The Glasgow Coma Scale (GCS) was devised in 1974 as a way of tracking the progress of neurosurgical coma patients. It is comprised of three components: eye movement, response to verbal commands, and motor function. Since then, it has become the primary tool in Emergency Medical Services (EMS) and emergency departments for assessing cognitive function and triaging patients in the setting of acute trauma. However, the GCS was never intended to be used in such a way. It has been demonstrated that there is a high degree of inter-rater variability when assigning GCS scores for trauma patients. Potential differences in GCS score assignments between different countries were examined. It was hypothesized there would be differences in mean total and component scores.

Using de-identified data from the Pan-Asian Trauma Outcomes Study (PATOS), the distributions of GCS scores from six countries were assessed: Japan, Korea, Malaysia, Taiwan, Thailand, and Vietnam. Using SPSS data analysis, a one-way ANOVA and Bonferroni post-hoc tests were performed to compare the means of the three GCS components and the total GCS scores reported by EMS personnel caring for trauma patients.

Data from 15,173 cases showed significant differences in mean total GCS score between countries (P <.001) as well as in mean component GCS scores (P <.001 for each of eye, verbal, and motor). Post-hoc tests showed that EMS personnel in Korea assigned significantly lower scores compared to all other countries in both component and total GCS scores. Field personnel in Japan, Malaysia, and Vietnam assigned the highest scores and significantly differed from the other three countries on component and total scores; Thailand and Taiwan had similar scores but significantly differed from the other four countries on component and total scores. Visual inspection of mean component and total GCS score histograms revealed differences in score assignment patterns among countries.

There are a number of significant differences in the mean total and component GCS scores assigned by EMS personnel in the six Asian countries studied. More investigation is necessary to determine if there is clinical significance to these differences in GCS score assignments, as well as the reasons for the differences.

While the overall survival rate for out-of-hospital cardiac arrest (OHCA) is low, ranging from 5%-10%, several characteristics have been shown to decrease mortality, such as presence of bystander cardiopulmonary resuscitation (CPR), witnessed vs unwitnessed events, and favorable initial rhythm (VF/VT). More recently, studies have shown that modified CPR algorithms, such as chest-compression only or cardio-cerebral resuscitation, can further increase survival rates in OHCA. Most of these studies have included only OHCA patients with “presumed cardiac etiology,” on the assumption that airway management is of lesser impact than chest compressions in these patients. However, prehospital personnel often lack objective and consistent criteria to assess whether an OHCA is of cardiac or non-cardiac etiology.

The relative proportions of cardiac vs non-cardiac etiology in published data sets of OHCA in the peer-reviewed literature were examined in order to assess the variability of prehospital clinical etiology assessment.

A Medline (US National Library of Medicine, National Institutes of Health; Bethesda, Maryland USA) search was performed using the subject headings “OHCA” and “Emergency Medical Services” (EMS). Studies were included if they reported prevalence of cardiac etiology among OHCA in the entire patient sample, or in all arms of a comparison study. Studies that either did not report etiology of OHCA, or that excluded all cardiac or non-cardiac etiologies prior to reporting clinical data, were excluded.

Twenty-four studies were identified, containing 27 datasets of OHCA which reported the prevalence of presumed cardiac vs non-cardiac etiology. These 27 datasets were drawn from 15 different countries. The prevalence of cardiac etiology among OHCA ranged from 50% to 91%. No obvious patterns were found regarding database size, year of publication, or global region (continent) of origin.

There exists significant variation in published rates of cardiac etiology among OHCAs. While some of this variation likely reflects different actual rates of cardiac etiologies in the sampled populations, varying definitions of cardiac etiology among prehospital personnel or varying implementation of existing definitions may also play a role. Different proportions of cardiac vs non-cardiac etiology of OHCA in a sample could result in entirely different interpretations of data. A more specific consensus definition of cardiac etiology than that which currently exists in the Utstein template may provide better guidance to prehospital personnel and EMS researchers in the future.

Carter RM , Cone DC . When is a Cardiac Arrest Non-Cardiac? Prehosp Disaster Med. 2017;32(5):523–527.

It is unclear which pediatric disaster triage (PDT) strategy yields the best accuracy or best patient outcomes.

We conducted a cross-sectional analysis on a sample of emergency medical services providers from a prospective cohort study comparing the accuracy and triage outcomes for 2 PDT strategies (Smart and JumpSTART) and clinical decision-making (CDM) with no algorithm. Participants were divided into cohorts by triage strategy. We presented 10-victim, multi-modal disaster simulations. A Delphi method determined patients’ expected triage levels. We compared triage accuracy overall and for each triage level (RED/Immediate, YELLOW/Delayed, GREEN/Ambulatory, BLACK/Deceased).

There were 273 participants (71 JumpSTART, 122 Smart, and 81 CDM). There was no significant difference between Smart triage and CDM. When JumpSTART triage was used, there was greater accuracy than with either Smart (P<0.001; OR [odds ratio]: 2.03; interquartile range [IQR]: 1.30, 3.17) or CDM (P=0.02; OR: 1.76; IQR: 1.10, 2.82). JumpSTART outperformed Smart for RED patients (P=0.05; OR: 1.48; IQR: 1.01,2.17), and outperformed both Smart (P<0.001; OR: 3.22; IQR: 1.78,5.88) and CDM (P<0.001; OR: 2.86; IQR: 1.53,5.26) for YELLOW patients. Furthermore, JumpSTART outperformed CDM for BLACK patients (P=0.01; OR: 5.55; IQR: 1.47, 20.0).

Our simulation-based comparison suggested that JumpSTART triage outperforms both Smart and CDM. JumpSTART outperformed Smart for RED patients and CDM for BLACK patients. For YELLOW patients, JumpSTART yielded more accurate triage results than did Smart triage or CDM. (Disaster Med Public Health Preparedness. 2016;10:253–260)

Disasters are high-stakes, low-frequency events. Telemedicine may offer a useful adjunct for paramedics performing disaster triage. The objective of this study was to determine the feasibility of telemedicine in disaster triage, and to determine whether telemedicine has an effect on the accuracy of triage or the time needed to perform triage.

This is a feasibility study in which an intervention team of two paramedics used the mobile device Google Glass (Google Inc; Mountain View, California USA) to communicate with an off-site physician disaster expert. The paramedic team triaged simulated disaster victims at the triennial drill of a commercial airport. The simulated victims had preassigned expected triage levels. The physician had an audio-video interface with the paramedic team and was able to observe the victims remotely. A control team of two paramedics performed disaster triage in the usual fashion. Both teams used the SMART Triage System (TSG Associates LLP; Halifax, England), which assigns patients into Red, Yellow, Green, and Black triage categories. The paramedics were video recorded, and their time required to triage was logged. It was determined whether the intervention team and the control team varied regarding accuracy of triage. Finally, the amount of time the intervention team needed to triage patients when telemedicine was used was compared to when that team did not use telemedicine.

The two teams triaged the same 20 patients. There was no significant difference between the two groups in overall triage accuracy (85.7% for the intervention group vs 75.9% for the control group; P = .39). Two patients were triaged with telemedicine. For the intervention group, there was a significant difference in time to triage patients with telemedicine versus those without telemedicine (35.5 seconds; 95% CI, 72.5-143.5 vs 18.5 seconds; 95% CI, 13.4-23.6; P = .041).

There was no increase in triage accuracy when paramedics evaluating disaster victims used telemedicine, and telemedicine required more time than conventional triage. There are a number of obstacles to available technology that, if overcome, might improve the utility of telemedicine in disaster response.

CiceroMX, WalshB, SoladY, WhitfillT, PaesanoG, KimK, BaumCR, ConeDC. Do You See What I See? Insights from Using Google Glass for Disaster Telemedicine Triage. Prehosp Disaster Med. 2015;30(1):1-5.

Many colleges and universities appear to exist in relative isolation from community-based emergency medical services (EMS) systems. In response, some have developed their own EMS systems.

To determine the extent of this phenomenon and to delineate the characteristics of these systems.

Questionnaires were mailed to 1,503 colleges/universities in the United States and Canada. The questionnaire asked whether the institution had an EMS system and included 19 questions about the characteristics of the system.

A total of 919 (61 %) responses were received. Of the institutions responding, 234 (25%) had an EMS system and 31 (3.4%) were considering starting a system. Characteristics of the systems were as follows: 1) Types of patients—the two most common call types were medical and trauma/surgical; 134 (57%) reported one-fourth of calls to be medical and 91 (39%) reported one-fourth of calls to be trauma/surgical. 2) Type of service—133 (57%) services transport patients; 195 (83%) respond only to the campus or other university property; the remainder also respond to the community; and 135 (58%) function all year. 3) Dispatch—178 (76%) are dispatched by the campus police, although most services are dispatched by several sources; 46 (20%) use 9-1-1. 4) Personnel—two systems (0.85%) exclusively employ paramedic; 141 systems (60%) have at least one emergency medical technician; the remainder use emergency care attendants and first-aid providers; 118 (50%) have medical directors, of these 76 (64 %) are student health physicians and 21 (18%) are community physicians. 5) Demographic Information—The majority of the campus-based EMS systems exist on small campuses in urban areas.

A significant number of colleges/universities have EMS systems and one-half transport patients. However, the level of training of the personnel and medical direction may be below the standard for the EMS systems in the communities in which these campus-based systems exist.

There is a growing interest in cases in which emergency medical services (EMS) providers evaluate a patient, but do not transport the patient to a hospital. A subset of these cases, the patient-initiated refusal (PIR) in which the patient refused care and transport, was studied and evaluated. The objectives of the study were to examine the adequacy of ambulance call report documentation in PIR, to examine the clinical outcome of these patients in one hospital-based, suburban EMS system, and to assess the potential impact of on-line medical command (OLMC) on cases of PIR.

The system studied is a hospital-based, transport-capable, advanced life support service in a suburban EMS system, with an annual call volume of 4,200 runs. During the 6-month study period, all ambulance call reports completed by the paramedics and medical command control forms completed by medical command physicians were examined, and cases of PIR collected. Each ambulance call report was examined for adequacy of documentation. Patient outcome was determined from emergency department records and telephone follow-up.

Eighty-five PIRs were documented during the study period. Four cases were excluded because of a missing ambulance call reports and/or medical command control forms, leaving 81 PIRs for analysis. Despite policy requiring OLMC in cases of PIR, OLMC was established in only 23 PIRs (28%). Of these, two (9%) had inadequate ambulance call report documentation. Of the 58 PIR in which OLMC was not established, 25 (43%) had inadequate ambulance call report documentation (p <0.001, Fisher's exact test). Follow-up was obtained for 54 (67%) PIR. Of these, 37 (68%) did not subsequently see a physician, and all needed no further medical care. Seven (13%) saw their own physicians within a few days of the initial refusal of prehospital care, and had no further problems. Ten patients were seen in an emergency department within a few days. Three (6%) were discharged, and did well. Seven (13%) were admitted to the hospital, with four (7%) admitted to monitored beds, and three (6%) to unmonitored beds. There were no deaths.

Ambulance call report documentation is better with OLMC than without. Patients who initially refuse care may be ill, and some ultimately will be hospitalized. Further research may elucidate a role for OLMC in preventing refusals by incompetent patients, convincing patients who are competent but appear ill to accept transport, and assisting paramedics with other difficult or unusual circumstances.