Neonatal resuscitation guideline adherence: simulation study and framework for improvement

Approximately 5–10% of all newborns require support of transition to initiate breathing and aerate the lungs; true resuscitation at birth, including chest compressions and medications, occurs significantly less often [1‐7]. Inadequate initial cardiorespiratory support may inflict pulmonary damage and may conduce to ongoing hypoxia/ischemia, possibly aggravating cerebral injury. The European Resuscitation Council (ERC) and American Heart Association (AHA) publish guidelines for the resuscitation of newborns at birth [1, 8, 9]. Healthcare professionals qualified to perform neonatal resuscitations are expected to abide by these guidelines. By attending the Newborn Life Support (NLS) and/or European Pediatric Advanced Life Support (EPALS) courses of the ERC, or the Neonatal Resuscitation Program (NRP) outside Europe, professionals can acquire and maintain the knowledge and skills required for guideline adherence. In the Netherlands, the NLS course is obligatory for pediatric residents. Many Dutch pediatricians also participate in this course. There is some, though limited evidence that structured neonatal resuscitation training improves patient outcomes [3, 4, 10].

A prerequisite for the positive effect of resuscitation guidelines on patient outcomes is adequate adherence to these guidelines. Previous studies looked at the adherence to neonatal resuscitation guidelines using video recording of delivery room management. Carbine et al. reported that 54% of neonatal resuscitations deviated from guideline recommendations [11]. Other studies also revealed insufficient compliance with the guidelines [2, 3, 6, 10, 12]. More complex resuscitations were associated with poorer adherence [2, 11]. Commonly reported deviations were overly vigorous stimulation, undue/inadequate suctioning, inadequate/unnecessary positive pressure ventilation, unjustified delivery of oxygen, prolonged/multiple intubation attempts, not completing tasks within the allocated time intervals, and overestimated Apgar scores. Most studies could not reliably report on the adequacy of chest compressions (CC), because resuscitations requiring compressions were infrequent. Crew resource management (CRM) skills during neonatal resuscitations also deserved attention [6, 10].

Video-based assessment of simulated or real-life neonatal resuscitations proved to be feasible and informative in the aforementioned studies. However, video recording, as the only means of assessment, has its audiovisual limitations [13], and does not provide objective, quantitative data on vital physiological parameters [14]. Therefore, we evaluated the adherence of Dutch pediatricians to the NLS guideline of the ERC in a simulated setting, using a scenario running on a model-driven, high-fidelity neonatal patient simulator. This allowed for the measurement of several resuscitation quality metrics, including airway patency, applied airway pressures, and CC. We also assessed pediatricians’ NLS knowledge and reviewed strategies to improve (neonatal) resuscitation guideline adherence.

Data were collected during a 3-day continuing medical education (CME) event for general pediatricians at our center. The main theme of the event did not involve (neonatal) resuscitation, so participants did not specifically prepare for participation in our study. This study took place just prior to the latest update of the ERC guidelines in 2015. Hence, we used the 2010 ERC guideline (and its Dutch equivalent) as a reference [1]. The only changes in the 2015 ERC guideline relevant to this study were (1) the more explicit recommendation to apply a higher peak inspiratory pressure (PIP) (i.e., 30 cm H₂O) during inflation breaths in term newborns, (2) the gestational age below which plastic wrapping is indicated (< 32 weeks instead of < 28 weeks), and (3) the option to start with 30% oxygen in preterm infants [8].

This study revealed that the adherence of Dutch pediatricians to the NLS guideline was suboptimal. In line with the publication by Yamada et al. [2], guideline deviations can be separated into errors of commission and errors of omission. We mainly encountered errors of commission as far as technical skills (compressions and ventilations) were concerned. Inflation breaths were started too late, they lasted too short, continuous airway patency was infrequently achieved, and the majority of CC were ineffective. On the other hand, inspiratory pressures and epinephrine dose were mostly correct. Considering the fact that CC rate and events/min were comparatively high, it appears that rapidness occurred at the expense of effectiveness. Errors of omission occurred less often. A quarter of the pediatricians did not increase the amount of oxygen when CC were initiated, as prescribed in the Dutch NLS guideline and suggested in the ERC guideline. Only half of them took full measures to maintain body temperature. Pulse oximetry was applied by the majority of our participants and all assessed initial heart rate by auscultation. Looking at the error types, healthcare professionals are apparently in need of reminders/prompts to ensure execution of all steps of the algorithm, whereas they require real-time quantitative feedback to guide their performance of technical skills.

Overall, it was interesting to witness a large variation in algorithm execution and skill performance among the pediatricians. For example, time to recovery varied between 4.5 and 12 min. Our findings substantiate the results of previous researchers in that neonatal resuscitation guideline adherence is rather low, and that ventilation errors occur frequently. In addition, we found that circulatory support also needs improvement. The ventilatory flaws are perhaps more intriguing than the circulation-related inadequacies. After all, pediatricians are more exposed/used to assisted ventilation than to compressions and medications, as (respiratory) support of transition is needed far more often than extensive resuscitation.

NLS knowledge was also suboptimal. Seventy-one percent of the multiple-choice questions were answered correctly. The cut-off point for passing the knowledge assessment of the NLS course is 80%. From a database, held by the Dutch Resuscitation Council and the Dutch Foundation for the Emergency Medical Care of Children, containing data from 2016 to 2019, we learned that 97% of all Dutch NLS participants passed the post-course MCQ. Median (IQR) test scores on the pre-course and post-course MCQ were 90% (86–96%) and 90% (86–94%), respectively. These percentages are higher than the 71% found in our study, probably because the pre-course and post-course results were achieved following manual reading and the NLS course, respectively, whereas pediatricians were tested ad hoc in the current study. Pediatric residents at the University of Colorado (USA) had a pass rate of 79 ± 3% on a pretest that assessed NRP knowledge [17]. All these residents reviewed NRP in the preceding 6 months, so they had a relatively recent booster. Some questions in our MCQ, e.g., the ones on laryngeal mask airway use and delayed cord clamping, were perhaps a little less essential for the practice of NLS by general pediatricians at that time. However, various pediatricians appeared to be unaware of important information regarding ventilations and compressions.

The NLS certification rate of our participants was comparatively low (46%). The majority (85%) of our participants did attend the APLS course. However, the Dutch APLS course only contains background information on and a demonstration of NLS, but does not incorporate practical rehearsal of NLS skills. Moreover, a median of 4–5 years had elapsed since our participants attended their last certified NLS or APLS course. The recertification interval for NLS and APLS is 4 years in the Netherlands. Unfortunately, we were not informed about the rate and nature of the neonatal resuscitation simulation training that our participants attended at their local hospitals. We know that they usually rehearse NLS scenarios at least once a year using a lower fidelity manikin and environment. In view of our results, pediatricians probably need to follow more frequent booster training to maintain their resuscitation capabilities. Annual refreshers may not be sufficient, for it is well known that resuscitation skills deteriorate within 3–6 months after initial training without regular practice [34].

The identification of guideline deviations is especially important in planning interventions to optimize algorithm adherence and quality of skill performance [2, 3]. Instead of only focusing on errors, we intended to come up with solutions as well. We therefore decided to create the framework as depicted in Table 4. This framework is meant to inspire clinicians, researchers, medical technicians, manufacturers, medical educators, policy makers, and course and guideline developers. It is not a formally validated tool to be implemented as a whole at any particular department. It should serve as a guide or “checklist” for institutions and departments that seek to enhance (neonatal) resuscitation guideline adherence. Departments may select the interventions that are relevant and feasible for them. The framework contains recommendations to improve both the stepwise execution of the algorithm and the performance of the skills therein, in order to redeem errors of omission and commission, respectively. For example, pocket cards, prompts, mnemonics, and decision support tools may ameliorate the stepwise process, while regular hands-on practice with feedback (devices) and the provision of correct equipment are measures to ensure adequate skill performance. Having incorporated a column with examples and/or extra information, we believe that our framework is self-explanatory. Some of the strategies are quite innovative and require advanced technology to effectuate them. However, many other strategies are less resource-demanding and can be realized in basically every hospital. Strategies that involve the use of video recording in the delivery room may raise ethical concerns [14]. These should be addressed before implementation. In spite of the framework’s size, we would like to emphasize that one should not lose sight of the pivotal role of repeated hands-on practice to ensure proper psychomotor skills. Also note that our framework does not exclusively pertain to NLS; it is probably also useful for the improvement of adherence to pediatric and adult resuscitation guidelines.

Our research group is currently working on follow-up studies, in which one or more strategies, taken from our framework, will be employed in an attempt to improve resuscitation guideline compliance. A very promising strategy, in our opinion, is the application of electronic decision support tools [4, 18]. We are planning to implement such a tool at our department, and will hopefully report on its use during delivery room resuscitation in the near future. A bundle of interventions from our framework will be used in a study that explores ways to enhance adherence to the airway, breathing, circulation, disability, and exposure (ABCDE) algorithm during the assessment of critically ill patients. Finally, we are investigating if use of a mixed reality device (i.e., the Microsoft® HoloLens™) will improve NLS guideline adherence.

The adherence of Dutch general pediatricians to the NLS guideline was suboptimal in our simulated setting. Their knowledge of important aspects of resuscitation at birth also requires attention. In an attempt to improve resuscitation guideline adherence, we constructed a comprehensive framework containing multiple suggestions to this end.