White JRM, Shugerman R, Brownlee C, Quan L. Performance of Advanced Resuscitation Skills by Pediatric Housestaff. Arch Pediatr Adolesc Med. 1998;152(12):1232-1235. doi:10.1001/archpedi.152.12.1232
Copyright 1998 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.1998
To describe pediatric housestaff resuscitation experience and their ability to perform key resuscitation skills.
Cohort study of 63 pediatric residents in a university-based training program.
Participants and Methods
Investigators observed, scored, and timed resident performance on 4 key resuscitation skills. Cognitive ability was tested with 4 written scenarios. Housestaff provided self-reports of the number of months since their last American Heart Association Pediatric Advanced Life Support course, number of mock and actual codes attended, number of times skills were performed, and self-confidence with respect to resuscitation.
A total of 45 pediatric residents (71%) participated. Median cognitive score was 5 (range, 1-5). Of all residents, 44 (97%) successfully bag mask–ventilated the mannequin; 24 (53%) and 36 (80%) used the correct bag and mask size, respectively. Thirty-nine residents (87%) placed a tube in the mannequin trachea, 12 (27%) checked that suction was working prior to intubation, and 30 (67%) chose the correct endotracheal tube size. Forty residents (89%) discharged the defibrillator, and 25 (56%) and 32 (71%) correctly chose asynchronous mode and infant paddles, respectively. Thirty-eight residents (84%) inserted an intraosseous line; 35 (78%) had correct placement. Median times for successful skill completion were 83 seconds for bag mask ventilation, 136 seconds for intubation, 149 seconds for defibrillation, and 68 seconds for intraosseous line placement.
Pediatric housestaff previously trained in pediatric advanced life support were generally able to reach the end point of 4 key resuscitation skills but less frequently performed the specific subcomponents of each skill. This poor performance and the prolonged time to skill completion suggest the need for greater attention to detail during training.
THE PROVISION of early, appropriate resuscitation improves the outcome for the patient in cardiac or respiratory arrest.1 Effective cardiopulmonary resuscitation requires the physician to apply detailed cognitive information and to perform technical skills. Thus, physicians who participate in inpatient resuscitations, including residents, are expected to have a ready fund of the requisite knowledge and skills.
To prepare pediatric residents to perform effective resuscitation, many pediatric residency training programs teach advanced life support.2 Formal training is provided through various means, such as the American Heart Association's Pediatric Advanced Life Support (PALS) course. Informal training occurs during bedside teaching and mock resuscitations.3 Still, despite the time, resources, and expense dedicated to such education, little is known about the effectiveness of these efforts and whether training programs for residents are meeting their educational objectives.
Several studies have evaluated physician performance of cardiopulmonary resuscitation after formal training; however, most have primarily focused on the provision of basic life support for adults.4 These studies have made clear the need for training. Not surprisingly, without an initial period of training the performance of resuscitation skills is poor.5,6 Moreover, retraining is imperative because the retention of resuscitation skills is also poor.7- 9
Fewer studies have evaluated physician performance of advanced life support, which is an even more complex set of skills and knowledge-based decision making. As with basic life support, most of the studies of advanced life support have addressed resuscitation of the adult patient.9- 13 These studies have shown that knowledge and performance of advanced life support skills vary amongst different specialties, likely based on the amount of training inherent to their practice. Without reinforcement, most physicians will have lost skills as early as 6 to 12 months after training.7- 9
While these studies demonstrate the need for specific training and retraining of providers of adult resuscitation, information specific to pediatric providers is lacking. Unfortunately, extrapolation from adult to pediatric settings is difficult because adult and pediatric resuscitations are very different.14 Pediatric resuscitation relies on recognition and treatment of primarily respiratory rather than cardiac events.15 Inpatient pediatric resuscitations occur infrequently, providing fewer practical practice opportunities for housestaff.16 In addition, formal advanced life support training courses for adults differ in content, format, and teaching style from those designed for pediatric caretakers.
To our knowledge, there have been no studies to evaluate performance of advanced pediatric cardiopulmonary resuscitation after formal and informal training. Little is known about the factors that influence either the performance or the retention of the knowledge and skills of pediatric resuscitation. Without this information, it is difficult to structure effective training for housestaff and, hence, to delegate resources to concentrate efforts effectively.
The goal of our investigation was to describe the resuscitation experience and performance of PALS-trained pediatric housestaff who participate in resuscitations at a tertiary care pediatric referral center. We sought to quantify the amount of resuscitation experience, both actual and simulated, experienced by an average pediatric resident. Most importantly, we sought to test their ability to perform key resuscitation skills.
During the period from April to June 1996, written requests for study participation were sent to all 63 members of the pediatric residency training program at the University of Washington, Seattle. Each resident included in the study had previously taken a formal advanced training course in pediatric resuscitation (PALS). In addition, all residents at this training center participated in mock and actual resuscitations in the delivery room, the emergency department, and on the ward.
The study was conducted in the resuscitation room of the pediatric emergency department of a tertiary care hospital. All equipment within this room was available for the residents' use.
After giving informed consent to the investigator or the study nurse, each resident completed a questionnaire. In this questionnaire, residents rated their self-confidence in their ability to resuscitate on a scale of 1 (cannot perform any component of pediatric resuscitation) to 10 (absolutely confident of my skill in resuscitating a pediatric patient). They then listed the number of mock resuscitations, actual resuscitations, and resuscitation skill workshops they had attended thus far in residency. Actual resuscitations were defined as settings in which bag valve mask ventilation, intubation, intraosseous line placement, administration of resuscitation medications, or defibrillation was required. Residents subsequently completed a brief written test to evaluate their knowledge base.
Residents were then asked to perform 4 skills, including bag valve mask ventilation, tracheal intubation, intraosseous line placement, and defibrillation. Each resident was requested to perform these skills on the most advanced Laerdal infant mannequin (Laerdal, Wappingers Falls, NY), which included inflatable lungs, an intubatable airway, and an artificial tibia for intraosseous line placement. Reading from a prepared, written script for each of the 4 skills, the investigator described the mock scenario, the age of the patient, and the specific task forthe resident to perform (ie, "Your sole role in this resuscitation is to place an intraosseous line."). The study investigator emphasized the need for the resident to be as fast and precise as possible when performing skills.
Performance was scored on a standardized checklist designed specifically for this study. We defined an end point for each skill. The end points were chest rise for bag valve mask ventilation, discharge defibrillator for defibrillation, tube in the trachea for intubation, and intraosseous line stable in bone for intraosseous line placement. In addition, we compiled a group of subcomponents that were considered important for the optimal performance of each skill. Each of these subcomponents is listed as a specific teaching point within the PALS instructor manual. For example, correct bag and mask size were 2 of the subcomponents assessed when residents performed the skill of bag valve mask ventilation.
Using the study checklist, 1 of 2 investigators scored resident performance. The investigator noted whether residents met the end point for each skill, and whether residents demonstrated specific subcomponents of each skill. If the end point was reached, the investigator documented the skill as successfully completed. Correct performance of the subcomponents was not required to achieve the end point for each skill. Skill performance was timed; timing ended when the resident either indicated that he or she believed that the skill had been correctly performed, stated their inability to complete the task, or attempted the skill 3 times without success. Once the trial was complete, the investigator gave the resident feedback about his or her performance.
The components of the residents' cumulative experience with resuscitation were tabulated. The frequencies of the group's ability to perform each skill and its subcomponents were compiled.
The institutional review board of Children's Hospital and Medical Center in Seattle approved this study.
Forty-five residents agreed to participate; all had completed the PALS course. Overall, 18 (40%) of 45 participating residents had completed the course within 12 months of the study. Eighteen residents did not participate, stating lack of time or inconvenience. Participants and nonparticipants were similar with respect to year of residency training.
Residents reported performing a median of 4 mock resuscitations either in the emergency department or in the hospital during the course of their residency training. Residents reported participation in a median of 20 actual resuscitations, including those in the delivery room, intensive care units, and the emergency department. There were 18 first-year residents, 8 second-year residents, and 19 third-year residents.
Regardless of experience or year of training, residents performed well on the cognitive portion; the median cognitive score was 5 (range, 1-5). Residents rated themselves as "average" in their ability to resuscitate a patient, and the median self-confidence rating was 5 (range, 1-10). Self-confidence did not correlate with amount of training, experience, or cognitive score.
More than 36 (80%) of the 45 pediatric housestaff were able to achieve the primary end point for each of the 4 skills (Table 1). Of note, however, was the extremely variable performance by residents on the subcomponents of each skill. For example, 39 residents (87%) were able to place the endotracheal tube in the mannequin trachea; however, only 12 (27%) checked to see that suction was turned on prior to intubation, only 15 (34%) checked to see that the bag valve mask equipment was available, and only 30 (67%) chose the proper endotracheal tube size. Similarly, 40 residents (89%) were able to discharge the defibrillator, but only 25 (56%) chose the asynchronous mode for a patient they were told was in ventricular fibrillation (Table 1).
Each potential determinant of skills performance, including time since the PALS course, number of mock resuscitations and actual resuscitations attended, amount of skills practice, and self-confidence was evaluated for its effect on resident ability to perform a skill or the subcomponent of a skill. Using multivariate analysis, we were unable to demonstrate factors in a resident's training that had a significant effect on the ability to perform either a skill or its subcomponent. We were, however, limited by small sample size. With 45 participants, we had only 35% power to state that there are no statistically significant determinants of performance.
An important goal of pediatric training programs is to provide residents with the knowledge and skills to deliver effective patient care including resuscitation. The approach at our institution, similar to that at other institutions, has been to provide formal advanced pediatric resuscitation training as well as skill workshops and mock resuscitation practice.
The results of this study suggest that we are not adequately preparing our residents to perform resuscitation. Our residents performed well on cognitive testing but less well on skill performance. This finding is consistent with evaluations of adult advanced life support: knowledge and skills are learned differently.10,11 Once trained, physicians maintain cognitive knowledge more easily than the skills of providing life support.7,10
In our evaluation, we chose to study 4 skills that either are used frequently in pediatric resuscitation17 or are immediately lifesaving.18 All 4 skills are taught in the PALS course.19 We evaluated each skill individually rather than using a "megacode" format because this allows easier critical evaluation of each resident's performance. A previous study by Kanter20 suggests that evaluation of a study participant who is performing a single skill correlates with the participant's ability to perform the same skill in a more complex resuscitation situation. We chose the subcomponents listed for each skill based on their emphasis as teaching points in the PALS instructor manual.18
For the 4 skills tested, the majority of residents achieved the defined skill end point. Notably, however, residents frequently either omitted or incorrectly performed key subcomponents of each skill. This study does not make clear whether such imprecision represents poor retention of information previously learned or a lack of initial teaching. For example, although each of the subcomponents tested is included as a specific teaching point within the PALS instructor manual, it is possible that teacher variability or lack of emphasis may have prevented these points from being conveyed to the participant. Videotaped recordings of resuscitation training sessions have shown instructor inconsistency in teaching resuscitation skills and in addressing errors in student performance.21 In fact, because the PALS course is not intended to certify performance of skills, the specific skills of students completing the course are not guaranteed.
During this study, we measured time to skill completion. Definitive conclusions from these results cannot be drawn because there are no criterion standards against which to compare our findings. Nonetheless, it is clear that several factors prolonged the time to skill completion. Investigators noted that residents often spent many seconds "deciphering" equipment to perform the skills requested. For example, many participants were able to place an intraosseous line only after disassembling and reassembling the intraosseous needle several times to familiarize themselves with it. Prior to intubating, residents spent several seconds searching for "code cards" (pocket-sized cards with resuscitation information) to determine or verify correct endotracheal tube size, laryngoscope blade size, and other information. The use of anesthesia bags for ventilation was especially difficult for residents, and many participants had difficulty coordinating bag inflation and oxygen delivery. In fact, the use of this apparatus has been shown to be problematic in other studies of trainees, and led Kanter20 to recommend that the bags be used only by those who use them routinely, such as anesthesiologists and respiratory therapists. Finally, many residents relied on the written instructions on the defibrillator machine to guide them through the steps of defibrillation. This approach sacrificed time and may have contributed to the omission of steps not included within the instructions (such as remembering whether to synchronize and to check pulse). These observations suggest that many residents lack familiarity with resuscitation equipment and that more frequent exposure to equipment during skill workshops may improve performance.
Our study was limited by the artificial nature of a mock situation in which no other members of a resuscitation team were present. Although investigators specifically instructed residents to act quickly and efficiently, it is unlikely that the study setting provided the tension of a real resuscitation. Although there is no substitute for a real patient to challenge skill performance, our use of the most advanced infant mannequin provided an improvement over previous testing simulations.
Other limitations included the size of our sample population. Participation in this investigation, which was designed as a small pilot study, was limited to the number of residents available in our program. Because these residents were at variable points in their training, it is certain that they had had different experiences with resuscitation. We relied on resident self-reporting to determine amount of experience. In addition, as a baseline, we confirmed that all participants had attended a course in pediatric resuscitation, although our study was not specifically designed to evaluate resident performance after the PALS course.
We conclude that most residents, all of whom had received formal training in the performance of pediatric resuscitation, had a good cognitive base and were able to achieve the end point for each skill performance. Nevertheless, many errors and omissions were made en route to each end point. To best prepare residents to perform pediatric resuscitation, educators need to emphasize precise key steps and speed at the initial training and should require detailed practice of resuscitation skills during training.
We believe that larger studies, whether multicenter evaluations or those performed at single institutions with consecutive groups of incoming interns, would provide sufficient statistical power to more clearly assess the effect of experience on skill performance. Our study of resuscitation performance by the pediatric resident provides a baseline for future studies and helps to direct resources for effective resident training.
Editor's Note: While the findings of this study need to be validated with a much larger number of residents, the results are disturbing but not surprising. Practice does not make perfect; only perfect practice makes perfect.—Catherine D. DeAngelis, MD
Accepted for publication July 8, 1997.
This work was supported, in part, by a grant from the Washington affiliate of the American Heart Association, Spokane.
Presented at the annual meeting of the Ambulatory Pediatrics Association, Washington, DC, May 3, 1997.
We would like to thank Peter Cummings, MD, PhD, for his statistical support and Laerdal, Wappingers Falls, NY, for the resuscitation mannequin.
Corresponding author: Jeanette R. M. White, MD, The Johns Hopkins Hospital, Division of Pediatric Anesthesia and Critical Care, 600 N Wolfe St, CMSC 7-110, Baltimore, MD 21287-3711 (e-mail: email@example.com).