Rubrum Coelis: The Contribution of Real-Time Telementoring in Acute Trauma Scenarios—A Randomized Controlled Trial

Introduction

Most deaths in military trauma care occur soon after the time of wounding, and demand the earliest possible interventions if death is to be prevented. ^{1 –3} Although hemorrhage predominates as the cause of potentially preventable death, airway obstruction, tension pneumothorax, and other severe wounds to the torso and head are also frequent. ^4,5 First responders caring for casualties in operational settings often have limited clinical experience and may be intimately familiar with the casualty for whom they are responsible. The continuing development of information technology (IT) and increasing access to the internet worldwide, including remote locations and combat zones, expand the possible uses of IT in medicine. ⁶ The ability to communicate through video from almost all over the world entices researchers from the fields of trauma, humanitarian aid, surgery, remote medicine, and military medicine to find clinical applications for medical telementoring. ^7,8 The basic tenet in the various forms of telementoring is the same—the operator in the field shows the patient, casualty, or surgical site to the mentor using video and receives vocal or graphical instructions on how to optimize care. ^{6,10 –14}

Gerhardt et al. have previously demonstrated that real-time telementoring of simulated trauma resuscitation was feasible and improved accuracy and efficiency of simulated nonemergency-trained resuscitators. ¹⁰ They aptly concluded, however, that validation and replicated study of these findings for guiding remote damage control resuscitation were warranted. ¹⁰ Our previous work with simulation of trauma medicine did not confirm efficacy in the models used, but revealed significantly improved confidence and mitigated subjective stress in the novice providers. ¹⁵ In a preliminary work carried out in preparation for this article, an emergency medical technician (EMT) in Israel was telementored from Canada. We found telementoring of a military EMT to be technically feasible, but were also concerned that telementoring hindered some of the EMT's abilities and prolonged treatment. ¹⁶

We thus undertook this study to follow through on this need for replication and to further investigate some of the perceived human factors related to nontechnical psychological factors involved in bringing advanced resuscitation further forward in the care of the catastrophically injured.

Methods

Ethical approval was obtained from the University of Calgary (REB15-0850-REN3) and the Israeli Defence Force. Thirty-three (n = 33) Israeli Defense Forces medics were randomized to perform a simulated trauma resuscitation scenario with either telementoring support (TMS) or no telementoring support (NTMS). All medics participating had basic life support training (3-month course). Before the study and allocation to study group, each medic completed a pretest demographic survey and assessment of previous experience. Two simulation suites at the Israel Defense Forces (IDF) Medical Academy (“Bahad 10”) were identically configured with simulation tasks based on the Trauma SimMan (SimMan Essential; Laerdal, Stavanger, Norway). This primary simulator was supplemented with the HapMed^® Leg Tourniquet trainer (CHI Systems, Inc., Plymouth Meeting, PA) as part of the simulation of an amputation requiring hemorrhage control. Whether assigned to TMS or NTMS, each medic was expected to perform a primary survey and initial stabilization of a simulated critically ill victim of a mortar blast (Fig. 1). Critical tasks to be accomplished in the course of the standardized scenario are presented in Table 1, and the on-site scoring form is presented in Table 2.

OPEN IN VIEWER

Fig. 1.

Medic performing a primary survey and initial stabilization of a simulated critically ill victim of a mortar blast.

In those randomized to TMS, the medic wore a standard Israeli combat helmet (F6 PASGT) augmented with a commercial off-the-shelf microphone/headset (Platronics^® PC Headset, Audio 310, Hoofdorp, The Netherlands) and a webcam (Microsoft LifeCam HD-3000 video chat webcam; Microsoft Corporation, Redmond, WA) (Fig. 2). For further recording and documentation, each entire resuscitation was recorded with time stamps using Leica^® Dicomar HC-VX870M WiFi video cameras (Leica Camera, Wetzlar, Germany).

OPEN IN VIEWER

Fig. 2.

Medic equipped with audio/video telementoring setup.

The TMS medics received voice direction from remote medical experts located at the Haifa Naval Base (230 km away). All mentors were actively serving IDF military physicians (see Table 3 for details on the mentors). Seven mentors were chosen to participate in the trial as opposed to having only one or two, to prevent them from becoming familiar with the scenario and performing the mentorship by rote. The remote mentors view included the “patient” and the medic's hands, as the image from the helmet-mounted cam was displayed to them using Microsoft Lync^® 2013 (Microsoft, Inc., Redmond, WA). The remote mentor and medic were in constant real-time voice communication using the same software. At the beginning of every mentored trauma resuscitation session, the remote mentors turned on their video camera to introduce themselves to the medics to be polite and to build rapport, but thereafter turned off their video display to minimize the unnecessary use of bandwidth.

Each medic performed a simulated resuscitation interacting and physically examining the SimMan. The initial vitals of the simulator were computer-set to a heart rate of 130, a respiratory rate (RR) of 30, and a blood pressure of 80/50, and an audio track representing markedly distressed breathing. These were duplicated for both simulators. The medic had no numerical representation of blood pressure however, and the SimMan was set to provide palpable femoral and carotid pulses but absent radial and pedal pulses. If a tension pneumothorax was not relieved promptly, the RR of the simulator increased by 2 breaths/min every 5 min. If the medic did not initiate an intravenous fluid infusion, the heart rate of the simulator increased by 5 beats/min every 5 min. If the medic relieved the tension pneumothorax (TPTX) and initiated fluid resuscitation, the peripheral pulses of the simulator were programmed to return. If the medic did not recognize the presence of the TPTX in either the TMS or the NTMS arm, at 15 min the simulation was halted and the medic was asked to perform a needle thoracostomy to the best of his or her ability. It should be noted that needle thoracostomies are not taught or trained at medic school, and are beyond the scope of medics' abilities and permissions. Any familiarity they may have had with it is by seeing it performed on mannequins by physicians or instructors during exercises. There was no rehearsal between the mentors and TM medics, and mentors were free to interact as much or as little as they wished with the medics.

Assessment of the scenarios consisted of scoring by an on-site judge in each room who scored the resuscitation according to the completion of primary survey and all critical treatment decisions as presented in Table 1. Further assessment was performed by review of the videos. Subjective post-test assessments were also completed separately by the medic and the mentor.

Results

Thirty-three medics participated in the study. Of these, 17 were randomized to the TMS group, and 16 were randomized to the NTMS group. The two groups had no statistically significant differences in terms of their age, skill level, etc. See Table 4 for complete data. Seven remote mentors each performed a mean of 2.4 exercises each.

After the trial, the HapMed trainer placed in the unmentored room was found to be faulty, thereby forcing us to disregard its numerical results (i.e., pressure, blood loss volume, and application time) despite finding some of these to be significantly in favor of the mentored group (mean combat application tourniquet (CAT) pressure: 372.3 mmHg vs. 240.0 mmHg in the TMS and NTMS groups, respectively, p < 0.001; mean calculated volume of blood loss: 435.6 mL vs. 590.9 mL in the TMS and NTMS groups, respectively. p < 0.001. Application times were not found to differ significantly, 1:26 vs. 1:46 min in the TMS vs. NTMS groups, respectively). We, therefore, opted to review all tourniquet applications on the video, and assess them for technique. This was expressed as both proper stretching of the CAT before applying rotation and proper placement on the limb in relation to the wound.

A comparison of the critical treatment decisions and characteristics we examined during the trial is presented in Table 5. Statistically significant differences were found in the time it took to complete the exercise (the mentored group had longer treatment times on average, even when needle thoracentesis is not taken into account). Complete treatment time was 12:56 ± 2:58 min for the mentored group, versus 9:33 ± 3:17 for the nonmentored group (p = 0.003). Needle thoracentesis time itself was slightly longer for the mentored group than for the nonmentored group (1:24 ± 1:00 vs. 0:49 ± 0:21 min, respectively, [p = 0.016], but the mentored group was significantly more successful than the unmentored group in diagnosing 82.35% vs 56.25%, p = 0.003) and treating pneumothorax. (52.94% vs. 37.5%, p = 0.035). Other hallmarks of the treatment were not found to differ significantly.

We examined whether being mentored changed the perceived treatment priorities of the medics. After the exercise, we asked each medic to rate the priorities of treating amputation, pneumothorax, airway obstruction, hypovolemic shock, and burns. Using both t-tests and ANOVA, no significant differences were found in the priorities assigned by the groups. In addition, no significant differences were found between the priorities assigned by the mentored medics and those assigned by their mentors. Table 6 lists the mean priorities assigned by the medics, mentors, and precepts.

The medics were also asked to rate their comprehension of the patients' injuries on a 5-point scale (1 = did not comprehend; 5 = comprehended completely). The results were 2.4 ± 0.63 for the mentored group versus 1.50 ± 0.63 for the nonmentored group, p = 0.0002, suggesting that the mentored group was more confident of understanding the patients' injuries.

Discussion

Telementoring of basic life saving-capable medics in a simulated critical injury scenario significantly improved the accuracy of diagnosis for diagnosing tension pneumothorax and to successfully treat these pneumothoraces. However, this improved success came at the cost of an overall elongated time for each operator to complete the treatment, even under ideal conditions (i.e., large bandwidth and no technical difficulties).

Remote consultations have been in existence ever since the invention of telecommunication. ¹⁶ Indeed, the inventor of radio, Guglielmo Marconi, was an honorary member of Centro Internazionale Radio Medico, an organization based in Rome providing free telemedical advice to ships without a doctor on board of any nationality and sailing in the entire world. ¹⁷ However, current technological improvements afford us an opportunity for real-time video-based (including ultrasound) consultation of first responders in civilian and military settings, and even in outer space. ¹⁸

Telementoring of underskilled operators may or may not in the future become a substitute for having skilled operators in the field. However, with the advance of broadband connectivity and the ubiquity of smartphones in all parts of the world, ^19,20 it is inevitable teleconsultations will take place, even if unintended. Looking toward the future, some military or civilian organizations may opt to put in place a telementoring program for first responders. The current literature on the topic may not yet be sufficiently robust for decisions to be made by policymakers regarding the need, the appropriate technology, proper technique, required training for the operators or consultants, relevant cases, and, especially, guidelines to maximize benefit and minimize harm by this new form of medicine. However, there is increasing evidence that smartphones can play an important role in telemedicine. ^21,22 In the spirit of pure science and curiosity, we sought to set up a robust controlled randomized trial of telementoring for the most basic level of professional operator.

Our findings show that the performance of rote, well-exercised, or basic skills by basic medical providers may not be affected by the remote presence of a mentor. However, in areas or skills in which the operator is not proficient and where there are significant potential complications, that is, diagnosis and chest drainage, the availability of the remote consultant made a measurable impact on treatment. We suggest that during telementoring, the mentor should be sufficiently familiar with the operator's basic skill set, and concentrate on the less familiar areas, so as not to delay performance of these skills and actions. The mentor's interventions should be highly selected for procedures or the operator is not adept in or critical treatment decisions not made by the operator. As to the finding of similar treatment priorities set by both the mentors and the medics, this, may reflect on the mentors and mentees having been educated in the same trauma system in a single school. It may prove to be different in the hands of unskilled operators, but this would have to be studied separately.

The primary limitation of this study was the use of simulation mannequins in sterile conditions. The uniformity of scenarios and the relatively stress-free atmosphere of the simulation laboratory may be a far cry from actual casualty treatment in the field. However, it is the authors' belief that a study of this type highlights some of the medical–technical details of the mentoring process. An additional limitation is the robustness of the internet connection used in the trial. Had this study been performed under real-world conditions, it is possible the treatment times would have been even longer, the mentor would not be able to formulate a reliable picture of the casualty, and that the thoracentesis procedure would not have been successful. Technical considerations such as this may be moot, however, with the continued and ceaseless advance of telecommunication technology. A further limitation is the use of subjective video-based assessments of tourniquet application due to a faulty or uncalibrated HapMed trainer. The choice to use several mentors may have also affected some outcomes of the trial, such as treatment times, as there may be a significant difference in proficiency between experienced and first-time mentors.

In contrast with the previous findings of Gerhardt et al., ¹⁰ we did not find telementoring to shorten treatment, rather, we found it prolongs it. This discrepancy suggests that a shortening of treatment times may not be relevant in all cases, and illustrates how the conduct of remote TMS requires more rigorous study. At the current time, remote TMS is in its infancy regarding the science of how and when to utilize these techniques. The nomenclature, communication protocols, and priorities all remain to be defined at their most basic levels. However, regardless of this initial prematurity in structure, a technique that can empower inexperienced point-of-care providers in catastrophic settings to more accurately diagnose and treat life-threatening injuries should be further explored and refined.

The findings of this study emphasize the fact that despite numerous attempts over the years, we have not yet succeeded in maturing this promising technology into a widely accepted applicable tool. Further studies are in place to assist in establishing protocols and guidelines that will not only allow the mentees to provide better more advanced care, but also allow the mentors to concentrate on those procedures and treatments that such mentoring will have a positive contribution.