What can be learned from the literature about intervals and strategies for paediatric CPR retraining of healthcare professionals? A scoping review of literature

Background Effective training and retraining may be key to good quality paediatric cardiopulmonary resuscitation (pCPR). PCPR skills decay within months after training, making the current retraining intervals ineffective. Establishing an effective retraining strategy is fundamental to improve quality of performance and potentially enhance patient outcomes. Objective To investigate the intervals and strategies of formal paediatric resuscitation retraining provided to healthcare professionals, and the associated outcomes including patient outcomes, quality of performance, retention of knowledge and skills and rescuer’s confidence. Methods This review was drafted and reported using the Preferred Reporting Items for Systematic Reviews and Meta-analysis extension for Scoping Reviews (PRISMA-ScR). PubMed, Medline, Cochrane, Embase, CINAHL Complete, ERIC and Web of Science were searched and studies addressing the PICOST question were selected. Results The results indicate complex data due to significant heterogeneity among study findings in relation to study design, retraining strategies, outcome measures and length of intervention. Out of 4706 studies identified, 21 were included with most of them opting for monthly or more frequent retraining sessions. The length of intervention ranged from 2-minutes up to 3.5 hours, with most studies selecting shorter durations (<1h). All studies pointed to the importance of regular retraining sessions for acquisition and retention of pCPR skills. Conclusions Brief and frequent pCPR retraining may result in more successful skill retention and consequent higher-quality performance. There is no strong evidence regarding the ideal retraining schedule however, with as little as two minutes of refresher training every month, there is the potential to increase pCPR performance and retain the skills for longer.


Introduction
Survival from paediatric cardiac arrest is dependent on medical interventions including high quality paediatric cardiopulmonary resuscitation (pCPR). [1][2][3] However, pCPR quality frequently does not meet current standards. Long interruptions and incorrect chest compression depth and rate are some of the challenges, potentially impacting positive outcomes. [4][5][6] Effective training and retraining may be key to pCPR quality. Previous studies demonstrated that learners acquire CPR knowledge and skills irrespective of the method it is delivered [7][8][9][10][11][12][13] however, evidence shows that CPR skills decay within weeks to months after training, demonstrating that the current retraining intervals of one or two years is ineffective. 9,[14][15][16] This, coupled with paediatric cardiac arrest being an uncommon event, with an incidence of 8.04/100,000 for out-of-hospital cardiac arrests and around 1/1000 admissions for in-hospital cardiac arrests, further perpetuates the challenge in retaining pCPR skill. [17][18][19] Current resuscitation guidelines recommend a distributed practice model for teaching and learning CPR skills, however, there is no clarity over the optimal gap between training or retraining sessions. [19][20][21] It has been suggested that monthly retraining can enhance retention 16,[22][23] yet, this may not be feasible in clinical areas due to associated high costs, staff motivation and drop outs. 16,24 Previous reviews have explored retraining intervals for laypersons and spaced learning for resuscitation training, however these reviews did not focus on paediatric CPR. High-quality CPR has been associated with improved survival outcomes after cardiac arrest in the adult population. [25][26][27] Establishing an effective retraining strategy that facilitates learning and maximizes retention of pCPR skills is fundamental to improve quality of performance and potentially enhance patient outcomes after cardiac arrest.
This scoping review aimed to provide a contemporary synthesis of the literature exploring intervals and strategies of formal paediatric resuscitation training/retraining provided to healthcare professionals, and the associated outcomes including patient-level outcomes, quality of pCPR performance, retention of knowledge and skills, and rescuer's confidence. Since interventions such as "low dose, high frequency" [28][29] is not considered full retraining but short exposures to the skills, the term "refresher" will be used alongside retraining when appropriate.

Study design and protocol
In order to achieve the above stated aim, a scoping review was the preferred method. It enables to determine the scope of evidence available, provide an overview on key aspects underpinning the research area and gaps in literature to be identified. 30 This scoping review protocol was drafted and reported using the Preferred Reporting Items for Systematic Reviews and Metaanalysis extension for Scoping Reviews (PRISMA-ScR). 31 To the best of the authors' knowledge, there are no existing scoping or systematic reviews exploring the same or similar research question based in the paediatric population. Ethical approval was not applicable to this study.

Research question
The research question was based on PICOST (Population, Intervention, Control, Outcomes, Study design and Timeframe) and defined as: "What can be learned from the literature regarding strategies and intervals of paediatric CPR retraining provided to healthcare professionals in relation to patient outcomes; good quality performance; better retention of knowledge and/or skills; and/or improved rescuer's confidence?" P -healthcare professionals including doctors, nurses, EMS providers, Allied Health Professionals or any other healthcare professional working in any geographic location and any setting (prehospital, community and hospitals) undergoing formal pCPR retraining.
I -any form of formal pCPR retraining. C -different retraining or refresher intervals.
O -patient outcome; ability to deliver effective pCPR -simulated or real; knowledge and skill improvement; retention of knowledge and skills; and rescuer's confidence.
S -Primary studies (quantitative, qualitative and mixed-methods) including randomised controlled trials (RCTs), non-randomised controlled trials, interrupted time series, controlled before-and-after studies, observational and cohort studies were included in order to consider different aspects of measuring outcomes.
T -studies published between January 2005 and March 2022 (since the first publication of the 2005 guidelines on resuscitation by the ILCOR process, feeding scientific literature to the different Resuscitation Councils).

Eligibility criteria
All studies addressing the PICOST question were eligible, including paediatric manikin and/or simulation; paediatric basic life support (BLS) retraining; paediatric advanced life support (PALS); European paediatric advanced life support (EPALS); advanced paediatric life support (APLS); paediatric immediate life support (PILS); prehospital paediatric life support (PHPLS); and advanced trauma life support (ATLS). Studies based on neonatal CPR training/ retraining; adult CPR training/retraining; healthcare students; unpublished studies and studies in a language other than English were excluded from this review.

Search strategy
The following databases were searched by three researchers (DA, LT, TP): PubMed/Medline; Cochrane; Excerpta Medica Database (Embase); Cumulative Index to Nursing and Allied Health Literature (CINAHL) Complete and Web of Science. A pre-defined search strategy was used combining Boolean operators 'AND' and 'OR' with medical search headings and subheadings (e.g. MESH) when applicable. The search terms (Appendix 1) were drafted by the research team and revised by an experienced librarian. The reference lists from included sources were manually searched to identify any further studies not yet captured.

Study identification and selection criteria
All articles initially identified were sent to the web-based bibliographic manager (EndNote Desktop X9) where duplicate references were removed. To increase consistency, two reviewers (DA and LT) screened the identified sources for relevance by evaluating the titles and abstracts according to the proposed eligibility criteria. Disagreement was resolved by consensus, moderated by a third reviewer (TP). If during abstract screening suitability could not be determined, further evaluation of the full text was performed, at which point, those studies that did not fit the eligibility criteria were excluded.

Data extraction and analysis
Data from included full text sources were extracted and organised in an Excel spreadsheet using a "descriptive-analytical" method within a pre-set framework 32 to ensure that variations between studies were uniformly captured and described. Study identification (first author, title, DOI, year of publication, geographic location); study design (context, sample size, intervention, duration); participants (profession, setting); training/retraining details (BLS, PALS, EPALS, PILS, etc.); outcomes (knowledge, confidence, ability to deliver effective CPR, skill retention, etc.); methods of assessment; results; and conclusion were summarised for further analysis.

Results
The initial search resulted in 6272 studies. Of those, 1566 were duplicates, resulting in 4706 titles post-deduplication. After reading titles and abstracts, 134 were selected for full text review. Of those, one could not be retrieved and 112 were discarded for not fully fulfilling the inclusion criteria, leaving 21 studies included in the analysis, as seen in Fig. 1.

Study characteristics
Study characteristics and interventions are summarised in Table 1. Over 3000 healthcare professionals were involved including nurses, paediatric residents, emergency medicine residents, EMS providers, physicians, respiratory therapists and pharmacists.

Retraining intervals and number of retraining sessions
The studies selected for this review used different timeframes for the first reinforcement session after initial pCPR training. Two studies performed the interventions straight after training 28,40 and one study  Target of high-quality CPR for more than 70% based on depth and rate As the number of training sessions increased, the percentage of CPR in the target range also increased, with less variability in performance. 29% with no training 46% after 1 session, 54% after 2 sessions, 68% after 3 sessions, 74% after 4 sessions (p = 0.001). Median percentage of time in the target area was 68% (interquartile range [IQR], 64-72) among the experienced trainees and 48% (IQR, 43-59) among novice trainees; p = 0.002).
Repeated short refresher with RTF significantly increase performance Braun et al. 38 (2015), USA RCT, 42 paediatric residents, hospital Baseline performance; Repeated scenario as needed until mastery-level performance was achieved (1 h to 2 h to achieve mastery); Retest 2, 4, or 6 months later Retention of masterylevel performance Percentage of residents maintaining mastery-level performance showed a significant linear decline (p = 0.039), with a drop at each retesting interval. 92% retained mastery at 2 months; Significant improvements in resuscitation performance after a single simulation-based mastery learning session. However, performance declined over time.  Control group: baseline knowledge assessment + self-confidence survey; mock code at 9-months; knowledge assessment and self-confidence survey at 11-months; Intervention group: baseline knowledge assessment and selfconfidence survey; simulation with debriefing at months 1, 5 and 9; knowledge assessment and selfconfidence survey at month 11.
knowledge, skills, selfconfidence No statistically significant difference between groups at baseline; Statistically significant difference in the post-test scores (p = 0.016) with the intervention group scoring higher than the control group. No statistically significant differences in self-confidence or final scenario between the groups.
Repeated paediatric mock code simulations with structured debriefing can be an effective method to educate CPR skills. "Rolling Refresher" bedside CPR skill training approach using "just-in-time" and "just-in-place" education is effective and well received by PICU staff.
Ojha et al. 48    Paediatric mock codes can improve resident confidence and selfassessment of their resuscitation skills.
requested participants to return one hour after initial training. 38 Four studies brought their participants back one week after training 33,36,43,49 and three studies two weeks after training. 22,42,48 Most study designs had the first refresher intervention at one month after training; 34,37,39,[45][46]51 one study after two months; 50 one study after three months 41 and two studies brought their participants back after four months. 44,47 The number of refresher interventions throughout the study periods also varied considerably, with most studies doing monthly refresher sessions. [34][35]37,[45][46]51 Five studies offered just one refresher session after initial training; 33,38,40,44,50 three studies had weekly re-exposure of pCPR skills 36,43,49 and further three studies fortnightly. 22,42,48 Two studies offered refresher sessions every-two or three months 28,41 and one study every-four months. 47 Interestingly, one study offered participants unlimited refresher opportunities, however, it resulted in no lasting improvement of the outcome measures. 39 Based on the number of refresher interventions throughout the study periods, it was noted that the outcome measures had similar positive results for most variables in each study included, apart from five of six studies that offered just one refresher intervention after training 38,40,44,50 Length of intervention during retraining sessions The length of intervention in each retraining session varied considerably, ranging from 2-minutes up to 3.5 hours. For easier identification and analysis, the studies were grouped as short duration (<1h) 22,28,[34][35][36][37][36][37][38][39][40][44][45][46][47][48][50][51] and long duration (1h). 33,38,43,49 Four studies 34,41,44,47 did not specify the length of retraining intervention, however, the interventions consisted of mock codes with/without debrief 34,44,47 and self-directed short skill training 41 and for this reason, the studies were added into the "short duration" interventions.

Long duration (retraining interventions lasting 1 h)
Four studies used longer lengths of interventions when retraining participants, with two of them lasting 1 h, 33,38 one lasting 90 minutes 43 and one lasting 3.5 h. 49 The outcome measures included skills; 33,43,49 knowledge; 49 retention of skills; 38,49 and confidence. 49 Only one study 38 did not improve the outcome measure at follow up. Results from interventions are described in Table 1.

Discussion
This research has broadly and systematically identified and analysed studies relevant to retraining schedules of paediatric resuscitation skills for healthcare professionals. The International Liaison Committee on Resuscitation (ILCOR) has stated that regular pCPR skills updates are important however, the ideal retraining interval has yet to be established as evidence is limited in both quantity and quality. 19 This review aimed to contribute to current knowledge for a further understanding of the challenges of pCPR learning and retention, from which future research can be planned. Although the researchers believe that this review will be the first step to map the gaps in knowledge and the consensus around pCPR retraining intervals for a broad overview of evidence, future research should investigate existing knowledge gaps associated with paediatric CPR training and retraining. Cost effectiveness is an important aspect that requires further exploration, particularly when retraining or refresher is delivered during clinical practice. Little is known around the impact of training or retraining strategies on patient-level outcomes as previous studies investigating variables such as survival to hospital discharge or neurological outcome are limited in both quality and quantity. 19 Another aspect that warrants further research is whether the same retraining interval is applicable throughout the career, and whether this should vary according to the skills being trained (e.g. chest compressions, ventilation, intubation). Furthermore, assessment of the optimal strategies to team-based training, nontechnical skills and leadership skills would be an important addition to the current evidence to paediatric CPR training and retraining.
In this review exploring retraining schedules of paediatric resuscitation skills for healthcare professionals, although the inclusion and exclusion criteria were well defined, the findings reveal complex data with studies that do not fit precisely into the categories. Despite internal quality assurance and transparency in reporting, identifying the time schedule and length of interventions was not simple, due to differing study methods and interventions. Additionally, the lack of clarity in some studies regarding the length of retraining interventions, made it more difficult to analyse the results.
The included studies demonstrated that the initial acquisition of pCPR skills is similar, irrespectively of the training model used. Different strategies were observed, including the traditional instructorbased training; 22,28,33,41,45,49 simulation-based mock code program; 34,47,51 simulation; 32,36,38,44,48 distributed practice; [45][46]48 and training with real-time feedback. 28,33,37,[39][40][42][43]50 Although the learning outcomes were similar, in the sense that learners acquired the skills, a better understanding of the impact of instructional designs on learning outcomes would enable researchers to design training programmes that translate into effective performance during real resuscitation attempts. 20 This is supported by evidence from recent reviews exploring training strategies to improve CPR performance and patient outcomes. [52][53] Lauridsen and colleagues 52 explored in their scoping review, different types of CPR training for healthcare professionals. They included aspects such as training approaches (e-learning, instructor-based, virtual reality, simulation, gamified learning); training duration and intervals; equipment and feedback (manikins, feedback devices, debriefing). The authors concluded that there is growing evidence advocating online learning and low-dose, high frequency CPR training to acquire CPR knowledge; the use of feedback devices to improve the quality of CPR skills; and team-based simulation with debriefing to enhance team perfor-mance manging a cardiac arrest. Additionally, Yeung et al. 53 conducted a systematic review comparing spaced learning with traditional massed learning to investigate whether spaced learning strategy improves educational and clinical outcomes. Although no conclusion can be made regarding patient outcomes, the results from their review suggested that spaced learning is more effective than massed learning for performance of CPR skill after training and at follow up. Our current scoping review adds to this body of evidence but provides additional contribution through the specific focus on paediatric CPR. Similar findings were determined suggesting brief and frequent practice enhances learning of paediatric CPR.
The best training and retraining designs should be tailored to specific learning objectives, learner type and needs, or context of learning. There are recommendations related to the use of deliberate practice, mastery learning, booster training, in situ education, realtime feedback and other strategies for training and retraining. 9,20 However, despite initial learning acquisition straight after training, CPR skills normally decay within weeks after initial training. 16,54 Although some providers retain CPR skills through recurring exposure to managing cardiac arrests as part of their clinical practice, 55 most paediatric providers go through long periods of clinical practice without performing pCPR due to the low incidence of cardiac arrest in this population. 18 Therefore, it becomes very important to establish retraining intervals to ensure that pCPR skills are maintained for longer.
In this review, all studies pointed to the importance of regular retraining sessions to the retention of pCPR skills. Although there is no consensus on the optimal interval, most of the studies opted for refresher sessions on a monthly basis [34][35]37,[45][46] or more frequently (weekly, fortnightly). 22,[36][37]39,[42][43][48][49]51 Their results indicate that frequent sessions enhance simulated pCPR outcomes with significant improvement in the outcome measures (survival rate, skills, knowledge, retention, and/or rescuer's confidence). Conversely, the studies with less than monthly refresher sessions 28,33,38,[40][41]44,47,50 suggest non-improvement, or improvement in one aspect but not others, or decline of skills at follow up, with one study suggesting that retention of skills was more likely with more refresher sessions. 28 The length of intervention in each retraining session varied considerably between the studies, ranging from two minutes up to 3.5 hours. Nevertheless, it was demonstrated that, with as little as two minutes of refresher session every month, there is the potential to increase pCPR performance and retain the skills for longer. However, this cannot be directly associated with patient-level outcomes.
Although regular updates are beneficial to retention of skills, frequent retraining sessions can be associated with high dropout rates. 16 This, aligned with significant increased costs of repeated retraining and backfilling of staff in clinical areas, may affect the viability of a high frequency training in practice. Therefore, an effective balance between retraining and sustainability has yet to be established. To reduce the burden and costs of moving practitioners away from clinical areas for lengthy pCPR retraining, short duration of intervention (as established by the current study) was adopted by most researchers included in this review. 22,28,[34][35][36][37][39][40][41][42][44][45][46][47][48][50][51] This is in alignment with other research exploring the benefits of low-dose and high-frequency or distributed practice, suggesting that retention of CPR skills may be optimised and costs reduced, by training sessions with short interventions. 14,23,28, 46 Lin and colleagues 56 investigated cost-effectiveness and outcomes of distributed paediatric CPR training using real-time feedback. Their results suggest that this strategy is associated with improved CPR quality and decreased training costs when compared with conventional annual mass CPR training. Despite this, further research is needed to investigate whether distributed practice affects the need for subsequent retraining intervals.
The use of feedback devices during retraining was observed in many studies. 28,37,[39][40][41][42][43]46,[49][50] Previous research has established the benefits of such devices during adult and paediatric CPR training. 57 This was also observed in this review, with the majority achieving an improvement in outcome measures. Nevertheless, although the use of real-time feedback devices has been associated with enhanced performance during CPR training, there are conflicting interpretations regarding its efficacy during real-life resuscitation attempts. 58 Therefore, whilst it may be intuitive to presume that real-time feedback devices can improve patient outcome, this is yet to be established.
This review has demonstrated that brief and frequent pCPR retraining using simulation and additional tools such as real-time feedback devices can potentially develop skills, knowledge and confidence in pCPR performance. It is also noted that increasing the frequency of retraining sessions may result in a more successful skill retention and consequent higher-quality performance. Despite this, there is no strong evidence regarding the ideal retraining schedule. It is suggested that a more nuanced approach to pCPR retraining, based on specific learning objectives, context and learners' needs and/or performance is recommended in an attempt to maximise skill retention and improve pCPR performance.

Limitations
This study has some limitations. First, potential biases were not systematically addressed like in a systematic review. Second, the heterogeneity among study design, retraining strategies, outcome measures and length of intervention, may impact the interpretation and synthesis of the results. Third, most studies were performed in a simulated, controlled environment, making it difficult to extrapolate the results to real-life CPR performance. Fourth, despite improvement in pCPR quality and retention of the skills, the results cannot be directly associated with patient-level outcomes. Fifth, this review only includes studies after 2005, therefore, it is not known if other important evidence exists prior to 2005. Finally, despite exhaustive attempts to locate every relevant resource, one study was identified but could not be retrieved.

Conclusion
Brief and frequent pCPR retraining may result in a more appropriate skill retention and consequent high-quality performance. There is no strong evidence regarding the ideal retraining schedule however, it was demonstrated that, with as little as two minutes of refresher training every month, there is the potential to increase pCPR performance and retain the skills for longer.