There is limited data on the optimum number of procedures emergency medicine residents need to perform to develop competency. The ACGME has set requirements for emergency medicine residents to perform specific procedures [3]. However, these numbers appear to be based on the expert opinion of the Residency Review Committee rather than literature derived.
On average, each resident performed 29 (range 9–52) intubations per year. Thus, our residents quickly met the 35 endotracheal intubation target requirement set by the ACGME. Although the purpose of this cross-sectional study was to provide a snapshot in time of procedure numbers, this result in particular begets the question how many procedures beyond the requirement lead to proficiency? Is 35 intubations per resident an adequate number? Does it mean adequate actual intubation skill or the recognition of difficult airways where advanced procedures are necessary or both? Is there clear learning benefit by more than doubling the requirement as in our program? Sagarin et al. found that as resident physicians increased in year of residency training, their success at intubation attempts improved [5].
Furthermore, video laryngoscopy has become a frequently available tool to complete the procedure. Nevertheless, it is important that residents obtain proficiency in video laryngoscope as well as direct laryngoscopy. We do not track video and direct laryngoscopy separately. How many intubations with and without video laryngoscopy assistance is necessary for proficiency also remains unclear. Sylvia et al. found that emergency medicine residents and pediatric residents had no difference in intubation success using video laryngoscopy in a simulated pediatric airway model, and overall, success was poor [6]. Perhaps a separate skill, video laryngoscopy, needs to be considered as an additional procedural competency by the ACGME for emergency medicine residents in addition to the minimum number of traditional direct laryngoscopy intubations [7].
Bernhard et al. studied intubation success in a prospective observation trial of German anesthesia residents. They compared intubation success and complications for every 25 intubations until each trainee had achieved 200 intubations and found that first past success increased (67 vs 83%, p = 0.0001) and total success and number of attempts required (1.6 vs 1.3, p = 0.0001) significantly improved by the time the trainee had amassed 200 intubations [4]. They concluded that their results should potentially influence the requirements for anesthesia and emergency medicine requirements. While this study was not performed on emergency medicine residents, nor was it performed in the USA, endotracheal intubation is a critical procedure that requires the best probability of first pass success, and our training requirements should reflect that. Moreover, just as importantly, some may question how many intubations are required to recognize procedural competency. This may have even more significance for EM residents and physicians.
On average, each resident performed six lumbar punctures a year. This just exceeds the minimum requirement of 15 procedures. Lammers et al. studied new postgraduate year (PGY) 1 resident proficiency in performing a lumbar puncture on a standard model by judging specific steps that needed to be performed [8]. Although 83% of PGY 1 EM residents reported having had performed the procedure on a patient during medical school, they found that PGY 1 residents had not attained competence. During the study, 69% of the attempts resulted in a failure to obtain cerebrospinal fluid. These findings suggest that more training in lumbar puncture may be beneficial.
Ultrasound numbers also proved interesting, with the average resident in our program performing and documenting over 370 scans in the various modalities, averaging 125 per year. This far exceeds the milestone 4 threshold of 150 ultrasounds, though meeting that number as a sole marker of competence has already been questioned, and a more nuanced pathway towards competence has been published [9]. There have been various studies published advocating or proposing competence with different numbers of studies for different ultrasound studies, varying from as little as 10 ultrasounds for FAST [10, 11] to 25–30 for gallbladder studies [12].
Each resident performed many adult medical and adult trauma resuscitations, averaging 59 and 18 a year respectively. The milestones delve deeper into resident competence, but the raw number of resuscitations does not measure team participation versus team leadership and developing a command presence, which are emphasized in the competency milestone.
There were many procedures that were performed infrequently, including arthrocentesis, pacing, lateral canthotomy, thoracotomy, pericardiocentesis, cricothyrotomy and others.
Thoracentesis is a relatively uncommon procedure performed in an emergency setting, and this, as well as our institutional tendency to defer this procedure to interventional radiology, likely resulted in its finish as the least commonly performed logged procedure. Educational efforts, such as simulation, cadaver lab, and models for these procedures may provide educational benefit.
The high utilization of bedside ultrasound at our program, including its use for peripheral venous access, could have impacted the use of other vascular access adjuncts including intraosseous access, one of the lowest performed procedures in our cohort.
Furthermore, the use of simulation can help to provide experiences to residents in high-risk low-frequency events, such as cardiac arrest or specific cardiovascular emergencies. Okuda reviewed the literature in relationship to resident training and cardiac arrest with the use of simulation and found there is a large variety of options to provide this training to the residents [7]. This type of experience can compliment resident training and can be targeted to lower frequency procedures as identified in our study.
Limitations
There are several limitations to our study. Our method of recording procedures switched after the first few months of the residency program’s start. Initial handwritten records were available and used for data collection but were less reliable compared to later computerized records due to handwriting and quality of the paper after years of storage.
Our results also include both actual and simulated procedures combined. For the first half of the residency program, simulated procedures were not recorded separately from actual procedures performed on patients. After the first 2.5 years, we began recording simulated procedures separately. Therefore, we kept these data combined for consistency as we had no reliable, efficient method to separate them. This can influence our results, especially for rarely performed procedures such as cricothyrotomy, lateral canthotomy, pericardiocentesis, and thoracotomy.
Furthermore, the study relies on individual residents self-reporting procedures completed. After the resident meets the required number of procedures for graduation, the incentive to continue recording every procedure performed decreases, and there may be a large number of procedures performed later in training that were not recorded. This may underestimate the actual number of procedures performed per resident. For example, the average number of intubations is 28.9 per resident per year (86.7 over 3 years), but the highest number of intubations performed is 157 by a single resident over 3 years, almost twice the average.
Perhaps most significantly, this was a single site study—there is likely significant variation in the types of patients, amount of penetrating vs. blunt trauma, deferral of procedures to other services—depending on the site/institution of the residency program.