The safety of term vaginal breech delivery has been the subject of much controversy. In 2000, Hannah and colleagues1 published a multinational randomized controlled trial of planned cesarean delivery compared with vaginal for term breech infants. The trial suggested that planned cesarean delivery reduced perinatal mortality, late neonatal death, and serious neonatal morbidity by one third. In response, the American College of Obstetricians and Gynecologists issued a Committee Opinion in 20012 stating that women with persistent singleton breech presentations at term should undergo cesarean delivery. Although there are some who argue that planned vaginal delivery for selected term breeches remains an option, American obstetricians have all but abandoned planned singleton breech vaginal deliveries. Planned vaginal delivery of the breech is now limited to breech extraction of the second twin in United States' training centers, and even this is not universal practice. Indeed in 2005, in response to declining experience, the Residency Review Committee for Obstetrics and Gynecology, dropped the requirement for tracking experience in vaginal breech delivery.3 Although uncommon, women may still present in advanced stages of labor with the fetus in breech presentation. The lack of breech deliveries available for training obstetric residents combined with the potential for significant fetal injuries resulting from incorrect delivery technique necessitates the need for alternative methods of instruction. During the past few years we have expanded the role of simulation training in the curriculum for our medical students and residents. Furthermore, we have shown that simulation training can improve resident competency in certain tasks and clinical management scenarios.4 The purpose of this current report is to evaluate whether targeted simulation training could improve resident competency in the management of simulated singleton vaginal breech delivery.
MATERIALS AND METHODS
The study population consisted of a convenience sample of obstetric residents from 2 university-based training programs at Georgetown University Hospital and the National Capital Consortium Residency in Obstetrics and Gynecology at the National Naval Medical Center, Bethesda. The investigation was performed in accordance with the guidelines set forth by the institutional review boards of both institutions. All residents participated as part of their scheduled academic teaching and were excused only if they were post-call, on leave, or scheduled for a rotation at an outside hospital.
The simulation used a combination of human actors in role playing and a modified NOELLE female simulator (Gaumard Scientific Company, Inc., Miami, FL), a full-sized female anthropomorphic robotic birth simulator (Fig. 1). All simulations were digitally recorded.
Without prior notice, participants were informed that they would be performing a simulated delivery and were given a standardized case scenario. Before simulation, the resident was instructed to use the appropriate maneuvers to accomplish the delivery as well as do any other important tasks they deemed necessary. The residents were also instructed to treat the simulation as real as much as possible. Just before entering the simulated delivery room, each resident was given the following scenario: a 26-year-old gravida 3 para 2 has presented to the triage room on Labor and Delivery. On examination, the fetus is in breech presentation with the buttocks visible at the introitus without pushing. The fetal heart rate is 150 beats per minute. The estimated fetal weight is 3.15 kg (7 lb). The women states she has had 2 vaginal deliveries of 3.6-kg (8-lb) infants and does not want cesarean delivery.
A standard delivery table was present in the room that included among other common instruments, clamps, scissors, and a towel. Piper forceps were made available if requested but were not initially visible to the resident. A human actor designated as a nurse was available for assistance if the resident requested this. As the resident entered the simulated delivery room, the fetal buttocks was made to deliver and the resident was then made to manage the delivery of a singleton breech. The scenario ended when the resident delivered the fetal head, by using the Mauriceau-Smellie-Veit maneuver (MSV), Piper forceps, or any other maneuvers.
After the initial scenario, all residents participated in a training session with the simulator to review the proper technique for vaginal breech delivery. The training session included a lecture about the risks and techniques of vaginal breech delivery, a review of the correct maneuvers and their sequence, and a practice session on the simulator. Two weeks after the training session, again without prior notification, participants repeated the vaginal breech birth simulation with a similar clinical scenario. All training and testing was performed by the same team of physicians to ensure consistency. A senior maternal–fetal medicine staff physician blinded to the resident's training status (ie, whether performance was before or after the training session) and experienced in performing vaginal breech deliveries viewed the scenarios and graded the participant's performance skills using a standardized evaluation sheet.
The evaluation sheet was tested by a small group of faculty and modified before the investigation. This evaluation sheet included a checklist of all maneuvers for vaginal delivery of a frank breech infant, as described by a major obstetric textbook,5 and a 5-point Likert scale to grade overall performance and safety. Because the delivery of the fetal head with the MSV maneuver is not specifically performed when Piper forceps are used, this component was evaluated separately in residents who did and did not use forceps for delivery. The total number of maneuvers performed correctly was then calculated for each resident.
The percentage of residents who completed each critical task successfully was compared between pretraining and posttraining groups, as were the scores for overall safety and performance before and after training. Statistical analyses included the Wilcoxon signed rank test, McNemar χ2, regression analysis, and the paired t test as appropriate. A P value of less than .05 was considered significant.
A total of 20 residents from 2 institutions completed all 3 parts of the study protocol, including the pretraining simulation, breech vaginal delivery training with the simulator, and then the posttraining scenario. The study group included 5 first-year, 6 second-year, 3 third-year, and 6 fourth-year residents. Although the majority of residents did not use Piper forceps for the delivery, 2 residents used them for both their pretraining and posttraining simulations, and 3 other residents used them after training.
Overall median performance scores were higher (2.0 compared with 3.0, P = .001), as was the rating for safety of the delivery (2.0 compared with 3.5, P = .001) for residents after the simulation training. A comparison of individual key performance components in the management of vaginal breech delivery before and after training is shown in Table 1. After training, residents were more likely to perform 8 of the 12 initial key components up to the delivery of the fetal head. Training was associated with knowledge of the need for assistance (20% compared with 80%, P = .002), refraining from applying premature traction (40% compared with 90%, P = .021), correct technique for the delivery of the first leg (30% compared with 75%, P = .004), appropriate placement of hands to avoid abdominal soft tissue trauma (15% compared with 55%, P = .021), correct direction of traction, (10% compared with 55%, P = .012), proper rotation of the fetus to deliver the arms (35% compared with70%, P= .039), and proper delivery of the initial arm (65% compared with 95%, P = .031). In addition, the mean (± standard deviation) total key component score (of 12) was significantly higher after training (8.0 ± 2.24 compared with 3.95 ± 2.1, P < .001). When we examined junior residents only (defined as PGY-1 and PGY-2), the total key component score was higher after training (3.0 ± 2.1 compared with 7.9 ± 2.5, P < .001), and the same held true when only senior residents (PGY-3 and PGY-4) were considered (5.0 ± 1.6 compared with 8.1 ± 1.9, P = .004). Regression analysis was performed to assess variance in resident performance between the 2 institutions and found no significant difference in the baseline or posttraining total component scores.
When residents who used Piper forceps for delivery were excluded, among the remaining 15 residents, they were more likely to perform the Mariceu-Smellie-Veit maneuver after training (12/15 (80%) compared with 6/15 (40%), P = .031). When only the 15 residents who did not use Pipers were analyzed, and the performance of the MSV maneuver was added to the total key component score (out of 13), the mean scores were again higher after simulation training (8.26 ± 2.3 trained compared with 4.13 ± 2.5 untrained, P < .001).
As vaginal breech deliveries become less common, the level of training in obstetric residency programs will suffer. What will not change and may increase is the risk to the fetus when a breech vaginal delivery is unavoidable and the physician inadequately trained. Simulation training allows for teaching and ensuring skill in today's litigious climate in that there is no risk to patient or fetus. Indeed, Dr. John T. Queenan6, in an editorial in Obstetrics & Gynecology, recommended that teaching vaginal breech delivery similarly to the manner in which cardiopulmonary resuscitation is taught would be beneficial. We agree and the investigation now presented supports the premise that hands-on sessions and mannequins are valuable for training these skills.
The performance of a term breech vaginal delivery is well-suited to simulation for the following reasons: 1) the situation is relatively uncommon and there is little opportunity to practice on real patients; 2) permanent injury may occur if the delivery is not performed correctly; 3) additional emergency situations, such as fetal head entrapment or a nuchal arm, may occur and require immediate intervention; and 4) the simulation may be easily accomplished with inexpensive and available birthing models.
Training residents with an obstetric birthing simulator at our 2 institutions resulted in better overall performance during a simulated term singleton vaginal breech delivery. After training, residents were likely to perform more critical maneuvers correctly. They also subjectively performed better overall and completed the delivery in a safer manner than before training. These findings are consistent with a previous report related to shoulder dystocia, demonstrating that practical application of “book knowledge” can subsequently be enhanced by simulation training.4 Currently, training residents in delivery of a term breech infant is typically done in a lecture or in small impromptu teaching session temporally related to a planned cesarean delivery of a breech fetus.
Although our study included a relatively small number of residents, a strength of the study is that each resident served as his or her own control and still demonstrated significant improvement in performance both objectively and subjectively by a blinded grader. There was also no difference in baseline performance between the institutions, and residents from both demonstrated similar improvement after training. Another benefit of the simulation training is that it not only allows the trainee to learn the proper techniques and maneuvers in a more real-life situation, but it also allows for these skills to be objectively evaluated. In addition, this training can be performed at a relatively low cost.
The mannequin used in our study can be purchased for approximately $2,600. Although there are much more expensive models available, the extra features they contain are not required to simulate several common obstetric emergencies. In addition, the cost of perishable supplies is also minimal, which means that the upkeep expenses are very low. We are aware of other residency programs that are beginning to use these relatively inexpensive mannequins and believe that they will become more commonplace in the future.
Our findings suggest that residents improved their competency in performing vaginal breech delivery after simulation training with a basic obstetric model and instruction. Importantly, tasks which could result in fetal injury if performed improperly were improved. Specifically, training resulted in improved ability to avoid placing traction across long bones, to avoid soft tissue compression, to avoid nuchal arms, and to apply traction at correct angles. The results support our belief that simulation training improves resident competency in the management of vaginal breech delivery, and we have incorporated this simulation into our yearly academic curriculum. We recognize that the predictive validity, ie, whether this training experience translates into equivalent improvement in real-life management of the vaginal breech, would need to be confirmed in large multicenter studies. The prevalence of these deliveries and training suggests that investigation is unlikely to be forthcoming. Our data support the notion that skills in the management of the rare but inevitable clinical scenario of vaginal breech delivery may be evaluated and enhanced by simulation exercises.