Page 70 - JSOM Spring 2023

hypothesized that motion metrics would exhibit an inverse FIGURE 1 Medic performing a RUSH exam using a portable
correlation with qualitative ratings, a traditional method of ultrasound probe.
assessment administered by attending anesthesiologists.

Methods
This study received Institutional Review Board approval for
exempt status with a waiver of documentation of informed
consent.

Course Logistics
Through a unique collaboration between civilian and military
team members, we developed a 5-day training course to teach
SOF combat and tactical medics POCUS techniques, includ-
ing transthoracic echocardiography, RUSH exams, US-guided
vascular access, and US-guided regional anesthesia. The mul-
timodal course included online modules, lectures, discussion
sessions, and hands-on practice on simulators and live human A medic acquiring an apical 4-chamber view while being observed and
models. Three cohorts of medics underwent the course led by a rated by an expert (attending anesthesiologist).
team of attending anesthesiologists (experts) with several years RUSH = Rapid Ultrasound for Shock and Hypotension.
11
of experience both using and teaching POCUS. RUSH ex-
ams were the cardinal focus of motion analysis in the course. Motion-based Feedback” below). Segmentation was per-
Therefore, the RUSH exam is the only POCUS technique pre- formed using a previously described method that allows for
sented in this article. the post-hoc analysis of predefined checkpoints in a proce-
dure. Participants acquired the following US views (check-
18
Course Participants points), consecutively: parasternal long axis, parasternal short
In total, 24 medics completed the course following the integra- axis, apical 4-chamber, subcostal 4-chamber, subcostal inferior
tion of motion metrics. The cohorts were divided into 10, 8, vena cava (IVC), Morison’s pouch, left-upper quadrant sple-
and 6 individuals. The sample size was a convenience sample norenal, bladder short axis, bladder long axis, left lung, and
size that was determined based on the availability of learners. right lung.
Cohort 1 was comprised of Navy Sea, Air, and Land (SEAL)
combat and tactical medics and Marine Raiders. Cohorts 2 Expert Ratings
and 3 were comprised of only Navy SEAL combat and tactical Instructors (attending anesthesiologists) who led the course
medics. directly observed and scored the performance of medics per-
forming each RUSH exam by using a 4-point Likert scale (1 =
novice level [needing attention], 2 = pre-trained novice [could
Motion Recordings
make adjustments], 3 = better than novice [significant improve-
Each cohort of medics underwent a 5-day training course in ment], and 4 = almost expert) to rate the exam on each of the
which they were trained in POCUS by a team of attending an- following items: image finding, image fine-tuning, speed, final
11
esthesiologists, as previously described. Each day began with image accuracy, and global assessment. Prior to each iteration
simulator-based training on a CAE Healthcare mannequin of the course, the instructors (experts) thoroughly reviewed
(https://www.caehealthcare.com/patient-simulation/), followed and agreed upon the criteria requisite for each item. Exams
by practice on live human models. Participants each performed were scored one at a time (one expert rater per exam) con-
a total of eight recorded RUSH exams on a live human model. secutively, using the same ultrasound and motion recording
These trials were divided into two exams per day for the lat- equipment across all cohorts. The expert rater for the last two
ter four days of the course. To record the motions of each days of the first cohort’s course and for all days of the second
medic, we attached an electromagnetic sensor to the Butterfly and third cohorts’ courses was the same. Due to clinical sched-
iQ+ (Butterfly Network, Inc., www.butterflynetwork.com) US ules, that expert rater was unavailable for the initial two days
probe used to perform each RUSH exam (Figure 1). Motions of the first cohort’s course. To ensure consistent scoring, all
of the portable US probe were recorded using a Polhemus Lib- instructors agreed upon and were trained in using the rating
erty (Polhemus, https://polhemus.com/) motion tracker. This system before participating in the course. Exam scores were
commercially available electromagnetic tracker costs about used to determine the correlation between traditional means
$16,000 and is portable with a simple procedural setup, allow- of observational assessment and motion metrics.
ing for feasibility of recordings in various settings. The stan-
16
dard model can track up to four sensors, but can be upgraded Objective Structured Clinical Exam (OSCE)
to a total of 16 sensors. A previously published OSCE was used to assess workflow
understanding at the terminus of the course. The OSCE was
19
Based on the motion data, four metrics were then calculated altered to exclude the transesophageal echocardiography sta-
for each exam based on previously described methods: total tion as the skill was not taught in this course and deemed not
distance travelled (path length), number of movements per- clinically relevant for the participants. The OSCE was com-
formed (translational motions), degrees of rotation performed posed of five stations, each consisting of clinical case scenar-
(rotational sum), and time. 17,18 Using video recordings, mo- ios and associated questions to assess participants’ workflow
tion metrics were later segmented to assess performance in understanding and diagnostic abilities for cardiac ultrasound
each US view and provide feedback (see “Observational and (transthoracic echocardiography), lung US, the RUSH exam,

68 | JSOM Volume 23, Edition 1 / Spring 2023