing one person does not guarantee that an organizational problem will be solved. Occup Health Saf 2007;76:55–56.
13. Mancini ME, Kaye W. Measuring cardiopulmonary resuscitation performance: a comparison of the Heartsaver checklist to manikin strip. Resuscitation 1990;19:135–141.
14. Kaye W, Mancini ME, Richards N. Organizing and implementing a hospital-wide first-responder automated external defibrillation program: strengthening the in-hospital chain of survival. Resuscitation 1995;30:151–156.
15. Hunt EA, Mancini ME, Smyth M, Truitt TL. Using the American Heart Association's National Registry of Cardiopulmonary Resuscitation for performance improvement. Jt Comm J Qual Patient Saf 2009;35:13–20.
16. Fleischhackl R, Losert H, Haugk M, et al. Differing operational outcomes with six commercially available automated external defibrillators. Resuscitation 2004;62:167–174.
17. Snyder D, Morgan C. Wide variation in cardiopulmonary resuscitation interruption intervals among commercially available automated external defibrillators may affect survival despite high defibrillation efficacy. Crit Care Med 2004;32(9 suppl):S421–S424.
18. Hedges JR, Sehra R, Van Zile JW, et al. Automated external defibrillator program does not impair cardiopulmonary resuscitation initiation in the public access defibrillation trial. Acad Emerg Med 2006;13:659–665.
19. Peberdy MA, Ottingham LV, Groh WJ, et al. Adverse events associated with lay emergency response programs: the public access defibrillation trial experience. Resuscitation 2006;70:59–65.
20. Riegel B, Mosesso VN, Birnbaum A, et al. Stress reactions and perceived difficulties of lay responders to a medical emergency. Resuscitation 2006;70:98–106.
21. Groeneveld PW, Kwong JL, Liu Y, et al. Cost-effectiveness of automated external defibrillators on airlines. JAMA 2001;12:1482–1489.
22. Caffrey SL, Willoughby PJ, Pepe PE, Becker LB. Public use of automated external defibrillators. N Engl J Med 2002;347:1242–1247.
23. Cram P, Fendrick AM, Vijan S. Public use of automated external defibrillators. N Engl J Med 2003;8:755–756.
24. Cram P, Vijan S, Fendrick AM. Cost-effectiveness of automated external defibrillator deployment in selected public locations. J Gen Intern Med 2003;18:745–754.
25. Christenson J, Nafziger S, Compton S, et al. The effect of time on CPR and automated external defibrillator skills in the Public Access Defibrillation Trial. Resuscitation 2007;74:52–62.
26. Page RL, Joglar JA, Kowal RC, et al. Use of automated external defibrillators by a U.S. airline. N Engl J Med 2000;343:1210–1216.
27. Valenzuela TD, Roe DJ, Nichol G, Clark LL, Spaite DW, Hardman RG. Outcomes of rapid defibrillation by security officers after cardiac arrest in casinos. N Engl J Med 2000;343:1206–1209.
28. Marenco JP, Wang PJ, Link MS, Homoud MK, Estes NA. Improving survival from sudden cardiac arrest: the role of the automated external defibrillator. JAMA 2001;285:1193–1200.
29. Richardson LD, Gunnels MD, Groh WJ, et al. Implementation of community-based public access defibrillation in the PAD trial. Acad Emerg Med 2005;12:688–697.
30. Destro A, Marzaloni M, Sermasi S, Rossi F. Automatic external defibrillators in the hospital as well? Resuscitation 1996;31:39–43; discussion 43–44.
31. Zafari AM, Zarter SK, Heggen V, et al. A program encouraging early defibrillation results in improved in-hospital resuscitation efficacy. J Am Coll Cardiol 2004;44:846–852.
32. Sweeney TA, Runge JW, Gibbs MA, et al. EMT defibrillation does not increase survival from sudden cardiac death in a two-tiered urban-suburban EMS system. Ann Emerg Med 1998;31:234–240.
33. Groh WJ, Birnbaum A, Barry A, et al. Characteristics of volunteers responding to emergencies in the Public Access Defibrillation Trial. Resuscitation 2007;72:193–199.
34. Moore MJ, Hamilton AJ, Cairns KJ, et al. The Northern Ireland Public Access Defibrillation (NIPAD) study: effectiveness in urban and rural populations. Heart 2008;94:1614–1619.
Keywords:

Defibrillation; Healthcare quality improvement; Medical emergency team; Research; Simulation

© 2010 Lippincott Williams & Wilkins, Inc.