Journal Logo


Board 114 - Research Study How Closely Does Simulation Approximate Reality? A Comparison of Key Resuscitation Metrics Between Simulated and Actual Cardiopulmonary Arrests (#16220)


Simulation in Healthcare: The Journal of the Society for Simulation in Healthcare: February 2016 - Volume 11 - Issue 1 - p 68
doi: 10.1097/SIH.0000000000000147
  • Free

A few data errors were discovered in this board submission. Please see the corrected submission here.

Hypothesis: Little is known about how key resuscitation benchmarks during simulated cardiac arrests compare to real-world events and outcomes. Discrepancies in these metrics may bear directly on the validity of simulated environments. In an effort to better clarify these differences the International Simulation Data Registry (ISDR), a consortium of simulation centers that contribute cardiac arrest simulation data to a centralized location, was queried for the rates of Return of Spontaneous Circulation (ROSC) and time to key resuscitation events among cardiac arrest simulations. These metrics were then compared to recently published American Heart Association Get With the Guidelines-Resuscitation (AHA GWTG-R) registry data to better assess similarities and differences in key event metrics and ROSC rates.1

Methods: The ISDR database includes information from 11 across the United States, Canada, and Europe. Participation was approved by each center’s Institutional Review Board. Operational definitions for key event metrics were developed via a modified Delphi process focused on the assessment of simulated pulseless events occurring as the first abnormal clinical state. Variables collected include session demographics, time to key event metrics, total arrest duration and arrest outcome. Data was queried after approximately one year of operation. Time-to-event variables were rounded down to the nearest minute and time to recognition was then subtracted to match, as closely as possible, the AHA GWTG-R definitions. Real-world arrest metrics and outcomes were abstracted from a recently published AHA-GWTG-R analysis.1 Key event metrics were reported descriptively. ROSC rates were compared to those in the literature using Chi-square.

Results: A total of 330 independent simulations were analyzed (28% adult, 72% pediatric). Programs conducted simulations in both in-situ (77.6%) and center-based (22.4%) formats. An average of 8 learners attended each session, although discipline specific attendance varied widely. Median time to recognition of pulseless state was <1 min (range <1 to 2) for adults and <1 min (range <1 to 3) for children. Table 1 displays the key event metrics (time to CPR, time to Epinephrine administration, time to Defibrillation, and total arrest duration) as compared between actual and simulated arrests. ROSC rates were 13% higher in simulated adult arrests (52% actual ROSC vs 65% simulated ROSC) and 42% higher in simulated pediatric arrests (47% actual ROSC vs 89% simulated ROSC) as compared between AHA GWTG-R and ISDR databases.1 This difference is significant at P < 0.001 by Chi-square in the pediatric group.

Conclusion: CPR and Time to Epinephrine metrics appeared similar in simulated and actual arrests. Time to Defibrillation and Total Arrest Duration appeared to differ slightly, however, though formal statistical comparison could not be carried out due to lack of reported standard deviations for times to key events in the AHA GWTG-R report. Outlier negative values in both databases, however, suggest incomplete adherence to operational definitions and thus potential error. ROSC occurred in simulated arrests at statistically higher rates than in actual arrests for children. This difference may be due to philosophical decisions made by individual simulation programs. Greater consideration of how realistic our portrayals of post-arrest survival may be warranted, as this could affect learner expectation of success during actual resuscitations. More work is needed to assure consistency in the application of operational definitions between the ISDR and other arrest registries.

Table 1
Table 1:
Descriptive Comparison of Key Event Metrics in Simulated vs Actual Cardiac Arrests

This table displays the time to key arrest metrics for both adult and pediatric arrests as compared between the simulated and actual environments. Times are given in terms of median minutes to the named event with ranges in parenthesis. Negative values denote events in which the time of event recognition was recorded as occurring after the time of therapy administration. Actual arrest values reported are derived from Donoghue et al. American Heart Association's Get With the Guidelines-Resuscitation I: Cardiopulmonary resuscitation for in-hospital events in the emergency department: a comparison of adult and pediatric outcomes and care processes. Resuscitation (2015).


1. Donoghue A, Abella B, Merchant R, Praestgaard A, Topjian A, Berg R, Nadkarni V, American Heart Association's Get With the Guidelines-Resuscitation I: Cardiopulmonary resuscitation for in-hospital events in the emergency department: a comparison of adult and pediatric outcomes and care processes. Resuscitation 2015.


Calhoun A, Maa T, Auerbach M, Overly F, Geeraerts T, et al. Board 114 - How closely does simulation approximate reality? A comparison of key resuscitation metrics between Ssimulated and actual cardiopulmonary arrests (#16220). Simul Healthc 2015; 10(6): 402–403.
    © 2016 Society for Simulation in Healthcare