Objective. To determine the compliance to Neonatal Resuscitation Program (NRP) guidelines in our institution, by the use of videotaped newborn resuscitations.
Background. NRP is the standard of care for newborn resuscitation. The application of NRP guidelines and resuscitation skills in actual clinical settings is undocumented.
Design/Methods. A video recorder, mounted to the radiant warmer in the main obstetrical operating room, was used to record all high-risk resuscitations. All members of the resuscitation team were NRP-certified. The videotapes were reviewed within 14 days of the resuscitation and then erased. This ongoing review was approved as a quality assurance (QA) project ensuring confidentiality under California law. The first 100 resuscitations were evaluated to assess NRP compliance. Each step in the resuscitation (positioning, oxygen delivery, ventilation, chest compressions, intubation, and medication) was graded. A score was devised, with 2 points being awarded for every correct decision and proper procedure, 1 point for delayed interventions or inadequate technique, and zero points for indicated procedures that were omitted or for interventions that were not indicated. The total points were divided by the total possible points for that patient. The scores for the first 25 resuscitations (group 1) and the last 25 resuscitations (group 2) were compared.
Results. Fifty-four percent of the 100 resuscitations had deviations from the NRP guidelines. Ten percent received overly aggressive stimulation and 22% had poor suction technique. Of the 78 infants given oxygen, this decision was considered incorrect in 15% and the delivery technique was poor in 10% of the infants given oxygen. Of those requiring mask ventilation (n = 18), 24% had poor chest expansion, 11% used an incorrect rate, and 17% had inadequate reevaluation. Twelve infants were intubated; only 7 were successfully intubated on the first attempt and only 4 were intubated in <20 seconds. The longest intubation attempt was 50 seconds. Naloxone was given to 2 patients. One was breathing spontaneously with a heart rate >100. Resuscitations receiving a perfect evaluation score were more likely to occur in infants needing less intervention. The level of resuscitation required for groups 1 and 2 were statistically similar. There was no difference in resuscitation scores between the 2 groups. Only the inappropriate use of deep suctioning improved, with 8 of 25 events in group 1, and 0 of 25 in group 2.
Conclusions. We have found a significant number of deviations from the NRP guidelines. Video recording of actual clinical practice is a useful QA tool for monitoring the conduct of newborn resuscitation. We are now conducting repeat video assessments of individual NRP providers to determine whether there is improved performance.
The Neonatal Resuscitation Program (NRP) was developed by the American Academy of Pediatrics and the American Heart Association and endorsed in 1987.1 The course has been widely taught in the United States and has become the standard of care in newborn resuscitation. Competency in the course material is demonstrated by successful completion of a written examination, a performance checklist, and a megacode. Renewal is required every 2 years. The assumption is that the NRP guidelines will be followed by individuals who have completed the course. In 1998, Kaczorowski et al2showed a significant deterioration of resuscitation knowledge and skills, when providers were retested 6 to 8 months after completion of the NRP course. The NRP program does state that completion of the course material does not ensure that the student can successfully resuscitate an infant in an actual clinical environment. The crucial step from completion of this program to actual clinical competence has not been adequately evaluated.
Video recordings were first reported as an educational tool in teaching emergency medicine in 1969 by Peltier et al.3 Hoyt et al4 reported effective use of videotaping trauma resuscitation in 1988, and demonstrated that videotape critique sessions significantly improved house staff performance over a 3-month period, compared with controls. Videotaping trauma and cardiac arrests in emergency departments has become accepted practice in the United Kingdom and Australia.5–7 To date there have been no reports of the use of videotaping as a means of assessing the quality of neonatal resuscitations.
We believed that the use of videotaping of neonatal resuscitation would allow us to: 1) determine the actual conduct of neonatal resuscitation in our institution; 2) compare that resuscitation against the standards set forth by the NRP program; and 3) reeducate and improve performance.
The UCSD obstetric service delivers ∼200 infants per month. This is a high-risk service with a disproportionate number of deliveries requiring intervention. A 40-bed level III regional neonatal intensive care unit (NICU) adjoins the labor and delivery suite. All high-risk deliveries are attended by at least 1 of the following: a staff neonatologist, neonatal fellow, neonatal nurse practitioner, or a senior pediatric resident (PL-2 or -3). The remainder of the resuscitation team includes a neonatal nurse, neonatal respiratory therapist, and a pediatric intern (PL-1). All members of the team are required to have a current NRP course completion card.
Resuscitation is performed on overhead radiant warmers in the each obstetric OR and delivery room. An 8-mm video recorder (Sony Hi8 Model #EVC200, Sony Corp, Tokyo, Japan), which also records sound, was permanently mounted to the radiant warmer (Ohmeda, Columbia, MD) in the main obstetrical operating room. The recorder was mounted ∼1 meter above the warmer and was zoomed so as to provide a field of view that included the entire infant and the hands of the resuscitation team. The video recorder displays a continuous date and time readout at the bottom of the recorded image allowing timing of performed procedures to the nearest second.
As video recording and evaluation are performed as a quality assurance (QA) exercise, confidentiality is assured under California evidence code sections 1156 and 1157. The project was approved by the University of California, San Diego Medical Center as a QA project without the requirement for parental consent.
The recorder is loaded with a fresh tape and switched on by the respiratory therapist when the resuscitation team arrives. The tape is removed after the resuscitation. Every 2 weeks these tapes are reviewed by an ad hoc committee whose representatives include at least 1 of the researchers, at least 1 physician, and representatives from nursing and respiratory therapy. The meetings are open to all NICU staff including house staff.
An evaluation sheet was designed to evaluate adherence to NRP guidelines and is completed for each resuscitation. Each step of resuscitation was evaluated. These included: the equipment check, positioning and stimulation, administration of oxygen, bag-mask ventilation, chest compressions, intubation, and administration of medications. Each step was scored as to whether an intervention was provided and whether that decision was appropriate. If an intervention was provided, the procedure was then evaluated for technique, success, and proper reevaluation of the patient. Problems that were not covered by the checklist were noted under the comments section. Items frequently noted in the comments section were the lack of communication regarding the infant's heart rate to the lead resuscitator, the number of deep suction procedures, the time for completion of intubation, and the number of intubation attempts. Current NRP guidelines were used as the objective measure by which scoring decisions were based. Each resuscitation was graded by the overall consensus of the committee members present.
The results of these reviews are discussed at the monthly NICU QA meetings. After formal review, these tapes are erased. Selected excerpts are retained for teaching purposes in monthly house staff lectures, the annual house staff NRP course, and for review with individual house officers when appropriate.
One hundred nonconsecutive resuscitations have been evaluated to date. All 100 resuscitations were evaluated to assess the relative frequency of NRP protocol deviations and to review the actual times and duration of procedures including intubation.
To assess the role of videotaping as an educational tool, the first 25 resuscitations (group 1) and last 25 resuscitations (group 2) were compared. Each level of resuscitation was evaluated and scored. A composite score was devised to assign a numerical score for each resuscitation. Two points were awarded for every correct decision and every properly performed procedure. One point was awarded if the intervention was delayed or the technique for a given procedure was inadequate. No points were awarded for indicated procedures that were omitted or for performed procedures that were not indicated. The sum of the awarded points was divided by the total possible points for that patient to obtain a percent score.
The scores for both groups were compared and tested for statistical significance using analysis of variance. The level of required intervention and individual components of the resuscitations were compared and tested for significance using the χ2 test. Statistical significance was defined as P ≤ .05.
One hundred nonconsecutive resuscitations were reviewed over a 9-month period (January 1999 to September 1999). Group 1 resuscitations were conducted and graded from January 1999 to March 1999. Group 2 resuscitations were conducted and graded from July 1999 to September 1999. Using the described scoring tool, 46/100 were deemed perfect resuscitations (ie, followed NRP guidelines without deviation). The likelihood of a resuscitation being conducted perfectly was significantly related to the level of care required. Less complicated resuscitations were more likely to be conducted without deviation (Table 1).
Because of this association, group 1 and group 2 were compared as to the degree of resuscitation required. The relative number of infants requiring blow by oxygen, bag-mask ventilation, intubation, chest compression, and medication were compared and were statistically similar. There was a nonstatistically significant trend for less intervention being required in group 2 (Table 2). There were no significant differences in test scores or number of perfect scores between the 2 groups. When each component of resuscitation was analyzed separately, only 1 aspect of resuscitation improved significantly in group 2. This was the use of deep nasogastric and/or orogastric suctioning with a suction catheter. This was frequently applied during resuscitation, often before the infant was fully stable, and repetitive deep suctioning was common. There were 8 infants who had deep suctioning in group 1. Five infants had 1 suction each, 1 had 2, 1 was suctioned on 4 occasions, and 1 infant was suctioned 8 times. There were no episodes of deep suctioning in group 2 (P = .002). No other aspects of resuscitation differed significantly between the 2 groups. The resuscitation scores are presented in Table 3.
The relative frequencies of NRP protocol deviations were analyzed for all 100 resuscitations. Those deviations that occurred >10% of the time are presented in Fig 1. The percent of incorrect actions for each resuscitation category is shown in the graph.
Common minor deviations included: overly aggressive stimulation, enthusiastic drying which interfered with attempts at suctioning, and the heart rate being checked but not clearly being communicated to the lead resuscitator.
There were a number of recurring problems with endotracheal intubation. The recommended allotted time for intubation is ≤20 seconds. We were able to time all intubation attempts to the nearest second. Of 12 attempts, only 4 (33.3%) were achieved in <20 seconds. The longest recorded attempt was 50 seconds. There was also a high rate of failed and or multiple attempts at intubation, with only 7 of 12 infants (58.3%) being successfully intubated on the first attempt. NRP protocol violations for intubation are presented in Fig 2.
As expected, there were very few neonates who required medication during the initial resuscitation. Only 1 infant required epinephrine. Of 2 infants whom were given naloxone, only 1 met appropriate criteria. The other infant was on free-flow oxygen with spontaneous respirations and an adequate heart rate and color at the time of naloxone administration.
Finally, we found that Apgar scoring often did not correlate with the observations of the reviewers. The Apgar assessment was not included as part of the score as the NRP decision sequence does not use the Apgar score. However, discrepancies were frequently observed between Apgar scores and the actual condition of the infant. These were noted in the comments section of the scoring sheet. The discrepancies were primarily because of errors in the estimation of actual time of assignment of the Apgar score, with observed delays of as much as 30 to 60 seconds.
We present this data as the first prospective real time analysis of neonatal resuscitation in a clinical setting. To date, there have been no other studies that have evaluated the actual application of neonatal resuscitation in the delivery room environment. By presenting our observations, we hope to illustrate the potential of this relatively simple QA tool.
The research by Kaczorowski et al2 demonstrated that retention of NRP skills deteriorates rapidly after completion of the course. Residents who had successfully completed and passed the NRP course were eligible for the study. The residents were divided into 3 groups. Group 1 received hands on training 3 to 5 months after NRP course completion. Group 2 was required to view a 26-minute video refresher course 3 to 5 months after initial NRP course completion. Group 3 received no follow-up instruction and served as the control. The standard NRP written examination and a practical examination (which included 5 clinical scenarios) was administered 6 to 8 months after initial NRP course completion. All 3 groups had significantly lower scores on the written examination. Only 59% of the residents passed the practicum. There were no differences among the 3 groups in the number of residents achieving a passing score. Periodic course renewal is not an effective means of keeping resuscitation skills current.
In 1996, Santora et al8 reported on use of videotaped trauma resuscitations over a 13-month period. Videotape reviews uncovered system problems in the study hospital that impacted on trauma resuscitations and revealed a lack of strict compliance with advanced trauma life support (ATLS) protocols. Use of the videotapes as a teaching tool reduced noncompliance with ATLS protocols from 30% to 20%. A 1998 survey showed an estimated prevalence rate of 20% in the use of videotaping as an educational tool in US Trauma Centers.9 The use of video recording in the current format was considered as a result of the use of videotaping in our institution as a QA tool for assessing trauma resuscitations.4
We selected the main obstetric operating room for placement of the video recorder, because this is where the majority of cesarean sections and complicated vaginal deliveries are performed. This site allowed us to evaluate a higher risk group of infants. We are in the process of adding a second camera, which is mobile and can capture resuscitations at other sites. Our initial intentions were to define the degree of NRP deviations and to use this information to improve the conduct of neonatal resuscitation in our hospital. Our use of the videotape format was designed to determine the actual practice of neonatal resuscitation at our institution.
Our current format has certain limitations. In this initial prospective cohort of resuscitations, individual resuscitators were not followed longitudinally. The turnover of housestaff in the NICU is very rapid with PL-1s having only two 4-week rotations during the year, and PL-3s having only one 4-week rotation. This limited our ability to demonstrate statistical improvements in resuscitation. The opportunity to follow their performance over time requires that they be captured on the videotape on more than one occasion. We are now beginning to collect such information and will be able to assess their performance on repeat resuscitations. The composition of the resuscitation teams changed randomly, with the likelihood of the same team being reviewed longitudinally being small. The identity of the resuscitators was kept anonymous during the review and scoring process. Only if significant problems were noted during a resuscitation, were the operators then identified. In these instances, the performance of the resuscitation was discussed with them, and in certain circumstances the videotape were replayed for them. Finally, the bulk of the resuscitations in group 2 occurred in July. Because this coincides with the onset of a new house staff-training cycle, we expected an increase in NRP deviations. Our results do show a slight improvement in resuscitations in group 2, although this was not statistically significant and could be explained by the population in group 2 requiring less intense resuscitation.
The current protocol is capable of detecting and correcting systems problems. The frequent and aggressive use of deep suctioning in our unit was previously undocumented. With the advent of video recording, the prevalence of this practice was noted and corrected. In group 2, the practice had been completely eliminated, with no episodes of deep suctioning being recorded. Another early problem was poor communication of heart rate to the lead resuscitator. After the first few review sessions, this concern was discussed with our caregivers and we have subsequently noted much better verbal and nonverbal communication of the heart rate to the lead resuscitator. This change was so rapid that it actually occurred within the first 25 patients.
The primary purpose of this study was to critically evaluate our current practice of NRP. Our review has demonstrated that a high prevalence of providers, who have successfully completed the NRP course, does not guarantee that the protocol is being implemented as taught. Our observations support the notion that NRP protocols are not always closely followed. The more complicated the resuscitation, the more likely that NRP deviations will occur. Continuous monitoring of resuscitations in a QA format should result in more consistent application of neonatal resuscitation. We believe that the current limitations of our QA process can be improved by making house staff attendance mandatory, conducting repeat assessments of individual resuscitators, and providing constructive feedback on individual performance when appropriate.
The 1999 International Liaison Committee on Resuscitation advisory statement on newborn resuscitation identified multiple aspects of current NRP practice, which need further research. The optimal approach for the meconium stained infant, the appropriate indications and technique for chest compressions, the indications and dosage of resuscitation drugs, and the ideal concentration of oxygen are aspects of resuscitation that require further evaluation.10 There are a number of other practices, such as the use of longer slower breaths to assist in the establishment of functional residual capacity,11 which have not been incorporated into NRP teaching. Should extremely low birth weight infants be approached in a similar manner as the term infant? Are the current recommendations for intubation appropriate and being followed for the very low birth weight infant? Is naloxone indicated and of benefit in the delivery room under any circumstances? Up until very recently, delivery room use of cardiac massage and epinephrine were said to be associated with a uniformly fatal outcome for infants of <750 g birth weight. Our recent reviews suggest that this is not the case.12,13 The current recommendations and teaching of NRP are not all evidence-based. There is a need for further prospective studies evaluating different modalities of resuscitation and their implementation.
Use of video recording of resuscitations not only serves as a valuable tool in QA and education, but also has a potentially invaluable role in assessing new approaches to resuscitation. We believe that the limitations of this system (ie, single view of patient or inability to capture all resuscitations in all delivery rooms) are minor relative to the strengths of the system. The system is very simple to use and we have found that our staff have adapted very quickly to the presence of the recorder. Although the presence of the video recorder may alter conduct in the delivery room, this is usually a change for the better. The recorder is unobtrusive and does not interfere with resuscitation. Permanent mounting of the camera has provided consistently high quality video and audio and has avoided accidental damage to the equipment. The recorded images are unalterable and provide extremely objective data. The simple addition of pulse oximetry during the resuscitation, captured on videotape, will provide even more objective data. This system can be easily duplicated at relatively low cost in any hospital and need not be limited to the delivery room or emergency department environment.
Neonatal resuscitation is probably the most frequently practiced form of acute resuscitation in any environment. NRP was designed to standardize and optimize newborn resuscitation. The use of video recording has demonstrated that, at least in our institution, we are not consistently following NRP guidelines. As we continue to teach and implement changes to NRP, there is a need to confirm that the knowledge and skills taught are consistently and reliably applied during actual clinical practice.
We thank all of the registered nurses, respiratory therapists, and physicians of the Infant Special Care Center at the University of California, San Diego Medical Center for their participation in and support of this project.
- Received December 17, 1999.
- Accepted February 16, 2000.
Reprint requests to (N.N.F.) University of California, San Diego Medical Center, 200 W Arbor Dr, MC 8774, San Diego, CA 92103-8774. E-mail:
- NRP =
- Neonatal Resuscitation Program •
- QA =
- quality assurance •
- NICU =
- neonatal intensive care unit •
- PL =
- pediatric intern •
- ATLS =
- advanced trauma life support
- Bloom RS, Copley C, and the AHA/AAP Neonatal Resuscitation Steering Committee. Textbook of Neonatal Resuscitation. Dallas, TX: American Heart Association; 1994
- Mann CJ,
- Heyworth J
- Kattwinkel J, Niermeyer S, Nadkarni V, et al. ILCOR advisory statement: resuscitation of the newly born infant. Pediatrics. 1999;103(4). URL: http://www.pediatrics.org/cgi/content/full/103/4/e56
- Finer NN,
- Horbar JD,
- Carpenter JH
- Finer NN, Tarin T, Vaucher YE, Barrington K, Bejar R. Intact survival in extremely low birth weight infants after delivery room resuscitation. Pediatrics. 1999;104(4). URL: http://www.pediatrics.org/cgi/content/full/104/4/e40
- Copyright © 2000 American Academy of Pediatrics