Rapid Cycle Deliberate Practice Improves Retention of Pediatric Resuscitation Skills Compared With Postsimulation Debriefing

Abstract

Introduction 

Rapid cycle deliberate practice (RCDP) for teaching team-based resuscitation is associated with similar improvements in immediate performance as compared with postsimulation debriefing (PSD). Limited studies compare skill retention between these 2 modalities. Our objective was to compare retention of team leader performance in residents trained with RCDP versus PSD.

Methods 

This was a cluster-randomized trial comparing RCDP and PSD from January 2018 to April 2019. Pediatric and emergency medicine residents participated in simulation-based pediatric resuscitation education, and teams were randomized to undergo either RCDP or PSD. Each participant's team leader performance was assessed 1 to 12 months after training via a simulated cardiac arrest. The primary outcome was time to defibrillation. Secondary outcomes included overall team leader performance and time to chest compressions.

Results 

Thirty-two residents (90.6% pediatrics, 9.4% emergency medicine) met inclusion criteria (16 RCDP, 16 PSD). Of the 32 residents, 40% returned in 1 to 3 months, 25% 3 to 6 months, 16% 6 to 9 months, and 19% 10 to 12 months. Participants in RCDP had more than 5 times the odds of achieving defibrillation versus those in the PSD group (odds ratio = 5.57, 95% confidence interval = 1.13–27.52, P = 0.04). The RCDP group had a higher mean Resident Team Leader Evaluation score (0.54 ± 0.19) than the PSD group (0.34 ± 0.16, P < 0.001).

Conclusions 

This study shows significant differences in subsequent performance in the team leader trained with RCDP and suggests that RCDP may improve retention of pediatric resuscitation skills compared with PSD. Future studies should focus on best applications for RCDP with attention to knowledge and skill decay.

Pediatric resuscitations are rare yet life-threatening events that require a coordinated and effective team performance. Simulation, when compared with traditional clinical education, has been shown to improve performance as well as technical and nontechnical skills.¹ Multiple studies have demonstrated the value of simulation-based medical education (SBME) to enhance education and to improve patient outcomes.^2–4 Simulation-based medical education that incorporates deliberate practice and mastery learning has been associated with improved learning outcomes, compared with SBME methods that did not include these features.^5–8 Research on these 2 features has focused on individual skills. For team-training simulation, the tradition has been scenario-based simulation with postsimulation debriefing (PSD).⁹ A potential weakness of the PSD design is that it is not conducive to long-term retention of skills.^10,11 As a result, there is a risk that these skills will not be applied when required. This is especially important for cardiac arrest, when interventions, such as defibrillation and initiation of cardiopulmonary resuscitation (CPR), are time-sensitive and directly affect patient outcomes.¹²

Rapid cycle deliberate practice (RCDP) is an SBME model for team-based training that is structured to quickly rotate from learner practice and expert feedback to achieve repeatable performance of a criterion standard choreography while providers use verbal scripts to improve communication during time-sensitive situations.¹³ Multiple forms of debriefing strategies are used including ongoing coaching, advocacy inquiry, and plus delta.^13–15 Teams are allowed enough time to practice these skills correctly until mastery is achieved.

A qualitative study from our group analyzed learners' experiences during RCDP simulation and found it to be well received by learners who reported increased confidence and decreased cognitive load.¹⁶ We hypothesized that a similar effect would be seen when skills were tested several months after initial training and that individuals taught with RCDP would have improved skill acquisition and retention.

Previous studies have also shown that RCDP improves learner performance and that it may be at least as effective as PSD in teaching complex resuscitation team skills in the short term.^13,15,17 However, there are limited data comparing the retention of pediatric resuscitation skills acquired by these 2 modalities. The relative importance of reflective learning and deliberate practice for resuscitation performance has not been clearly defined, and different structures of simulation need to be compared.

The aim of our study is to determine whether future performances of residents acting as team leaders of pediatric resuscitation scenarios are better after training with RCDP simulation compared with residents trained through PSD simulation with time to defibrillation serving as the primary outcome.

METHODS

Design

This was a cluster-randomized trial comparing the effect of RCDP and PSD education on follow-up performance in simulation.

Study Setting and Participants

The study was conducted in the in situ simulation suite of the emergency department at an urban, tertiary care children's hospital from January 2018 to April 2019.

Institutional board approval (Reference #H-41413) was obtained before the start of enrollment.

As part of their pediatric emergency medicine (PEM) rotation, pediatric and emergency medicine residents participated in simulation-based resuscitation training and had the opportunity to enroll in the study. If a participant did not consent to the study, simulation training continued that day, but the team's data were not collected or used. Depending on the residency program and postgraduate year, residents were scheduled to have 1 to 3 PEM rotations during their residency. On each training day for every PEM rotation, a team was formed comprising 3 to 4 residents and 2 nurses. The team was randomized to either RCDP or PSD. Two PEM faculty simulation instructors and 1 nurse educator led these training sessions. All educators were trained to teach both RCDP and PSD. They completed an instructor course as well as an additional workshop for RCDP to teach this specific curriculum. There was ongoing monitoring of instructors by the experts who taught the workshops to ensure consistency, and feedback was provided to the instructors. Roles included team lead, first responder, airway, CPR coach, bedside provider, and recorder. For teams of 5, the CPR coach was removed because this role is not always available in every resuscitation, whereas all the other roles are critical to every code.

After completion of their initial training session, every resident who returned during the study period, on a subsequent PEM rotation, was eligible for enrollment in this study. Learners excluded from this study were those who failed to complete the entire simulation protocol and those whose video recording could not be used because they were either improperly recorded or missing critical information. In addition, although they were part of the intervention and initial training session, our returning nurses were excluded from this study.

Randomization.com was used to develop a scheme with equal numbers of groups in each arm for blocks of 6 teams. In addition to randomization to the type of training, the roles of each member of the team were also randomized. The research assistant was unaware of that day's randomization until consent had been obtained. Once consent was obtained, the research assistant determined to which preassigned type of training that the group was randomized.

Simulation Curriculum Design

Simulation With PSD

Teams randomized to RCDP or PSD training and debriefing were taught the same topics in the same timeframe. In the PSD group, the training session included 1 uninterrupted 20-minute scenario of an unresponsive child presenting in pulseless electrical activity (PEA) arrest. The case started with the first responder entering the room with the remainder of the team arriving 10 seconds after the call for help. The case proceeded through basic life support, monitor placement, intraosseous (IO) access, and epinephrine administration until the patient had return of spontaneous circulation and required postarrest stabilization with intubation and treatment of shock. This was followed by a 40-minute reflective debriefing session. Instructors used a scripted debriefing including choreography and learning objectives (see Doc, Supplemental Digital Content 1 https://links.lww.com/SIH/A654, which demonstrates RCDP and PSD teaching scenarios). Debriefing for traditional simulation was conducted using the PEARLS (Promoting Excellence and Reflective Learning in Simulation) framework with a scripted debriefing.¹⁸

Rapid Cycle Deliberate Practice Simulation

Participants in RCDP engaged in a 1-hour teaching session with predetermined learning objectives. The session included multiple rounds of progression of the same scenario in rapid sequence with predesigned “hard stops” and “soft stops” with scripted learning points. Learners achieved predefined goals in each round before progressing to the next level of difficulty. Instructors focused on debriefing with direct feedback, using a pause and rewind/restart methodology.

In RCDP, feedback was interspersed throughout the simulation and was short and concentrated on team actions that had just been performed. With each round, team members rotated roles, giving everyone a chance to be the team lead. With predefined goals in place, learners had the opportunity to rewind before a key action or entirely restart the scenario and achieve those goals before moving onto the next round. For example, the first round presented an unresponsive child with apnea but with a pulse. The team's objectives were to quickly assess the patient and recognize the need for additional help and a crash cart, apply monitors, and reposition the airway. Once those objectives were achieved, the team moved onto the second round, an apneic patient requiring bag-mask ventilation. The third round was of a pulseless patient requiring quick recognition and initiation of chest compressions. The fourth round required placement of a backboard, pads, step stool, and charging the defibrillator to the appropriate energy setting in an effort to reach the first pulse and rhythm check in a coordinated way. Once the team identified the rhythm as PEA, the fifth round required access via IO placement and administration of epinephrine. The final round required pulse and rhythm checks every 2 minutes, multiple doses of epinephrine, the switching of compressors, and mental modeling by the team leader who directed adjunctive therapies for the treatment of shock until the patient achieved return of spontaneous circulation. The objectives and skills expected for this final round of RCDP were the same as those for the PSD scenario (see Document, Supplemental Digital Content 1 https://links.lww.com/SIH/A654, which demonstrates RCDP and PSD teaching scenarios).

Although the technique of debriefing in RCDP is more directive than in PSD, instructors were encouraged to use advocacy and inquiry methods to explore persistent performance gaps in both types of simulation.¹⁹ In contrast to RCDP, the PSD model allowed more time for reflection on individuals' frames and sought to correct the underlying frame through feedback during an active discussion. Even in the PSD simulations, instructors would provide direct feedback focusing on actions and, specifically, on ideal choreography of pediatric resuscitation. Moreover, both groups had the opportunity to review microprocedures, such as IO placement, manipulating the defibrillator, which included review of shockable rhythms, and drawing up epinephrine; however only the RCDP group was allowed to deliberately practice these skills. Both RCDP and PSD used a blended approach to debriefing using the PEARLS framework and a debriefing script.¹⁸ The differences in debriefing techniques lay in how much emphasis was placed on direct feedback versus advocacy inquiry technique, as well as the timing of that feedback. Rapid cycle deliberate practice debriefing used the technique of direct feedback more than advocacy inquiry, whereas PSD debriefing allowed more time for reflective advocacy inquiry over direct feedback.

Study Protocol and Data Collection

After obtaining informed consent, a brief orientation to simulation and the SimJunior mannequin (Laerdal Medical; Stavenger, Norway) was provided. Teams of residents and nurses participated in a 1-hour training simulation focusing on resuscitation of a child with PEA, using either RCDP or PSD based on randomization. After the teaching scenario, the team completed a test case of a child in ventricular fibrillation cardiac arrest, requiring initiation of CPR, defibrillation, and administration of epinephrine.

Participants returned for a second simulation day on a subsequent PEM rotation, as determined by their residency schedule, ranging from 1 to 12 months after their initial training session. Before the start of this session, participants completed a demographic survey administered through Research Electronic Data Capture (REDCap), a secure institutional database.²⁰ This survey queried each learner's level of training, type of residency, whether they were pediatric advanced life support (PALS) certified, and estimated number of real-life resuscitations attended. They were tested individually on a pulseless ventricular tachycardia (PVT) arrest case before the start of their second day of simulation training. These individuals were tested in the team lead role. This case was scripted and preprogrammed with 3 confederates (2 PEM faculty, 1 nurse) playing the other roles (1 person to give bag-mask ventilation, 1 person to perform CPR, and another to assist with all other tasks; see Document, Supplemental Digital Content 2 https://links.lww.com/SIH/A655, which contains the retention test scenario). These confederates performed bedside tasks based on the team leader's directions. Performance data for participants were measured in real time by study coinvestigators using a stopwatch and recorded on an encrypted spreadsheet, and the cases were video recorded. The initial call for help was for an unresponsive and apneic child for which bag-valve mask was initiated. Resuscitative efforts would reveal that the child is also in PVT arrest requiring defibrillation. This case lasted 3 minutes or until a shock was delivered. Two different defibrillators were used during the study. Most participants were trained and tested with the Zoll-R Series defibrillator (Zoll Medical Corporation, Chelmsford, MA), which replaced all the Lifepak 20 defibrillators (Physio-Control, Redmond, WA) across the institution in March 2018. Because of the switch, several participants were tested with a different defibrillator than the one with which they were trained.

Video recordings were used to evaluate resident team leader performance. Five coinvestigators served as video reviewers. One blinded reviewer (S.K.W.) confirmed the times recorded in real time to chest compressions and time to defibrillation for all cases to minimize bias and ensure consistency; the start time was determined by the beginning of a verbal prompt given by the nurse upon each participant's entry into the room in response to the call for help. In addition, each team leader performance was assessed by 2 of the 5 reviewers for each video recording; their respective scores were recorded as Resuscitation Team Leader Evaluation (RTLE) 1 and RTLE 2 for each case. Each reviewer was given a number 1 through 5. For every 10 videos, every combination of 2 of the 5 reviewers was used so that each video was reviewed twice and each reviewer was assigned to 4 videos. Reviewers were blinded to study group allocation of the participants.

The RTLE is a validated tool that assesses elements of team leader competency in pediatric resuscitation, and each participant was scored using a modified RTLE.²¹ The original tool includes 26 items, 12 that evaluate leadership and communication and 14 that evaluate knowledge and skills. Because of the brief length of the test case in our study, the application of the full-length tool was limited as several items could not be applied or adequately measured. Therefore, a priori, the study coinvestigators limited the components of this scoring tool to those that could be assessed in the 3-minute test case, before data collection and analysis (see Doc, Supplemental Digital Content 3 https://links.lww.com/SIH/A656, REDCap Survey). The reviewers together evaluated every component of the original RTLE and determined whether it could be applied to the brief test case. Together, they adjusted elements of the RTLE to fit the test case and reached an agreement on the grading of every component.

Outcomes

The primary outcome was time to defibrillation in a pediatric cardiac arrest case by participants, previously trained in either RCDP or PSD simulation, in the team lead role at delayed follow-up. Secondary outcomes included time to chest compressions and team leader performance based on the modified RTLE.²¹

Statistical Analysis

Using data from a prior study,²² we estimated time to defibrillation to be approximately 110 seconds. We wanted the power to detect a difference of 30 seconds between teams, so we calculated the mean of the populations to be approximately 110 and 140 and a standard deviation of 30. Using a 2-tailed α value of 0.05 and a β value of 0.2, we calculated a sample size of 16 per arm via https://www.stat.ubc.ca/~rollin/stats/ssize/n2.html.

Differences between the 2 study groups and learner demographics were assessed using the Pearson χ² test for categorical variables and the Mann-Whitney test for skewed, continuous data. If any of the categorical variables included a value of less than 5, the Fisher exact test was used. Frequencies, column percentages, medians, interquartile ranges, and P values were reported.

To determine the odds of achieving defibrillation in the intervention group, an odds ratio (OR) with a 95% confidence interval (CI) and a P value was calculated. A time-to-event analysis (Kaplan-Meier curve/log rank P value) using 1 minus survival for time-to-shock was also conducted to show the proportion of those who achieved defibrillation for the 2 groups. If the teams did not reach defibrillation by 3 minutes (stopping point), they were censored (contributed time without an event).

The distributions of the continuous outcome variables were normal, so independent t tests were used to determine time differences. Power, means, standard deviations, and P values for each outcome were reported.

Lastly, to determine the reliability of the independent scores generated from video viewing, an intraclass correlation coefficient (ICC) analysis was conducted to find the agreement between 2 raters. A 1-way random model for absolute agreement using single measures was chosen. For this analysis, the ICC, 95% CIs, and P values were reported. Statistical significance was defined as a P value of less than 0.05. All analyses were conducted using the Statistical Package for the Social Sciences (SPSS), Version 25 (IBM Corp, Armonk, NY).

RESULTS

Participants

Two hundred sixty-four residents were eligible to participate in this study. Once enrolled, they were randomized as a group to receive either RCDP (n = 127) or PSD (n = 137) on their initial simulation day. We expected a significant percentage of enrolled participants to be lost to follow-up because of the nature of residents' schedules during the study period. Some residents rotated through PEM once and never returned for a subsequent session. Most pediatric residents rotate through PEM twice a year in their second year (ranging 1–11 months between rotations). Our study was aligned with our residents' schedules; thus, most participants could not be captured in our study period, which began in the middle of the academic year. After a variable washout period of 1 to 12 months since the initial assessment, 36 residents returned for a second simulation day during the study period and met inclusion criteria. Four residents were excluded because of either video recording failure or technical error during the case where the incorrect rhythm was displayed on the monitor. Ultimately, data from 32 residents (90.6% pediatrics, 9.4% emergency medicine) were analyzed for this study. Within the follow-up period of 1 to 12 months, 21 residents returned within 6 months (9 RCDPs, 12 PSDs) and 11 residents returned between 6 and 12 months (7 RCDPs, 4 PSDs). More specifically, 40% returned in 1 to 3 months, 25% 3 to 6 months, 16% 6 to 9 months, and 19% 10 to 12 months. For both study groups, pediatric residents were the majority (Fig. 1).

FIGURE 1: Flow of study.

Learner characteristics are presented (Table 1). There were no significant differences between baseline characteristics (training program, year of training, prior simulation lead, and number of codes attended) and study groups. All participants were PALS certified throughout the study period.

TABLE 1 - Learner Demographics (N = 32)

	PSD (n = 16, 50.0%), n (%)	RCDP (n = 16, 50.0%), n (%)	P
Training program
Categorical pediatrics	14 (87.5)	11 (68.8)	0.19
Emergency medicine	0 (0.0)	3 (18.9)
Pediatric specialty tracks	2 (12.5)	2 (12.5)
Year of training
PGY1	1 (6.3)	1 (6.3)	1.00
PGY2	15 (93.8)	15 (93.8)
PALS certified	16 (100.0)	16 (100.0)	*
Prior simulation experience	15 (93.8)	15 (93.8)	1.00†
Prior simulation lead experience	12 (75.0)	15 (93.8)	0.33†
Actual codes attended	3.0 (2.0, 3.75)	3.0 (2.0, 8.75)	0.82†

*Cannot be calculated.

†Fisher’s exact test was used when any cell value was less than 5.

‡P value calculated using the Mann-Whitney test.

PGY: postgraduate year.

Performance

Significantly more members of the intervention group achieved defibrillation within 3 minutes: 81.3% in the RCDP group versus 43.8% in the PSD group. Therefore, the RCDP group had more than 5 times the odds of reaching defibrillation than those in the PSD group (OR = 5.57, 95% CI = 1.13–27.52, P = 0.04; Table 2). When using a time-to-event analysis, the percentages to defibrillation increased at a higher rate in the RCDP group than the PSD group (log rank P = 0.02; Fig. 2). Further analysis was done to assess skill retention in 3-month intervals and found our results to be inconclusive (see Table, Supplemental Digital Content 4 https://links.lww.com/SIH/A657, which demonstrates performance data in 3-month increments since time of training).

TABLE 2 - Odds of Achieving Defibrillation Within 3 Minutes by Study Group (N = 32)

	Shock, n (%)*	Did Not Shock, n (%)*	OR	95% CI	P
PSD	7 (43.8)	9 (56.3)	5.57	1.13–27.52	0.04
RCDP	13 (81.3)	3 (18.8)

*Row percentages are provided.

PSD, postsimulation debriefing (control).

FIGURE 2: Kaplan-Meier curve of time elapsed from onset of pulseless VT to defibrillation. Rapid cycle deliberate practice showed a higher rate of increase in percentages to defibrillation than PSD. VT, ventricular tachycardia.

All participants in both study groups initiated CPR during the test case. The RCDP group required less time than the PSD group to initiate compressions (22.75 seconds vs. 35.00 seconds, respectively; Table 3), but this difference was not statistically significant (P = 0.07). For the modified RTLE, the RCDP group had a higher score than the PSD control group (0.54 vs. 0.34, respectively, P < 0.001; Table 3).

TABLE 3 - Comparison of Study Groups and Outcomes Using the Independent T Test

	PSD, n	PSD, Mean (SD)	RCDP, n	RCDP, Mean (SD)	P
Time to shock, s	7	138.14 (38.42)	13	126.85 (33.59)	0.50
Washout period, mo	16	4.23 (3.17)	16	5.99 (3.73)	0.16
Time to CPR, s	16	35.00 (23.12)	16	22.75 (9.83)	0.07
RTLE (1) score	16	0.34 (0.16)	16	0.54 (0.19)	<0.001

Interrater Reliability

To assess interrater reliability, ICCs were calculated for both RTLE 1 and RTLE 2 scores. Interrater reliability had excellent agreement, having a significant correlation of 0.88 (95% CI = 0.77–0.94, P < 0.001).

DISCUSSION

Educational Significance

Our study found that individuals who were previously trained in pediatric resuscitation skills using RCDP were more likely than those trained using PSD to defibrillate within 3 minutes of cardiac arrest. In addition, RCDP was associated with an improvement in time to chest compressions and team leader performance, but this was not statistically significant.

In previous research, RCDP has been shown to be at least as effective as PSD in imparting immediate knowledge and skills.²³ However, there are limited data comparing the long-term retention of skills acquired by these 2 modalities of SBME, and the current data comparing these 2 teaching methods are limited and inconclusive.²⁴ Hunt et al¹³ showed that RCDP improves the time-sensitive “first 5-minute” performance of resident learners and introduced the concept of providing direct feedback paired with repetitive and deliberate practice. Rapid cycle deliberate practice was associated with key measures of resuscitation quality, including improved times to CPR and defibrillation, but without direct comparison to PSD. Lemke et al¹⁵ built upon these initial findings and directly compared RCDP with PSD simulation methods. Their study showed improvement in team performance after RCDP sessions, which was similar to the improvement seen after a PSD session. Their study also demonstrated a significantly greater improvement for human factors (establishing roles, closed-loop communication, shared mental model, mutual respect, knowledge sharing, constructive intervention, work distribution, and appropriate disposition) with RCDP compared with PSD simulation. In addition, another study by Cory et al²⁵ showed that although both RCDP and PSD were effective in training residents in the management of septic shock, RCDP was superior immediately after training; there was no difference on follow-up assessment 3 to 4 months after initial training.

It is possible that repetition with deliberate practice enhances memory consolidation, a key to skill retention. A large meta-analysis comparing SBME with deliberate practice versus traditional clinical education found superiority in SBME with deliberate practice as a teaching model for acquiring specific clinical skills.⁷ Similarly, RCDP mirrors the learning in these studies and emphasizes repetitive practice over reflective debriefing.

Our findings are consistent with and build upon those of previous studies by examining skill retention over time. While residents were taught pediatric resuscitation skills based on a case of PEA, they received training on the defibrillator that would allow them to be successful on a subsequent test that included a potentially more challenging algorithm in a case of PVT. Despite the difference in training and testing scenarios, our data suggest that RCDP may be more effective than PSD at solidifying learned skills and creating muscle memory for time-sensitive choreographed events.

The timing and quality of resuscitation matter when caring for critically ill patients. In cardiac arrest, the delivery of appropriate defibrillation for a shockable rhythm is crucial. The PALS guidelines continue to emphasize immediate CPR and early defibrillation in pediatric cardiac arrest as soon as a shockable rhythm is recognized, with the goal of maximizing the chance of return of spontaneous circulation.

Our results reflect the beneficial effects of RCDP as a teaching modality, as more residents trained with RCDP were achieving critical actions faster and having measurably better team performances than those taught using the traditional debriefing model. One possible explanation for this difference is that the cognitive load is lower during RCDP than in PSD. In RCDP, participants receive real-time feedback in digestible portions and in smaller time increments, followed by an opportunity to practice deliberately and accomplish specific tasks.¹⁶ The deliberate practice is key to creating an effective connection between behavior improvements. This also allows for a more meaningful understanding of one's actions, which may further improve retention. Possibly, one of the most beneficial effects of RCDP is that the technique reflects Kolb experiential learning theory, an educational theory in which much of medical training is modeled after. It includes a 4-step cycle of concrete experience, reflective observation, abstract conceptualization, and active experimentation. When these steps are done sequentially as they are in RCDP, it is intended to truly solidify learning.²⁶

This is in contrast to PSD in which general and specific feedback is given all at once without an immediate opportunity to implement behavioral change through deliberate practice. Attempts to recall specific circumstances that need improvement may not be successful, and some aspects of the feedback may become less effective. The other difference is how much time is spent on different aspects of pediatric resuscitation. Our RCDP residents spent more time practicing initial actions of pediatric resuscitation at the expense of less time on actions farther down the algorithm; residents in the PSD group spent more time performing or discussing subsequent actions.

The cohort of instructors who taught both RCDP and reflective debriefing sessions received formal training in both methods of debriefing and used scripted debriefs for both. However, these instructors may have varied in their individual teaching styles and delivery of feedback, arguably making our results more generalizable because we then captured a real-life application of the educational interventions, rather than their use under purely laboratory conditions. Moreover, despite having multiple video reviewers, the ICCs for resident team leader performance scores had excellent agreement.

Limitations

This study had several limitations. First, the return test case was designed to include 3 confederates with the learner serving as team lead. However, in 4 cases, the instructor team did not have a third confederate to assist, either because of scheduling conflicts or functionally because 1 instructor had to operate the computer for the simulation. Of those 4 cases, 3 of the learners had received simulation training with RCDP and only 1 with reflective debriefing. This may have altered the performance of the team lead, as the learner might have been occupied by tasks that they normally would not be responsible for if a third teammate were present. Secondly, there was some variability in team members for the initial training session, with some teams comprising of 5 and others 6 learners. Depending on the size of the team, some teams did not have a learner participant as a CPR coach, which may have altered the dynamics and choreography of the team. In addition, there was an institution-wide switch in defibrillators, from Physio-Control Lifepak 20 defibrillators to the Zoll R Series, which took place in March 2018, 2 months into our study. This switch affected 9 participants (28%), as they were tested using a defibrillator with which they were not initially trained. They were, however, evenly distributed between the 2 groups (5 RCDPs, 4 PSDs). Another limitation lies in the modified RTLE tool for team leader performance, because modification may affect its validity. Furthermore, the amount of time spent leading the team was less for individuals who achieved defibrillation earlier because the scenario ended once defibrillation was performed. However, our study showed that the RCDP group performed significantly better than the PSD group based on their RTLE scores. Although this study was not powered to compare RTLE and defibrillation times, future research studies should seek to understand how team leader performance affects outcomes. Moreover, this study tested team leadership. Most of the residents in the PSD group kept their assigned roles throughout the simulation and therefore did not receive the chance to be team lead, whereas RCDP gave residents the opportunity to rotate roles and thus practice in real team as team lead. This may suggest that the benefit seen in RCDP lies more in rotating the learners and offering first-hand experience than from the difference in training structure itself. Lastly, although the initial simulation education came in the form of team-based training, the follow-up testing focused on individual performance. Therefore, the first and second time points could not be directly compared for each individual, thus raising the question if retention is truly measured. Future studies could directly compare individual scores before and after a period in which decay of knowledge and skills might occur.

Future Directions

Future studies should focus on best applications for RCDP and its effectiveness with varying levels of learners and differing content. Rapid cycle deliberate practice should be tested against PSD after integration into the PALS curriculum. These should also assess knowledge and skill decay. In addition, research is needed to determine whether the benefits observed from RCDP are translated to clinical practice, resulting in clinically significant improvement in resuscitation performance and ultimately survival in children.

CONCLUSIONS

A skill is not considered truly learned until retention and/or transfer of that particular skill is demonstrated. This study demonstrated a significant difference in performance metrics in team leaders who were previously trained with RCDP compared with PSD. Residents trained using RCDP were more likely to achieve defibrillation faster and perform more effectively as team leader than those trained using PSD debriefing methods. Rapid cycle deliberate practice may be a superior method for training teams when choreographed and time-dependent actions are required. Such an advantage should encourage wider adoption and study of RCDP, which may provide better learning outcomes for trainees and improved clinical outcomes for patients.