Introduction:

Severe trauma remains a major issue for public healthcare worldwide. Management of these polytrauma patients is mostly performed in intensive care units (ICU) that provide mutlidisciplinary care. The radiological evaluation is a vital part of their management and is essentialy based on performing a whole-body computed tomography (WBCT). However, less attention has been paid to the management of non-vital polytrauma patients in the Emergency Departments (ED). The aim of the study was firstly to evaluate the management of non-vital polytrauma patients who had a WBCT in our ED. We also performed this study to find predicting factors of severity at hospital admission.

Methods:

We present here a retrospective and monocentric study. We reviewed the chart of all patients who had a WBCT (for trauma) in 2014. We collected epidemiological, clinical and biological parameters and all therapeutic measures during the ED stay. A long-term survival follow-up was also performed. All patients directly admitted to the ICU were excluded.

Results:

A total of 210 patients were included for statistical analysis and 64% (CI95% : [57.8-70.8]) of them had one or more lesion(s) in the WBCT (36% normal WBCT). The mean ISS score was 10.1±8.8. 42 patients (20% ; CI95%: [14.6-25.4]) underwent urgent surgical procedures or were admitted to the ICU. We have defined these patients (n=42) as severely injured trauma patients. The mean ISS score for these patients was 16,1±10.8 compared to others 8,5±7.5 in the remaining cohort (p<0.0001). The mortality rate predicted by the TRISS model was 3,1% compared to 1,5% in the whole cohort (p<0.0001). The average length of stay in the ED was 5,4±2.9  hours for the severely injured trauma patients versus 7,2±4.6  hours for the other patients of the study (p=0.003) and the average lenght of stay was 16,2±18.9 days for the severely injured group versus 3,1±6.4 days for the non-severely injured patients (p<0.0001). In multivariate analysis, heart rate (>100/min) and Vittel score (³2 criterias) were related to the probability of belonging to the severely injured trauma group (p=0.03). The 24-hour mortality rate was 0.5% in the ED and the 30-day mortality rate was 1.5%.

Conclusion:

The development of a specific network in the ED hosting non-vital but severly injured polytraumas remains crucial. The primary goal of this future network will be to meet technical and time requirements and establish new in-hospital triage algorithms based on clinical variables (such as heart rate), in order to detect these patients at an early stage and offer them priority care in our overcrowded EDs.

Introduction. Traumatic brain injury (TBI) is very common in emergency department. Most of them are the results of mild head injury defined by a Glasgow coma scale score greater or equal to 13. In case of head trauma under antiplatelet agent, guidelines of the French Society of Emergency Medicine required to perform a CT scan to detect brain lesions. In this indication, 90% of CT scan are normal. The utility of CT is still debate given controversial and scarce number of studies.

Methods. We used the RATED registry (Registry of patient with antithrombotic agents admitted to an Emergency Department, NCT02706080) to assess factors of cerebral bleeding related to antiplatelet agent after head trauma. RATED is a monocentric, ongoing (from January 2014), observational registry of consecutive patients under antithrombotic drugs admitted to an emergency department. For this study, only patients under antiplatelet drugs at admission, with a head trauma who benefit a CT scan, were considered. Those under anticoagulants drugs were excluded.

Results. From January 2014 to December 2015, 993 patients under antiplatelet agent were recruited. Of these, 293 (29.5%) underwent a CT scan for trauma brain injury. Intracranial bleeding was found in 26 patients (8.9%). After multivariate analysis, these patients were more likely to have a history of severe hemorrhage (OR = 8.47, 95%CI: 1.56-45.82), a dual antiplatelet therapy (OR = 6.46, 95%CI:1.46-28.44), headache or vomiting (OR = 4.27, 95%CI: 1.44-2.60) and less frequently a glasgow coma scale of 15 (OR = 0.11, 95%CI: 0.03-0.35) than those without intracranial bleeding. The predictive model derived from these variables had a specificity of 98.9% and a Negative predictive value of 92%. The area under curves of the ROC curve was 0.85 (95%CI: 0.77-0.93).

Conclusions. Our study showed that the absence of a history of severe hemorrhage, dual antiplatelet therapy, headache or vomiting, and normal Glasgow coma scale score seems to predict a normal CT scan after trauma brain injury under antiplatelet. This founding need to be confirmed by prospective studies.

Background

Falls constitute a significant challenge in health care, particularly in the oldest age groups. We aimed to investigate the incidence of fall-related injuries among the older Finns.

Materials and methods

In this retrospective population-based study, we registered all consecutive emergency admissions due to fall-related injuries in a high-volume emergency department (ED) during a 2-year study period (January 2015 to December 2016). The catchment area of the study hospital exceeds 1 million inhabitants in tertiary care services, and it provides both primary and tertiary care ED services for a population of 226,696 inhabitants (10,991 aged 80 years or more). Patients were eligible for this study if they were 80 years or older. Patient demographics and diagnoses were retrieved from hospital records. The key factor under analysis was the incidence of fall related injuries.

Results

During the study period, 2,951 patients (median age 87, range 80-104; 73 % females) had 3,802 emergency admissions due to fall-related injuries (2.2% of all ED visits and 11% of all visits in this age group). The incidence of these injuries increased from 49/1,000/year in inhabitants aged 80-89 years to 87/1,000/year among those aged ≥90 years (p<0.001). The risk was also higher among females (p<0.001), the incidence was 60/1,000/year among females and 45/1,000/year among males. Typical injuries caused by falls included, fractured femur (21%), intracranial injury (19%, with intra-cranial hemorrhage in 5% of patients), open head wound (15%) and forearm fracture (7%). The number of injurious falls of all ED visits varied from 253-358 from month to month, with no association with time of year. Recurrent falls were registered in 39% of patients (range 1-5 falls). When single fallers and recurrent fallers were compared, older age and female gender were not associated with higher risk of recurrences (p>0.05). In community-dwellers, most injuries (64%) occurred in domestic setting. Ten percent of all patients were admitted to hospital from nursing homes.

Conclusions

The observed incidence of fall-related injuries is lower than in earlier reports. Nevertheless, these incidents cause a significant burden to emergency services. While the risk of single falls increased with age and was also higher among females, these risk factors were not associated with higher risk of fall recurrences.

 Introduction:

One of the most important factors for total morbidity and mortality in traumatized emergency patients is chest trauma; the complexity of injury in trauma patients makes it challenging to provide an optimal oxygenation while protecting the lung from further ventilator-induced injury to it. Ultrasound has a well-known established role in the diagnosis of a traumatic pneumothorax.

Case series: We are reported 46 patients with traumatic pneumothorax who require positive-pressure ventilation   recruited from Alexandria Main University Hospital Emergency Department who examined pre and post CT-Chest scanning by the same operator between November 2015 and September 2016.

Inclusion criteria: Any patient with either blunt or penetrating chest trauma regardless the age or gender who require positive-pressure ventilation.

Exclusion criteria: patients with open or tension pneumothorax.

The chest scanned (using the superficial probe 7.5 MHz type L7M-A of our CHISON device model ECO 2) at three lines and two views for each hemithorax as the following (1)anterior second through sixth intercostal spaces at the parasternal line, (2) anterior second through sixth intercostal spaces at the mid-clavicular line, (3) fourth through sixth intercostal spaces at the anterior axillary line,(4) fourth intercostal space at the mid-axillary line, (5) fourth intercostal space at the posterior axillary line, to assess for the presence of a sliding lung.

We analyzed 46 patients, 85% were men, with median age of 27 for the total number of patients. The results of ultrasound scanning of intubated patients’   pre CT scan in comparison to the result of ultrasound scanning after the CT scan of the chest to detect the behavior of traumatic pneumothorax. Six patient (13%) had expanding pneumothorax detected by CT chest and comforted by bedside US. One patient developed right tension pneumothorax after 6 hours, in which pre CT US scanning was Right minimal pneumothorax and Left minimal pneumothorax, post CT after 35 minute was Right mild pneumothorax and left minimal pneumothorax. Another patient developed massive left pneumothorax discovered accidental during CT abdomen, pre CT US scanning was Right mild pneumothorax and free left pneumothorax, post CT was Right mild pneumothorax and left minimal pneumothorax.                                                                                                                                                     

Conclusion: Bedside US lung could be used a predictive tool in detection of early expanding traumatic pneumothorax in-patient on positive-pressure ventilation.

M.D., PhD-student Egholm JWM^1,2, PhD Pedersen B^1,3, M.D. Oppedal K⁴, Professor M.D. Lauritzen JB⁵, M.D. Madsen BL⁶, Professor M.D. Tønnesen H^1,3

¹WHO-CC, Clin Health Promotion Centre, Bispebjerg-Frederiksberg Hospital, University of Copenhagen, Denmark; ²Orthopedic Dept, Hospital of Southern Jutland, University of Southern Denmark; ³Clin Health Promotion Centre, Dept Health Sciences, Lund University, Sweden; ⁴Alcohol and Drug Research Western Norway, Stavanger University Hospital, Norway; ⁵Dept Orthopaedic Surgery, Bispebjerg-Frederiksberg Hospital, University of Copenhagen, Denmark; ⁶Dept Orthopaedic Surgery, Hvidovre Hospital, University of Copenhagen, Denmark.

Introduction:

Patients with hazardous alcohol intake are overrepresented in emergency departments and surgical departments. In elective surgery, preoperative alcohol cessation interventions can reduce postoperative complications[1] but no studies have investigated the effect of alcohol cessation intervention at the time of surgery for acute fractures.

Purpose:

To evaluate the effect of the Gold Standard Programme for alcohol cessation intervention (GSP-A) for patients undergoing acute ankle fracture surgery regarding postoperative complications.

Methods

Our RCT design included a total number of 70 patients from Hvidovre and Bispebjerg Hospitals with an excessive intake of alcohol and an ankle fracture that required osteosynthesis. They were allocated to either standard care or a 6-weeks GSP-A aiming to complete alcohol abstinence peri- and postoperatively.

GPS-A involved a patient educating programme and weekly sessions at the orthopedic outpatient clinics. Furthermore, patients were provided with thiamine and B-vitamins, prophylaxis and treatment for alcohol withdrawal symptom and disulfiram to support abstinence.

Biochemical validation of alcohol intake was carried out. Follow-up took place after 6 weeks and 3, 6, 9 and 12 months.

The main outcomes were postoperative complications (requiring treatments), alcohol intake and cost-effectiveness.

Results:

In the GSP-A group 12 patients (34%) developed complications compared to 14 patients (42%) in the control group (p=1.0).

Interestingly, 14% in the control group versus 51 % in the in the GSP-A group had abstained completely from alcohol at 6 weeks follow up. (p=0.001).

Conclusion:

We found no statistically significant differences in the number of complications between the intervention and control group.

We are looking forward to evaluate the cost-effectiveness of the study to see if there is a difference between the two groups. Biochemical analysis is still ongoing.

However, the majority of the intervention group completely abstained from alcohol in the intervention period.

Contact information: This email address is being protected from spambots. You need JavaScript enabled to view it.

Background

Head, neck, extremity, thorax and abdominal injuries are the main injury sites for trauma in Hong Kong. Return to work (RTW) status is an indicator of their outcome and an important element in their social reintegration. Besides physical recovery, RTW may be affected by the psychological characteristics of patients, available technologies that can assist their functional recovery and socioeconomic factors. Some of these are modifiable and may vary between injury sites.  The aim of this study was to compare the time to return to work (RTW) and time to reach Hong Kong norm of SF36 of patients who sustained moderate to severe injuries at these sites. This may help to identify the type of injury where research is most needed to bridge the gap between time to recovery and return to work.

Methods

This was a multicenter, prospective cohort study of patients entered into the trauma registry of the three regional trauma centres in Hong Kong. Patients were included if they were aged between 18 and 70, with an ISS³9 and were working or seeking employment before injury. Outcome measures included physical component summary (PCS) and mental component summary (MCS) scores in SF36 and their return to work status over 5 years. The Hong Kong norm for PCS is defined as >52.83 and that for MCS is >47.18.

Results

189 patients were recruited to the study from 1 January to 31 September 2010 (mean age 41.4 years; 80.4% male). 99 (52.4%) patients had head or neck injuries, 44 (23.3%) had extremity injuries, 39 (20.6%) had thoracic or abdominal injuries, and 7 (3.7%) had spinal or other body region injuries. 5-year mortality rate was highest in head injury patients (12/99, 12.1%, 95%CI 6.4-20.2%), followed by extremity injury (1/44, 2.3%, 95%CI 0.1-12.0%). There was no death in patients with injury in other parts of the body. For patients that are still alive at 5 year follow-up (n=87, 43 and 39 for head/neck, extremity and thorax/abdomen injuries respectively), Kaplan-Meier curves of time to RTW were different between the three principal injury sites (log-rank test p=0.041). Mean time to RTW for head/neck injury was 2.2 years (95%CI 1.6-2.8 years), 2.3 years (95%CI 1.5-3.1 years) for thoracic/abdominal injury and 3.6 years (95%CI 2.8-4.5 years) for extremity injury. However, Kaplan-Meier curves of time to reaching the Hong Kong normal PCS or MCS showed no significant difference between the three injury sites (log-rank test p=0.386 and 0.482 respectively; mean time to PCS>52.83 ranged from 3.3-4.0 years and mean time to MCS>47.18 ranged from 2.0-2.3 years).

Discussion

Among the factors that affect the RTW status besides health status, some of them are specific to extremity (limb) injuries. This may represent a specific group for targeted rehabilitation to shorten the time to RTW.

Funding

The study has been conducted with the support of the Health and Health Service Research Fund from Hong Kong government (HHSRF 07080261).

Background: whiplash Associated Disorders (WAD) are very frequent reasons for request of health care interventions in acute setting. They involve about 30-40% of people after a car accident, and they are a growing problem in terms of both health and insurance costs. The “gold standard” of WAD in acute management is not established and the application of cervical collar continues to represent a common practice, although it is known that it does not to improve the course of disease.

Aim: the goal of this study is to assess whether the use of cervical collar in patients with WAD is associated to a higher risk of readmission within 90 days from trauma and the likely onset of post-traumatic associated syndrome

Methods: we retrospectively evaluated all the patients observed in the Emergency Department (ED) of the University Hospital of Verona for WAD, according to the Quebec Task Force definition. We considered only the patients with isolated WAD within 48 hours from a car accident, excluding those with associated head trauma. At the ED evaluation time we registered for each patient: the demographic and clinical features; the time and mechanism of trauma; the grade of WAD and if the application of cervical collar was performed or not. Moreover, we excluded the patients in case of rx assessment of cervical spine fracture (WAD 4). We carried out a multivariate analysis to verify the impact of the cervical collar on the outcome adjusted for grade of WAD. Finally, we submitted our series of patients to propensity score matching, in order to reduce the bias of a retrospective study, and then we repeated the statistical analysis on the selected cohort of patients.       

Results: from January 2013 to December 2014 we observed 2156 patients with WAD (grade 0 to 3). In most of the cases (85.5% of the patients) a cervical collar was applied. An overall number of 162 patients (7.5%) had a readmission within 90 days from trauma and 154 (7.1%) out of them used a cervical collar (p = 0.001). Cervical immobilization resulted to be a significant risk factor for readmission both in univariate (OR 3.663, 95% IC 1.684-7.122) and in multivariate analysis (adjusted OR 3.561, 95% IC 1.255-7.349). After the propensity score matching we selected 482 patients equally divided (50% with and 50% without cervical collar). Even in the selected series readmission rate was higher in patients with cervical collar (10.4% vs 2.9%, p < 0.001). The risk was confirmed both in univariate (OR 3.869, 95% IC 1.640-9.127, p = 0.002) and in multivariate analysis (adjusted OR 3.878, 95% IC 1.643-9.153, p = 0.001).

Conclusion: the application of cervical collar in WAD appears to be an independent risk factor for the onset of post-traumatic associated syndrome. These results are not related with the grade of WAD and they have been confirmed even when propensity score matching was performed.

Introduction:  Anaphylactic shock (AS) is a life-threatening condition in which blood circulation fails to meet oxygen demands of organs in the body. Restoring effective circulation is the major objective of treatment which can sometimes be resistant to conventional treatments available. We evaluated novel drugs targeting specific pharmacological pathways to define therapeutic alternatives or supplements to epinephrine. The role of potassium voltage-dependent channels (Kv) and hydrogen disulfide (H₂S), a potent vasodilator, was not explored in AS.

Objectives: Our objective was to demonstrate that the inhibition of Kv with 4 aminopyridine (4-AP) alone or combined with the inhibition of the production of H₂S with dl-propargylglycine (PEG) or betacyanoalanine (BCA) restores blood pressure and improves survival.

Methods: Rats were sensitized with ovalbumin (OVA, 1mg s.c), and AS was induced by i.v. injection of OVA (1mg) through jugular vein catheter. Experimental groups included NA=non-allergic rats (N=6); Controls=allergic rats (N=10); allergic rats treated with 4-AP (1 mg/kg) (N=10); epinephrine (EPI)=allergic rats treated with EPI (10 µg/kg) (N=10), allergic rats treated with PEG (N=10), allergic rats treated with BCA (N=10). Treatments were administered 1 minute after induction of AS. Mean arterial blood pressure (MAP), heart rate (HR) were measured through carotid artery catheter

Results: MAP, HR, and survival were measured for 60 minutes. MAP was normal in the NA group; severe hypotension and high mortality were observed in controls; normalization of MAP, HR, and increased survival were observed in 4-AP, DPG, BCA, 4-AP+DPG or BCA, and EPI groups. Survival time was: controls=22±3 min, EPI=59±1 min, 4-AP=60 min, DPG=49±5 min, BCA=44±7 min, 4-AP+DPG=58±2, 4-AP+BCA=60 min. All allergic 4-AP and 4-AP+BCA-treated rats survived after the induction of AS (p<0.05).

Conclusion: We demonstrated that the allergen-activation of Kv is a probably a new pathway involved in the vasodilation induce by AS in a rat model. It seems that H₂S is a major mediator released during AS in endothelial and vascular smooth muscle cells and could play a role in the induction of AS. Inhibition of Kv alone or combined with the inhibition of H₂S production improves significantly survival and restore blood pressure. 

Background: Inflammation and thrombosis are involved in the progression of acute aortic dissection (AAD). The aim of this study was to assess the prognostic significance of the Simplified Thrombo-Inflammatory Prognostic Score (sTIPS) in patients with early phase type B AAD.

Methods: We retrospectively reviewed 491 patients with type B AAD between November 2012 and September 2015. sTIPS was calculated using data obtained at the time of admission, and patients were assigned a score of 0, 1, or 2. Kaplan-Meier curves and multivariable Cox regression analyses were used to investigate the associations between the score and hospital survival.

Results: Of 491 type B AAD patients included in this analysis, 24 patients (4.9%) died while hospitalized. Compared to those with lower sTIPS, patients with higher sTIPS had higher rates of in-hospital mortality (P = 0.001). Kaplan-Meier analysis also showed that cumulative mortality was significantly higher in patients with higher sTIPS (P = 0.001). Multivariable Cox regression analysis further revealed that scores of 2 or 1 (versus 0) were strong predictors of in-hospital mortality (sTIPS 2: hazard ratio: 5.620, 95%; confidence interval [CI]: 1.320-16.167; P = 0.017; sTIPS 1: hazard ratio: 2.012, 95%; CI: 1.254-3.204; P = 0.043) after controlling for all of the confounding factors. Subgroup analysis showed sTIPS was also positively associated with the hazard of in-hospital mortality in patients with different therapies.

Conclusions: sTIPS was a useful tool for risk-stratifying type B AAD patients at admission for outcomes such as in-hospital mortality in the early phase.

Introduction. The morbidity and mortality of patients with pulmonary embolism is high with 8 to 23 deaths per 100,000. In order to make the diagnosis, the emergency physician uses clinical scores that involve the presence or absence of a tachycardia. The objective of our study was to assess the impact of beta-blockers on the heart rate of patient with a pulmonary embolism.

Methods. We performed a retrospective, observational, monocentric study from June 2014 to May 2015. All consecutive patients admitted to our department with an objectived pulmonary embolism were included.

Results. Among the 117 patients included, more than a quarter was under beta-blocker (n = 31). The mean age was 68+/-17 years. Patients with beta-blockers were older than those without (respectively, 76+/-11 vs. 66+/-18 years, p = 0.03). The mean heart rate was lower in patient with beta-blockers than in those without (76.714 vs. 90.719 bpm, p <0.001, respectively). Moreover, regarding the heart rate item in the revised Geneva score, those with beta-blockers were more likely to have a heart rate < 75 bpm (42% vs. 19%, p = 0.015, respectively) and less likely > 95 bpm (13% vs. 41%, p = 0.04, respectively) than those without. However, for the sPESI score, there was no difference between the mean score of those with and without beta-blockers (respectively, 1.2+/-0.9 vs. 0.9+/-0.8, p = 0.104). Interestingly, regarding the heart rate item of sPESI score, none of the patient with beta-blocker had a heart rate > 110 bpm whereas 9 (10.5%) of those without beta-blocker had it (p = 0.11).

Conclusions. Our study showed that 26.5% of PE patient admitted in an emergency department were under beta-blockers. Moreover, PE patient with beta-blockers had a lower mean heart rate than those without. This could have an impact on the different scores used by emergency physicians to diagnose PE. Thus, 42% of PE patients with beta-blockers did not obtain the point assigned to tachycardia in the revised Geneva score. This could therefore have an impact on emergency physician diagnosis which could underestimated the PE score of some patient with beta-blocker. This results should be validated by multicentric and prospective studies.

Introduction. The use of direct oral anticoagulants (DOACs) is increasing due to an easier used and a decreased bleeding risk. The objective of our study was to describe the clinical characteristics of patients under DOACs, the type of hemorrhage and their management.

Methods. We performed a retrospective, monocentric and descriptive study on patients under DOACs and admitted to our emergency department between January 2014 and June 2015. We used the RATED registry (Registry of patient with antithrombotic agents admitted to an Emergency Department, NCT02706080) which is a monocentric, ongoing (from January 2014), observational registry of consecutive patients under antithrombotic drugs admitted to an emergency department.

Results. Of the 357 patients under DOACs included, 51 (14.3%) were under apixaban, 109 (30.5%) under dabigatran and 197 (55.2%) under rivaroxaban. Atrial fibrillation was the most frequent indication for DOACs (79.3%) with an average CHA2DS2-VASc score of 3.9 ± 1.8. The mean age was 73 ± 15 years with 78 ± 12 years for apixaban, 74 ± 13 years for dabigatran and 71 ± 16 years for rivaroxaban. Our cohort identified 211 (59%) prescribing errors related to the combination of a treatment which could increase the bleeding risk (33.6%), a dose not adapted to the age of the patient (28.0%), a dosage not adapted to creatinine clearance level (15.9%) or an additional treatment with antiplatelet agent for more than 1 year (13.3%). Of the 357 patients under DOACs, 64 patients (17.9%) were admitted for hemorrhage with 23 major bleeding (6.4%). Bleeding occurred in 8 patients (15.7%) under apixaban, in 13 under dabigatran (11.9%), in 43 (21.8%) under rivaroxaban with, respectively, 2-5-16 major bleeding. The management of those with major bleeding (n=23) was transfusion for 9 (39%) of them (5 under rivaroxaban, 3 under dabigatran and 1 under apixaban), reversion by prothrombin complex concentrate for 1 (4.3%) under rivaroxaban or by FEIBA for 4 (17.4%) under rivaroxaban.

Conclusions. Our study showed that more than half of patients under DOACs admitted to the emergency department had a prescription error and that only 1 major bleeding over 5 were treated with a reversal agent. Even if DOACs are easier to use, clinicians must be aware of interaction and contraindication, and must improve the management of major bleeding.

Introduction: In the treatment of ST-segment–elevation myocardial infarction (STEMI), faster times to reperfusion result in better outcomes. Primary prehospital Helicopter Emergency Medical Service (HEMS) interventions may play a role in reperfusion therapy. We analyzed data from our Cardiologic Regional Interventional Observatory Registry involving patients with STEMI aimed at the evaluation of the potential benefit of HEMS interventions as compared with EMS ground transport with MICU (Mobile Intensive care Unit).

Materials and Methods:Retrospective multicentric study conducted from January 2014 to January 2015. All successive patients with STEMI eligible for percutaneous coronary intervention (PCI) were included. Access times were computed allowing the estimation of dispatch French System (SAMU) delay from emergency the phone call to admission to one of the 6 cath labs using standard data collection from the French region Centre Registry of Acute Coronary (CRAC). We analyze pre admission transport time (FMC-DTB) according to  distance between FMC to cath lab location and mode of transport (HEMS vs MICU)

Results: During the study period, 1699 patients with STEMI were included in the Registry.In the overall population, Mean age was 63,2 y-o, sex ratio male was 2,7. The mean time from the emergency phone call to the dispatch center (Number 15 or 112) to the ECG as the First Medical Contact (FMC) is 1h40 mn. Of these patients 82 % were transferred for primary PCI, 2% fibrinolytic therapy, 7% secondary angioplasty.The mean response time FMC to Door-to Balloon (DTB) time was longer using the helicopter HEMS (2h20 mn) than road transport with MICU ambulances (2h:06 min).For short distances (25-74 km) the median delay using road transport was shorter (1h: 54 min) whereas this time by air transport (2h: 12 min).The median transport delay gain was shorter (15 min) for distances over 75 km by using HEMS.

Discussion: In many European countries, patients with STEMI , guidelines have called for device activation within 90 minutes of initial presentation. Our results offer important insights of the impact of transfer time and mode of transport on achievement of guideline goals for D2D time in the treatment of STEMI patients as HEMS transport did not offer D2D time advantages for STEMI patients better than ambulances for short distances. Several explanations: (1) time for air transportation, (2) cases with more complexity (3) distance .In our study for patients with STEMI,  benefits of air transportation with HEMS are not significant for short (< 50 km) or long (> 75 km) distances and may encourage the dispatch center to use the MICU ambulances for a rapid access to PCI .

Conclusions In our CRAC registry for management of STEMI, strategy of decision was associated with longer DTB times with HEMS versus MICU ambulances.Our findings suggest the need for evaluation and continued systems improvement of the use of effective resources for transport of STEMI to the PCI.

 Background

The purpose of this study was to assess the prognosis for patients with suspected acute coronary syndrome (ACS) after the implementation of high sensitivity troponins (hstn). The first objective was to investigate if there had been a change in the incidence of  patients  diagnosed with ischemic heart disease and non-specific chest pain(NSCP) after the implementation of hsTn. The second objective was to determine the mortality in the NSCP group before and after hsTn and the third objective  to evaluate the prognosis and mortality in patients discharged with the diagnosis ACS, stable angina pectoris (SAP), other heart related conditions (OHC) and NSCP after hstn implementation.

 Method

This study is a register based prospective multicentre cohort study. All patients aged 18 and older in the region of southern Denmark with an acute contact to the emergency department or cardiology department in 2013 and who had at least one troponin measurement were included. Depending on their discharge diagnosis and the Troponin value, the patients were stratified into 4 groups, consisting of

a)      Patients with myocardial infarction and elevated Troponin measurement.

b)      Patients with other serious heart disease than ACS

c)      Patients with clinical assessed stable SAP and normal troponin

d)     Patients with normal troponin and without acute heart disease, who were discharged without an explanatory diagnosis of the chest pain, NSCP.

 Patients arewere identified by their troponin measurements from the biochemical system. Endpoint during follow up was obtained from the National Patient Registry and Civil Registration System. Endpoints were readmission with myocardial infarction, ventricular fibrillation/cardiac arrest and all-cause death. Endpoints were reported for one year follow up or until emigration.

 Results

6037 patients were eligible for final analyses. The group consisted of 2 513 patients before and 3 491 after the hstn implementation. Comparing the diagnosis before and after hstn, there was a relative increase in MI of 25% and OHC with 9%, NSCP and SAP decreased with 6% and 43% respectively. The 12 months-mortality for NSCP was 3.4% before hstn and 2% after. No significant difference was shown between in mortality before and after hstn. The prognosis for the four groups after implantation of hstn showed that MI had the highest risk of future MI with 4% event rate during one year. NSCP had 5 times lower risk of future MI compared to the MI group. No significant differences  in mortality was shown when comparing the MI group with the NSCP, OHC or SAP groups.

 Conclusion

The number of NSCP patients decreased after the hstn implementation and had a low risk for future cardiac events No significant differences in mortality was shown comparing the mortality rate in this group before and after hstn implementation. The risk of future MI is significantly lower in NSCP than in the MI group

Background

Atrial fibrillation (AF) increases the risk for stroke and other thromboembolic complications. Properly carried out anticoagulation (AC) is the most important treatment for AF patients in stroke prevention, and it has been found to prevent up to 60-70 % of strokes. In clinical practice, the need for anticoagulation in AF is estimated by using CHA₂DS₂VASc and HAS-BLED scores. CHA₂DS₂VASc score assesses the thromboembolic risk and HAS-BLED score the risk for bleeding.

According to recent national and international studies, anticoagulation is underused among AF patients. In particular, the Finnish FinFib study showed that as many as 29% of high risk patients did not receive anticoagulation (1). Therefore, we wanted to examine how well the AF treatment is carried out at Tampere University Emergency Department (ED).

Materials and methods

From 1 October 2014 to 30 November 2014 all patients with AF or atrial flutter (ICD-10 code I48) at ED were identified from hospital records. We collected the following data: age, gender, what type of AC was used and whether the diagnosis was new or previously existing. We also investigated if the CHA₂DS₂VASc and HAS-BLED scores were calculated during the patients stay. Additionally, we counted both scores for all the patients, regardless of whether they were calculated during the ED visit. For this purpose, we collected data on all patients with any type of risk factors. Finally, we investigated whether the bridge therapy was started with warfarin, and furthermore, what were the INR values of all the patients on warfarin therapy.

Results

A total of 470 patients with 537 unique ED visits were included in the study. Patients had a mean age of 70 years (range 21 -100 yrs). Out of the unique ED visits, CHA2DS2VASc and HAS-BLED scores were calculated in 19% and 3% of the cases. 87% of all the high-risk and 72% of new high-risk AF patients had AC therapy when leaving the ED. 57% of the patients on warfarin had an INR in therapeutic range (2.0-3.0). Furthermore, only 53% of patients with newly diagnosed AF referred to cardioversion had received bridging AC therapy with small molecular heparin (LMWH).

Discussion:

We found that although CHA₂DS₂VASc and HAS-BLED scores were calculated in a minority of the cases, relevant AC therapy for high-risk patients was well executed. However, over every fourth of the high-risk patients with new onset AF did not receive AC therapy when leaving the ED. Furthermore, only half of the patients who began warfarin therapy were prescribed to use bridge therapy with LMWH. Based on these results, we have developed a standardized treatment protocol for AF patients. This protocol includes evaluation of the optimal treatment (rhythm or rate control) and execution of suitable AC treatment for each patient. We have also developed a special program to our ED’s patient data system in which the CHA₂DS₂VASc and HAS-BLED scores shall be filled in for each patient.

Introduction: The HEART score is a validated risk score for chest pain patients presenting at the Emergency department (ED). The HEART score  consists of five elements; history, ECG, age, risk factors and troponin. All of the elements are scored zero, one or two points, depending on the severity of the abnormality. Two of its elements, history and ECG, are subjective for interpretation by the ED physician. Little is known about the interobserver variability of history and ECG scoring and how this influences the interobserver variability of the final HEART score. The purpose of this study is to assess if the HEART score can be calculated reliably by different physicians.

Methods: For this study data from 125 patients was used. Each patient was scored by two cardiologists, two emergency physicians and two residents. Each physician scored the history and ECG of the 125 cases independently and blinded for the other elements of the HEART score. Interobserver agreement was measured by calculation of the intraclass correlation coefficient (ICC), using R statistics.

Results: Mean HEART score was 4.7 (95% CI 0.7-8.7). The analysis of the separate components yielded an ICC of 0.617 (range 0.0-1.0) for the history and an ICC of 0.512 for the ECG. On top of that, the agreement of the total HEART score between physicians was even higher, with an ICC for the HEART score of 0.888. The ICC for the total HEART score was 0.887 for the cardiologists, 0.882 for the emergency physicians and 0.986 for the residents.

Conclusion:  This study shows a very high ICC between different physicians at the ED, demonstrating a very high interobserver reliability of the HEART score. This supports the use of the HEART score by  several physicians. 

Abstract

Purpose: The purpose of the study is to detect whether using modified valsalva maneuver (VM) be more effective than standard VM in terminating SVT. 

Material and Method:This prospective randomized control trial, was conducted in an emergency department with patients who diagnosed SVT between 01.12.2015 - 31.12.2016. Participants were divided into two groups, randomly as standard VM or modified VM, as the first treatment with two-dimensional permutation blocks; in the order of arrival of the patients. The randomization was performed by envelope method. In both groups; the determined procedure for standard or modified VM were repeated up to three times in unresponsive patients. In both groups; if the maneuver is unsuccessful after three attempts, rescue medication with anti-arrhythmic treatment used. The primary outcome was defined to compare the success rate of achieving sinus rhythm after standard VM and modified VM.

Results: Totally, 56 patients were included randomly to this study; 28 were assigned to the standard VM, and 28 were assigned to the modified VM. Three of 28 patients(10.7%) in VM group and 12 of 28 patients(42.9%) in modified VM group were returned to sinus rhythm after intervention (p=0.007). Number of patients who need rescue treatment was lower in modified VM group, 16 (57.1 %) of 28, than in standard VM group, 25 (89.3%) of 28 (p=0.007).

Conclusion: Modified VM therapy is more effective than standard VM for terminating of SVT. It also indirectly reduces the need for anti-arrhythmic medication and indirectly causes fewer side effects. Therefore, we believe that modified VM can be considered as a first line treatment option in management of SVT according to results of this and previously studies. 

Background

Chest pain is the most common reason for emergency hospital admission, although the majority could be avoided with improved diagnostic technology. The Troponin-only Manchester Acute Coronary Syndromes (T-MACS) decision aid ‘rules in’ and ‘rules out’ acute myocardial infarction (AMI) with one blood test at the time of arrival in the Emergency Department (ED). T-MACS estimates the probability of AMI using basic data about a patient’s symptoms, signs, electrocardiogram and a single concentration of high sensitivity cardiac troponin (hs-cTn), a central laboratory assay.

Successful validation with a point of care (POC) cardiac troponin (cTn) assay would reduce turnaround time, helping to unburden crowded EDs. We aimed to prospectively validate T-MACS with a contemporary POC cTn assay.

Methods 

In this multi-centre prospective diagnostic accuracy study nested within the Bedside Evaluation of Sensitive Troponin (BEST) programme, we recruited patients with suspected cardiac chest pain presenting to nine EDs. Patients with another medical condition requiring hospital admission and those whose symptoms peaked >12h ago were excluded. Ethical approval was obtained and all participants provided written informed consent. 

Blood samples drawn on arrival were analysed for cTnI using the POC i-Stat assay (Abbott Point of Care, New Jersey, 99th percentile 80ng/L, LoD 20ng/L). The primary outcome was a diagnosis of AMI, which was defined in accordance with the Third Universal Definition based on central laboratory cTn analysis. To provide an adequate reference standard, the protocol required that all patients also undergo serial central laboratory cTn testing over at least 3 hours (for high sensitivity assays) or at least 6 hours (contemporary assays). 

T-MACS was computed using the original reported formula. We calculated sensitivity, specificity, positive and negative predictive values and positive and negative likelihood ratios.

Results 

A total of 622 patients were included in this analysis, of which 78 (12.5%) had AMI. Key results are summarised in Table 1. With a single POC cTn test, T-MACS would have ‘ruled out’ 41.6% (n=259) patients with a sensitivity of 97.4% (95% CI 91.0–99.7%) and a negative predictive value of 99.2% (95% CI 97.0–99.8%). Two AMIs were missed based on serial laboratory cTn concentrations. One patient did not receive a clinical diagnosis of AMI, received no treatment, did not undergo further investigation and had no adverse events within 30 days.

T-MACS would have ‘ruled in’ 7.1% (n=44) patients with a specificity of 99.2% (95% CI 98.0–99.8%) and a positive predictive value of 90.9% (95% CI 78.5–96.5%).

Conclusions

To our knowledge this is the first successful validation of a single test ‘rule out strategy’ using a POC cTn assay. Its use could enable almost immediate reassurance and discharge for >40% of patients with suspected cardiac chest pain.

BACKGROUND: In the last two decades we have observed a dramatically increase of the CT scans use at the emergency department (ED). As a result, there has been growing concern around the increase of medical radiation exposure and cancer. The positive rate of head CT in non-trauma patients presenting to the ED is low. Currently, indications for imaging are based on the individual experience of the emergency physician, which contributes to overuse and variability in imaging indication. In adult patients with syncope, without cranioencephalic trauma and with no alarm signs, there is variability and inadequacy in deciding to request a CT scan. The aim of this study is to ascertain the predictors of a positive head CT scan in patients with syncope and demonstrate the feasibility of clinical decision rule (CDR) to improve the adequacy.

METHODS: A systematic search for evidence was made based on a PICO question (Haynes pyramid). After the search PubMed and Embase (2005-2014) was consulted. Four reviewers reviewed all citations and select eligible ones for inclusion. Two authors independently appraised the quality of the studies and their degree of recommendation (GRADE). All disagreements among reviewers were discussed and resolved by a third independent reviewer. After the systematic review CDR was developed and included in the Electronic Clinical Record System at the Emergency department. We retro/prospectively reviewed all ED syncope patients 6 months before and after implementing the CDR. Medical records were assessed for analyzing the adequacy of the CDR and a telephone call was made after 30 days from discharge, in order to ensure that a related event had not occurred.

REUSLTS: The CDR has 5 items (alarm signs) and if any of them was present the CT scan should be performed. The items were: abrupt/severe headache after vigorous exercise, neurological focus, alteration of the level of consciousness, meningism and hypertension emergency. Both in the pre and post-implantation groups, 10% of the syncope had alarm signs. From patients with alarm signs, 10% had a positive CT result. In the cases that did not show alarm signs (90%) an evident differentiation was detected resulting in the Pre-CDR group being 46% unsuitable indications, compared to 7.6% after the CDR implantation, which shows that the goal of adapting CT indications was achieved. None of the patients that were discharged from the emergency department without a CT scan had a new event related with the syncope (CDR had a safety of 100%). Finally, a cost analysis was carried out. We found a reduction of total expenditure of CT scan from 50% to 20% post-intervention, which accounts a total of 85,848 euros savings annually.

DISCUSSION: We can conclude that, the implementation of the CDR for the indication of a CT scan, being the reason of the consultation the syncope at the emergency department, is safe and increases the adequacy reducing not only the costs but also the medical radiation exposure.

Background

Approximately 2300 in-hospital cardiac arrests (IHCA) occur annually in Sweden and the prognosis for those affected is poor with 29% survival to hospital discharge. A do-not-attempt-resuscitation (DNAR) order is issued when it is against the wishes of the patient that cardiopulmonary resuscitation (CPR) is performed, or when CPR is considered medically futile; that is when the chances of good quality survival are minimal. Emergency physicians are required to address the question of futility in the emergency department, but scientific support to fulfill this task is sparse. The Good Outcome Following Attempted Resuscitation (GO-FAR) score was developed in 2013 and has not been validated on a population basis outside the index population. The GO-FAR score is a summed score consisting of 13 prearrest variables with values ranging from -15 to 11 points reflecting the likelihood of good neurological survival at discharge measured as CPC 1 (patient is alert, able to work and lead a normal life, may have minor psychologic or neurologic deficits). A prearrest prediction tool for good outcome after IHCA would substantially contribute to daily clinical practice and reduce barriers for discussing DNAR orders. 

Methods

This validation study is based on a retrospective cohort of adult IHCA in Stockholm County 2013 to 2014 identified through the Swedish Cardiopulmonary Resuscitation Registry (SCRR). SCRR provides patient and event characteristics, personal identification numbers and CPC score at discharge. Data for the GO-FAR variables was obtained from manual review of hospital electronic patient records. The model performance was evaluated by quantifying discrimination and calibration, calculating the area under the receiver operating curve (AUROC) and evaluating the calibration plot with calibration-in-the-large and calibration slope.

Results

The final cohort included 717 patients with a 30-day survival of 27.5% and survival with good outcome at discharge of 22.3%. 61.9% were male, mean age was 72 years (SD 14 years) and 22.1% presented with ventricular fibrillation or ventricular tachycardia. In complete case analysis (526 cases) AUROC was 0.82 (95% CI 0.78 to 0.86) indicating good discrimination. The calibration slope was 1.38 (95% CI 1.08 to 1.68) and calibration-in-the-large -0.84 (95% CI -1.05 to -0.63) indicating that the GO-FAR score systematically underestimates the probability of good neurological survival.

Conclusion

The GO-FAR score shows satisfactory discrimination but dissatisfactory calibration in a cohort representing a Swedish population. It has the ability to distinguish a patient with good outcome from a patient with adverse outcome, but good outcome is systematically underestimated. Recalibration of the GO-FAR score is suggested before taken into clinical practice in Sweden.

Background

Triage systems have been implemented in most emergency departments (EDs) worldwide to minimize crowding and treatment delays that may adversely affect outcomes in acutely admitted patients. Triage systems are designed firstly to identify patients in need of immediate care, and second to provide risk stratification and ensure the optimal distribution of resources.

However, pre-existing triage systems are time consuming, supported by limited evidence, and could potentially be of more harmful than beneficial.

This trial prospectively compared a new simplified triage system with emphasis on clinical judgement with a traditional triage system with focus on clinical endpoints.

Methods

The Copenhagen Triage Algorithm (CTA) study was a large prospective, two-center, cluster-randomized, parallel, cross-over, open trial comparing CTA to a traditional triage system, Danish Emergency Process Triage (DEPT), which is a local adaptation of the internationally used ADAPT system.

All patients ≥17 years admitted to the ED in two large hospitals in a 10-month period were randomly cluster allocated to either CTA or DEPT triage with subsequent crossover. Based on vital signs and a subsequent clinical assessment by the ED nurse, CTA stratifies patients into 5 acuity levels. 

The study had a non-inferiority design with 30-day all-cause mortality as the primary endpoint. The non-inferiority margin was set at 0.5%. As a secondary endpoint, the study aimed to assess if CTA was superior in predicting 30-day mortality as assessed by C-statistics.

Results

A total of 45,977 patient visits were included. Of these 23,415 (50.9%) visits were triaged using CTA and 22,562 (49.1%) visits using DEPT.  Patients were well matched on baseline characteristics. The non-inferiority criteria was met, with the 30-day mortality among patients triaged with CTA and DEPT at 3.35% and 3.28%, respectively (p=0.68), a difference of 0.07% (95% CI: -0.26-0.40). Comparable results were observed for mortality at 48 hours (0.63% and 0.68%, p=0.58 using CTA and DEPT, respectively) and at 90 days (6.18 % and 6.36 %, p=0.41 using CTA and DEPT, respectively).

The triage level of patients in the CTA group was significantly lower (P<0.001) and CTA was superior in predicting 30-day mortality with an AUC of 0.670 (95% CI 0.650-0.690) compared to 0.638 for DEPT (95% CI 0.618-0.659) (P=0.03). Still there was no significant increase in mortality among low risk patients. A sensitivity analysis including vital signs, age, and sex showed no added risk among patients in the CTA arm for mortality at 48 hours (HR 1.07, 95% CI 0.97-1.18), 7 days (HR 0.98, 95% CI. 0.98-0.99), 30 days (HR 0.98, 95% CI 0.89-1.08) or 90 days (HR 1.03, 95% CI 0.96-1.1).

Conclusion

A new triage system based on vital signs and a clinical assessment by an ED nurse was non-inferior to a traditional triage algorithm with regards to mortality and superior at predicting 30-day mortality. 

Terapat Chantawong*, Pilaiwan Sawangwong*, Warawut Khangmak*, Chaiyaporn Yuksen MD,  Yuwares Sittichanbuncha ,MD. Department of Emergency Medicine, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.

 Backgrounds: Out of hospital cardiac arrest (OHCA) is one of the main causes of death in Thailand. Chest compression in narrow space may occur in ambulance or aircraft. Straddle chest compression is one of the technique to help the stability of chest compression.

Objective:  To compare the quality of chest compressions and exhaustion of healthcare provider between straddle chest compression and conventional chest compression in manikin model.

Methodology: This is an experimental study randomization into two different group, by using the SNOSE and block of four randomization by dividing them into group A (Straddle chest compression) and group B (conventional chest compression). Each participants were performed maximum 4 minutes hands only compression, quality parameters were recorded: compression rate and depth. The blood pressure, heart rate and respiratory rate of each participant were recorded both before and after chest compression.

Result: 124 participants (mean age 25.8 years and 27.6 percent are male, 40), the rate of chest compressions in conventional CPR and straddle CPR (126.2±17.1 and 127.0±21.0, p = 0.811), the average depth (43.8±9.6) and 43.4±9.0), p = 0.830). The blood pressure, heart rate and respiratory rate before and after chest compression in both method was no clinical significantly. 
  
Conclusion: The quality of straddle chest compression was as good as conventional chest compression. The exhaustion of healthcare provider between Straddle and conventional chest compression was no clinical significantly.

Background: Trauma is a main cause of death among young adults worldwide. Patients experiencing a traumatic cardiac arrest (TCA) certainly have poor prognosis but population-based studies are sparse.

Aim: Primarily to describe characteristics and 30-day survival following a TCA as compared with a medical out-of-hospital cardiac arrest (medical CA).

 Material and methods: A cohort study based on data from the nationwide, prospective population-based Swedish Registry for Cardiopulmonary Resuscitation (SRCR) between 1990 -2015. The definition of  a TCA in the SRCR is, a patient who is unresponsive with apnoea where cardiopulmonary resuscitation and/or defibrillation have been initiated and in whom the Emergency Medical Services (EMS) reported trauma as the aetiology. Outcome was overall 30-day survival. Descriptive statistics as well as multivariable logistic regression models were used.

Results: In all, between 1990-2015, 1 710 (2,4%) cases had a TCA and 69 123 had a medical CA. Overall 30-day survival gradually increased over the years and was in total among TCAs 3,8% compared to 8.0% following a medical CA (p<0.01). Factors associated with a higher 30-day survival were bystander witnessed TCA and having a shockable initial rhythm (adjusted OR 2.65, 95% C.I. 1.13–6.21 and OR 9.38 (4.44-19.85, respectively) while those given adrenaline had a poorer survival (adjusted OR 0.40, 95% C.I. 0.19 - 0.83).

 Conclusion: Even if the prevalence of TCAs is low and survival is poorer than in medical CAs, many affected individuals are young. Therefore, resuscitation in TCAs should not be seen as futile, but rather an area considered for improvements.  

Objective:

To determine which factors had the most impact on outcomes after pediatric out-of-hospital cardiac arrest (POHCA) now that protocol guidelines have become more aligned with those for adults, including emphasis on quality CPR.

Methods:

In an EMS jurisdiction using a comprehensive Utstein-style database, all POHCA cases over the previous 5 calendar years (1/1/12 through 12/31/16) -- since implementation of the latest international guidelines -- were analyzed to identify current predictors for return of spontaneous circulation (ROSC), hospital admission (HA) and survival to successful hospital discharge (SURV). Logistic regression models of traditional predictors were performed using JMP 12.0 for Mac.

Results:

Of 133 consecutive POCHA cases studied (61 % male), the interquartile range (IQR) for response intervals was 16 to 47 minutes (range: 0-490). As traditionally predicted, shorter times from arrest to EMS arrival were associated significantly with ROSC, HA and SURV (all p<0.0001) whereas witnessed arrest cases (only 13%) were not (p=NS). Still, in 95% of cases, the arrest was identified by a bystander prior to EMS arrival and, contrary to previous studies, chest compressions were performed by bystanders in 59% of cases. While the earlier CPR was provided by EMS personnel was itself significantly associated with ROSC, HA and SURV (all P<0.0001), some form of treatment before EMS arrival was provided in 54% of cases and such actions were strongly associated with ROSC, HA and SURV (p<0.0001 for all) whereas AED placement (50% of cases) was not.

Conclusion:

Whereas witnessed arrests and AED placement were not identified as contributing factors in this subpopulation of arrests, 1) shorter elapsed intervals from the moment of arrest to EMS arrival, 2) performance of CPR prior to EMS arrival and, in fact, 3) any treatment provided before EMS arrival, resulted in significantly higher rates of ROSC, hospital admission and survival beyond hospital discharge.

Goal: The aim of this study was to assess the survival of victims of out-of-hospital cardiac arrests due to ventricular fibrillations of three successive cohorts over a 10 years’ period.

Materials and methods: This was a retrospective observational study that compared three successive cohorts. The study was approved by an ethics committee. Inclusion criteria were: non-traumatic out-of-hospital cardiac arrests due to ventricular fibrillations shocked by an automated external defibrillator, and having benefited from prehospital advanced life support by a team managed by an emergency physician. Collected variables were: the period of occurrence of the cardiac arrest (1^st period = P1 - September 2005 to March 2007, 2^nd period = P2 - June 2011 to December 2012, 3^rd period = P3 - June 2014 to December 2015), the patient’s age and gender, the location of the cardiac arrest, the presence of a witness, the initiation of chest compressions by a bystander, the number of external electric shocks delivered by the automated external defibrillator, and the number of adrenaline injections administered by the advanced life support team. The primary endpoint was the hospital discharge. The secondary endpoint was the admission to the hospital with a palpable pulse. We used a logistic regression model to estimate the relationship between hospital discharge and the variables that were collected. We show the median [interquartile range] or the rate (%) according to the quantitative or qualitative nature of variables.

Results: A total of 2,648 patients (843 for P1, 830 for P2, 975 for P3) were included (62 years-old [51–74]; 1,608 (60.7%) men). Patients admitted to the hospital with a pulse during P1, P2, P3 were respectively 361 (43%), 464 (56%), 555 (57%) (p < 0.001). Patients discharged from the hospital during P1, P2, P3 were respectively 101 (12%), 166 (20%), 204 (22%) (p <0.001). In the multivariate analysis, factors associated with hospital discharge were: being a woman, a younger age, the occurrence of the cardiac arrest in a public area, and chest compressions by a witness. The rate of witnesses who performed chest compressions increased considerably over time (28% of cases for P1, 50% for P2, and 67% for P3). Adjusted odds-ratio associated with hospital admission were respectively: 1 for P1, 2.3 [1.8–3.0] for P2, and 1.8 [1.4–2.3] for P3.

Discussion: The survival of out-of-hospital cardiac arrests due to ventricular fibrillations shocked by an automated external defibrillator has improved over the past ten years. The increase of the rate of victims who benefited from chest compressions provided by a witness is one of the explanatory factors.

Conclusion: Efforts undertaken to improve the chain of survival, especially the early recognition of out-of-hospital cardiac arrests, an early cardiopulmonary resuscitation, and an early defibrillation must be continued.

Background: In the case of out-of-hospital cardiac arrests, 50% of ventricular fibrillations reoccur at least one time, mainly after the resumption of chest compressions. The aim of this study was to identify factors associated with the time between the resumption of chest compressions and the recurrence of the ventricular fibrillation (“CC-VF”). 

Materials and methods: We conducted a retrospective observational study. Collected variables were: patients’ age and gender, the presumed origin of the cardiac arrest, its location, the presence of a witness, the initiation of chest compressions by a bystander. We extracted from automated external defibrillators the rate of chest compressions, the time and length of each chest compressions’ interruption, the time between external electric shocks and the resumption of chest compressions (“EES-CC”), and “CC-VF”. Inclusion criteria were: out-of-hospital cardiac arrests, being over 18 years-old, one or more recurrences of a ventricular fibrillation shocked by an automated external defibrillator of basic life support teams. We performed a univariate analysis followed by a multinomial regression on repeating data; only variables associated with “CC-VF” with p < 0.2 in the univariate analysis were included in the multivariate analysis. We show the median [interquartile range] or the rate (%) according to the quantitative or qualitative nature of variables.

Results: Between 2010 and 2013, we recruited 266 patients (62 years-old [51.5–76]; 212 (80%) men) for a total of 1,047 episodes of recurrent ventricular fibrillations. A witness was present in 129 (48%) cases and performed chest compressions in 57 (21%) cases. The number of external electric shocks by automated external defibrillator ranged from 1 to 19. Concerning recurrences of ventricular fibrillations, 342 (32.7%) episodes occurred before the resumption of chest compressions, 129 (12.3%) were concomitant of the resumption of chest compressions, 170 (16.2%) occurred between 2.5 and 10 s after the resumption of chest compressions, and 406 occurred more than 10 s after the resumption of chest compressions. In the univariate analysis, factors associated with a shorter “CC-VF” were: being a male, an older age, a presumed cardiac origin of the cardiac arrest, the presence of a witness, a longer “EES-CC”, faster chest compressions’ rates, and greater chest compressions’ ratios. In the multivariate analysis, an older age and a faster rate of chest compressions were associated with a shorter “CC-VF”. A longer time between the first alert to the dispatch center and the initiation of chest compressions was associated with a longer “CC-VF”.

Discussion: This is the first study to identify factors associated with the time between the resumption of chest compressions and the recurrence of a ventricular fibrillation. If preliminary results presented above are confirmed, new approaches could be suggested to deal with the recurrence of ventricular fibrillations.

Background: Circadian variation in occurrence of out-of-hospital cardiac arrest (OHCA) is an observation which has been reported from several parts of the world. Mostly, diurnal variation was shown exhibiting low incidence of OHCA at night and a two daytime peaks, in the morning and late afternoon. However, this variation can be related to geographical regions and the validity of the results is extremely dependent on the quality of the data collection. Therefore we have analyzed the Czech and Slovak data from the EuReCa ONE study to investigate whether there is any local significant variation of OCHA events treated by Emergency Medical Services (EMS) in time and if so, whether it depends on geographic variables.

Methods: In an international clinical study EuReCa ONE (European Registry of Cardiac Arrest), data on all EMS treated OHCA events were collected from the entire territory of the Slovak Republic (5421352 inhabitants) and several administrative regions of the Czech Republic (4350000 inhabitants) in the period from 1.10.2014 to 31.10.2014. Data were processed and analyzed for circadian and infradian variability.

Results: For the selected period, a total of 659 cases of confirmed resuscitated OHCA events was reported. Significant circadian variation was observed, with very low occurrence in the night (approximately 2% of all episodes each hour) followed by three peaks in the daytime, in the 9^th, 16^th and 20^th hour (8.2, 6.4 and 7.4 % of all episodes, respectively, p<0.05). During the week, OHCA events were the most frequent on Fridays while the least common on Tuesdays (16.8 versus 12.1 % of all events, p=0.019). In the Czech Republic, OHCA was more frequent at weekends than in Slovakia (31.2 versus 22.4 % of all episodes, p=0.013). The lowest thirty-day survival or survival to hospital discharge was observed in OHCA events that occurred on Tuesdays, while highest in episodes that occurred on Thursdays (6.2 versus 20.6 %, p=0.011). Time dependent variation of survival patterns were We did not found any differences between the Czech and Slovak republics in survival variation.

Discussion: In the selected regions and time interval we have identified a marked circadian and infradian variability of OHCA events occurrence and of their survival as well. While the variability of the events in time was partly country-dependent, survival was not. Further investigation of this phenomenon may lead to a better understanding of the circumstances leading to cardiac arrest and improve prevention of this cardiovascular catastrophe.

References: Gräsner JT et al. EuReCa ONE-27 Nations, ONE Europe, ONE Registry: Aprospective one month analysis of out-of-hospital cardiac arrest outcomes in 27 countries in Europe. Resuscitation 2016;105:188-95.

Objective

To compare mortality, length of stay and cause for admission in adult emergency department (ED) patients at the Royal Infirmary of Edinburgh with a delay in intensive care unit (ICU) admission of up to 48h with a group of patients admitted directly from the ED to the ICU.

Background

There had been no study of the differences in patient cohort, or patient outcome between patients being referred directly for critical care support, or those being admitted from a ward within 48 hours. We aimed to identify a cohort of patients who had not been referred in ED, but who required critical care support early on during their hospital admission, and therefore may have benefitted from an earlier referral to ICU.

Sample & method:

We performed a retrospective cohort study in a 900-bed university teaching hospital over a 3 month period analysing data for all ITU admissions over the time period. One hundred and twenty-four adult emergency department patients admitted to the intensive care unit either directly from the emergency department (direct group) or within 48h of ward admission (delayed group) were identified.  The main outcome measures investigated were mortality, length of hospital stay and cause for admission to ITU. Exclusions included those transferred to the hospital directly to a ward, those who were admitted from recovery or theatres, and those assessed in the Primary Assessment Area.

Findings: 

1. Mortality in the delayed group was 8% lower than the direct group (34% vs 26%).
2. Length of hospital stay was 11.5 days longer in the delayed group compared with the direct group (25 days vs 13.5 days).
3. Of the patients in the delayed group, 59% of the cohort presented with sepsis.

Conclusions

Our study showed that patients in the delayed group had a lower mortality compared with the direct group which we hypothesize to be secondary to a less critically ill patient cohort. We noted an almost double length of hospital stay in the delayed group, which holds significant implications for increased morbidity in that cohort. Of these delayed patients requiring ITU care, we noted that over half presented with sepsis. These results raised the question of whether our sepsis identifiers in the ED are being utilised appropriately and/or are clinically effective. There are two sepsis tools used: Systemic Inflammatory Response Syndrome  (SIRS) criteria “ and “Quick Sepsis Organ Failure Assessment” (qSOFA).

We performed further work to assess whether the sepsis tools of SIRS criteria and qSOFA would have helped identify the delayed patients with sepsis earlier. It identified that 60% of patients scored positive based on the SIRS criteria, but only 10% of patients scored positive based on the qSOFA criteria. It highlights that while there is a strong evidence base for qSOFA in risk stratification, it may be a poor indicator of the presence of sepsis in these patients and therefore should not be the sole sepsis screening tool within the department. 

Background: Though oxygen is one of the oldest drugs available, it is still the most inappropriately administered drug. It is a very commonly used therapeutic agent and is the treatment for hypoxaemia. There has been increasing criticism of the unrestricted use of O2 therapy over the past few years, and it is still controversial. This has led to over utilization of this very expensive resource.

Methodology:This was a cross sectional observational study done in the Emergency Department (ED) of a large tertiary care hospitals in South India. All patients who were administered oxygen over a 3 week period in April 2016 were included in the study. Details of oxygen administration and outcome were analyzed. Oxygen administered without hypoxia was considered as inappropriate.A basic cost analysis was also done. This study was approved by the ethics committee of Christian Medical College, Vellore and patient confidentiality was maintained using unique identifiers.

Results: 15.4 % (363/2356) of the patients presenting to the ED were administered oxygen. The mean age was 50.2 ± 17 years. There was a male predominance (66.4%). Majority (67.2%) were triaged as priority 1 patients. The common reasons for initiation of oxygen therapy were dyspnea (56.7%), low sensorium (14.0%), intubated elsewhere (11.0%), polytrauma (7.2%) and seizures (4.4%).The mean duration of Oxygen therapy was 7.3 ± 4.6 hours. 36.6% of patients were administered oxygen inappropriately. This resulted in wastage of 1376 euros in three weeks through treatment cost. 65.3% were admitted, 20.1% were discharged stable from ED and 11.8% left against medical advice. ED mortality rate was 2.8% and an additional 11.6% expired during hospitalization.

Conclusion:Oxygen is inappropriately administered in 1/3 of patients presenting to ED and increases the treatment cost for patients. Oxygen therapy should be strictly regulated to minimize its wastage.

Background: Sepsis-induced myocardial dysfunction (SIMD) is established in about 50% of septic patients; aim of this study was to assess if SIMD is a reversible condition by mean of strain echocardiography.

Methods: Unselected patients affected by severe sepsis and septic shock admitted consecutively to a High Dependency Unit from the Emergency Department between October 2012 and December 2016 were prospectively enrolled. An echocardiogram was performed within 24 hours from the diagnosis of sepsis (ECHO1). LVEF was calculated using LV volumes derived by LV planimetry by manual tracing (Simpson’s rule) and was considered normal if >55%. The global longitudinal strain (GLS) was evaluated from apical LV views, with a commercially available system (Philips Q-LAB ver. 8.1) and was considered normal if <-14%; RV systolic function was evaluated through Tricuspidal Annular Systolic Posterior Excursion (TAPSE). In a consecutive group of survivors we repeated an echocardiogram 3 months after hospital discharge (ECHO2).

Results: Among 177 patients who underwent an echocardiogram within 24 h from sepsis diagnosis, 127 (72%) survived at 28 days and 44 patients (35% of survivors) accepted to repeat an echocardiographic evaluation after three months and they represent our study population; no significant differences were found between participants and non participants in term of LV (LVEF: 54 ±14% vs 50 ±15%; GLS -12.9 ±3.7 vs -11.8 ±3.4, tutti p=NS) and RV systolic function (TAPSE: 2.1 ±0.5 vs 1.9 ±0.5, p=NS). ECHO1 showed an LV systolic dysfunction in 26 (59%) and RV systolic dysfunction in 9 (21%); at ECHO2 LV systolic function returned to normal values in 13 patients and RV systolic function in 7, but a new systolic dysfunction was appreciated in 2 patients for LV and in 7 for RV. Considering LVEF analysis, the prevalence of LV dysfunction at ECHO1 was 49% and dropped to 23% at ECHO2; according to GLS it fell from 59% to 29% p=0.010 between ECHO1 and ECHO2, p <0.001 between evaluation by mean of GLS or by LVEF). Overall 50% of patients presented at ECHO2 a mono- or biventricular systolic dysfunction: patients with persistent dysfunction did not have a higher prevalence of coronary heart disease (14 vs 9%) or, during the acute phase, they did not develop more frequently a septic shock (29 vs 20%) or needed high-dosage vasopressors (11 vs 9%, all p=NS). Biomarkers levels in the acute phase were comparable between patients with reversible or irreversible dysfunction (Troponine: 0.43 ±1.57 vs 0.83 ±1.61 microgr/L; NTproBNP 6615 ±6501 vs 7501 ±12785 pg/mL).

Conclusions

SIMD has a significant incidence and it persists beyond the acute phase of the septic process in a relevant proportion of patients, but we did not find any useful parameter to predict SIMD reversibility; strain echocardiography was superior to conventional methods in identifying systolic dysfunction. 

Background: Left (LV) or right (RV) systolic ventricular dysfunction has been found in about 50% of septic patients. The aims of this study were:  1) to evaluate biomarkers’ diagnostic accuracy in identifying patients who develop SIMD; 2) to evaluate prognostic significance of biomarkers.

Methods: In 177 patients diagnosed with severe sepsis/septic shock and admitted in our ED-High Dependency Unit between August 2012 and December 2016; an echocardiogram was performed within 24 hours from admission. We evaluated LV systolic function using Global Longitudinal Strain (GLS) and Ejection Fraction (EF) measurement and RV systolic function with Tricuspidal Annular Plane Systolic Excursion (TAPSE). We divided our population in two subgroups:  patients who had a mono- or biventricular systolic dysfunction (D+) and those who hadn’t (D-). We referred to a GLS > -14% for LV systolic dysfunction and TAPSE <16 mm for RV systolic dysfunction. Biomarkers’ levels were measured both at the time of admission (T0) and after 24 hours (T1), considering them both as continue and dichotomized values (TnI: ≤0.1 or > 0.1 µg/L; NTproBNP: > or ≤ 6000 pg/mL). Day-7 and day-28 mortality were our primary end-point.

Results: Most frequent comorbidities were diabetes (27%), hypertension (55%) and neoplastic disease (31%); lung was the most common primary infection site (54%). One hundred twenty-seven patients (72%) showed an LV dysfunction and 54 (30%) a RV dysfunction; overall D+ group included 136 patients.

TnI T0 and TnI T1 levels were significantly higher in D+ patients compared with D- subjects (T0: 0.78 ±2.43 vs 0.15 ±0.29; T1: 1.00 ±2.60 vs. 0.19 ±0.48). T0 NTproBNP value was significantly higher in D+ than in D- patients (18292 ±34019 vs 9329 ±15616); dichotomized values did not show a significant different distribution between D+ and D- patients. ROC analysis showed an area under the curve (AUC)= 0.64 for T0

TnI,0.67 for T1 TnI, 0.60 for T0 NTproBNP and 0.65 for T1 NTproBNP.

Day-28 mortality was 28% (n=50). Biomarkers’ values did not show any significant association with an

increased mortality rate at univariate analysis; a more compromised value of GLS, TAPSE and EF was significantly associated with an increased day-7 and day-28 mortality; ; after adjustment for age and SOFA, an increased GLS was associated with an increased day-7 mortality (RR 1.18, IC 95% 1.04-1.35, p=0.010) while abnormal TAPSE and GLS were associated with increased day-28 mortality rate (respectively, RR 0.30, IC 95% 0.12-0.72, p=0.007 and RR 1.11, IC95% 1.01-1.25, p=0.041) while EF did not demonstrated any prognostic significance in a similar multivariable model.

Conclusion: SIMD has a significant incidence and is associated with an increased mortality rate; the levels of biomarkers, commonly considered as a result of myocardial damage, are higher in patients who present ventricular systolic dysfunction than in others but the prognostic discrimination ability is poor.

Background: Aim of this study was to examine the feasibility and diagnostic accuracy of VCCI and velocity time integral variation after passive leg raising (PLR) in an unselected population of critically ill patients admitted to a sub-intensive clinical setting.

Methods: This is a prospective, observational, pilot study. Unselected critical patients admitted in an Emergency Department High-Dependency Unit (ED-HDU) were evaluated by transthoracic echocardiography to measure vena cava collapsibility index (VCCI) and aortic velocity (AoV)  variation during PLR. Conventional LV and RV diastolic dimensions and systolic function (LV ejection fraction, EF, and Tricuspid Annulus Systolic Posterior Excursion, TAPSE) were measured. According to VCCI, patients were considered fluid-responders when the value was ≥40%. According to AoV variation after PLR, a positive hemodynamic response was defined as an increase in AoV ≥ 10%. Whenever possible, both VCCI and AoV variation during PLR were evaluated. According to echocardiographic evaluation, three therapeutic options were considered: no intervention, administration of fluids or diuretics. Any change in the therapeutic strategy by the treating physician in the following 12 hours was annotated into the clinical records.

Results: we enrolled 53 patients, mean age 73±14 years; the two most frequent reasons for ED-HDU admission were sepsis (75%) and COPD re-exacerbation (8%); in 5 (10%) patients echocardiographic evaluation was not feasible. VCCI was feasible in 35 (66%) patients, while PLR could be performed in 33 (62%). Eighteen patients were managed according to VCCI: 13 were non FR, while 5 were FR and were treated with fluid boluses. In the following 12 hours, in 4 non FR patients and in 3 FR patients therapeutic strategy was modified (7/18, 39%). Thirty-one patients were treated according to PLR: among 18 FR patients, 16 received a fluid bolus while 13 non FR did not receive fluids and this therapeutic strategy was maintained in all but one FR patient in the following  12 hours (1/31, 3%, p=0.002). In the group of patients managed by PLR 18 also underwent VCCI evaluation which was discordant with PLR in 3 patients. Finally we compared LV and RV dimensions and systolic function between patients in whom VCCI correctly identified FR (n=26) or it did not (n=10): presence of LV dilatation  (LV diastolic diameter >55mm; 10% in both groups), RV dilatation (At least 2 of the three conventional RV diameter over normal limits; 62% in patients correctly identified vs 46% in patients not correctly identified), LV systolic dysfunction (LVEF <50%; 44% vs 33%) and RV systolic dysfunction (TAPSE

 Conclusions: we confirmed a poor diagnostic accuracy for VCCI independent to LV and RV dimensions and systolic function; VTI variation during PLR showed a very good diagnostic performance.   

Introduction: The aim of this study was to compare the prognostic value of MEWS (Modified Early warning System) score and lactate dosage absolute value and trend over 2, 6 and 24 hours after admission, in order to identify the most appropriate timing to evaluate score’s evolution.

Methods: In the period November 2011-December 2016, 269 patients enrolled in a prospective study aiming to find reliable biomarkers for an early sepsis diagnosis. Patients admitted to our High-Dependency Unit from the Emergency Department with a diagnosis of severe sepsis/septic shock were eligible. At ED-admission (T0), after 2 hours (T2), 6 hours (T6) and 24 hours (T24) from the initial diagnosis, we evaluated lactate and MEWS score; score differences over 2-hour (ΔMEWS-2H), 6-hour (ΔMEWS-6H) and 24-hour time interval (ΔMEWS-24H)were calculated. Lactate absolute values (analyzed as continuous values and ≤ or >2 meq/L) and lactate clearance (dichotomized as ≤ or >10%) were evaluated at the same time intervals. The primary end-point was in-hospital mortality.

Results: Mean age of the study population was 74±14 year, 59% male gender; main comorbidities were arterial hypertension (61%), diabetes (33%), neoplasia (22%) and chronic kidney disease (24%). The most frequent infection source was respiratory (45%) and 41% of patients developed a septic shock. Overall in-hospital mortality was 26%. Mews score was significantly higher in non-survivors compared with survivors at all evaluations (T0: 4.3±2.1 vs 3.6±2.0, p=0.028; T2: 4.1±1.6 vs 2.9±1.7, p<0.001; T6: 4.3±2.1 vs 2.6±1.7, p<0.001; T24: 3.9±2.6 vs 2.3±1.7, p<0.001); repeated measures analysis confirmed a significant difference within subjects (p<0.001) and between survivors and non survivors, with a continuous decrease in the first group and a flat trend in the second one. Score variation was negligible in non survivors (T2: -0.15±1.51 vs -0.71±2.03, p=0.062; T6: -0.19±2.15 vs 0.97±2.01, p=0.023; T24: 0.11±2.61 vs -1.31±2.01, p<0.001); after dichotomization of the score variation on the basis of the median value of this study population (≤ or >-1), only at 24-hour evaluation a variation>-1 was significantly more frequent among non survivors (60 vs 34%, p=0.004). Lactate dosage was significantly higher in survivors at all evaluations except for T2 (T0: 3.8±3.9 vs 2.7±2.5, p=0.043; T2: 3.5±3.8 vs 2.4±2.6, p=0.067; T6: 3.4±3.8 vs 1.7±1,3, p=0.003; T24 4.1±5.7 vs 1.4±0.9, p=0.002); a value>2 meq/L was significantly more frequent among non-survivors only at T24 (43 vs 19%, p=0.001). A lactate clearance >10% was significantly more frequent among survivors at T6 (71 vs 48%, p=0.006) and tendentially at T24 (67 vs 50%, p=0.053), not significant at T2 (59 vs 47%, p=.208).

Conclusions: Vital signs aggregated into MEWS score and lactate dosage were significantly worst in non-survivors compared with survivors at the moment of sepsis diagnosis; a 2-hour interval appears too short to allow a prognostic evaluation. 

Introduction: The aim of this study was to compare the prognostic value of score trend at 6 and at 24 hours after admission, in order to identify the most appropriate timing to evaluate score’s evolution.

Methods :In the period November 2011-December 2016, 269 patients enrolled in a prospective study aiming to find reliable biomarkers for an early sepsis diagnosis. Patients admitted to our High-Dependency Unit from the Emergency Department with a diagnosis of severe sepsis/septic shock were eligible. Exclusion criteria included presence of severe cognitive impairment inducing associated with immobilization syndrome lasting from more than three months; age

Results: Mean age of the study population was 74±14 year, 59% male gender; main comorbidities were arterial hypertension (61%), diabetes (33%), neoplasia (22%) and chronic kidney disease (24%). The most frequent infection source was respiratory (45%) and 41% of patients developed a septic shock. Overall in-hospital mortality was 26%. SOFA score was significantly higher non-survivors compared with survivors at all the evaluations (T0: 6.1±2.7 vs 5.0±2.7, p=0.013; T6: 7.8±3.1 vs 6.1±2.9, p<0.001; T24: 8.5±3.5 vs 5.3±2.6, p<0.001). Discriminative analysis by ROC curves showed an improving prognostic stratification ability in following evaluations (T0: area under curve, AUC, 0.62, 95%CI 0.54-0.70, p=0.007; T6 AUC 0.67, 95%CI 0.59-0.775, p<0.001; T24 AUC 0.77, 95%CI 0.70-0.85, p<0.001). Based on ROC curve analysis, we identified the value 3.5 as that having a good sensitivity a specificity (98 and 74%): a SOFA score lower than 3.5 was significantly more frequent among survivors at all evaluation points (T0: 32 vs 17%, p=0.039; T6: 21 vs 6%, p=0.022; T24 26 vs 2%, p=0.001). ΔSOFA-T6 (1.8± 2.3 vs 1.1± 2.0, p=0.025) and ΔSOFA-24H (2.5±3.3 vs 0.3±1.9, p<0.001) were significantly higher in non-survivors compared with survivors. A ΔSOFA value at either evaluation point >1 (median value in our study population) was significantly more frequent among non-survivors (T6: 57 vs 35%, p=0.006; T24: 61 vs 28%, p<0.001). Patients with a SOFA score >3.5 and a score variation >1 showed a significantly higher mortality rate at either T6 and T24 evaluation, compared with patients who presented only one of the previous values or neither (T6: 57% vs 38% vs 6%, p=0.003; T24: 61% vs 37% vs 2%, p<0.001).

Conclusions: Prognostic value of SOFA score was modest at the moment of sepsis diagnosis; at a 6-hour interval, useful prognostic information could be obtained both from absolute score values and score variation, which were further confirmed at the 24-hour evaluation.

OBJECTIVE

To analyze the education in pediatric emergency medicine (PEM) given to pediatric residents and the research carried out by them in Spain.

METHODS

Descriptive cross-sectional study based on web surveys. First survey, regarding characteristics of PEM education and research in the Emergency Department (ED), was distributed to the directors of pediatric EDs included in the Spanish Society of Pediatric Emergencies. Respondents were asked to distribute a second survey to their residents and pediatric assistants. Only EDs with more than 30% of respondents were included for the descriptive analysis of all the variables.

Main outcome variables were the level of satisfaction within resident education (on a scale from 0 to 10) and the number of papers published in a peer-reviewed journal in the last 5 years. Multivariate analysis was made to assess associated factors between them.

RESULTS

First survey was sent to 83 directors and 42 (50.6%) answered it. In 33 (78.6%) EDs more than 30% of respondents fulfilled the second survey, including finally 376 (92.8%) for analysis (196, 52.1%, fulfilled by residents).

Median value of resident’s satisfaction with PEM training in each hospital ranked from 5 to 9. Factors associated with higher values were having education quality indicators, closer supervision of clinical practice and a structured evaluation of resident’s PEM skills when finishing the working shift or the rotation in the ED period.

In the previous 5 years, the average of research training activities by ED was 1 (IQR 0-3), with 11 EDs (33.3%) having no activity. Around 50% of respondents considered that research was not adequately supported at their EDs.

Level of resident’s satisfaction with research in PEM in each hospital are shown in figure 1. Sixty-eight respondents (18.1%) had published at least one paper on a peer-reviewed journal (residents, 17, 25%). Associated factors with having a paper published were the existence of a research director in the ED, having research quality indicators, self-perception of residents that the research was supported in the ED and having performed any research training activity in the previous 5 years.

Those residents with at least one paper published rated higher their education [mean= 8 (CI95% 7.34-8.66) vs those with no publications, 7.1 (CI95% 6.87-7.33)]

CONCLUSION

Significant variability in PEM education and the research was noted among Spanish pediatric residents. An adequate organization of the EDs seems to be essential to improve education and research. 

Background:

Diabetic ketoacidosis (DKA) is an acute and potential life-threatening complication of diabetes mellitus. The mainstay in the treatment of DKA involves the administration of regular insulin. However, the route and the dose of insulin remains controversial. This study was designed to compare the safety and the efficiency of two protocols of intravenous insulin (IV): Protocol (A) Intravenous bolus of regular insulin 0,1UI/Kg followed by a continuous IV infusion at the dose of 0,1UI/Kg/H; Protocol (B) a continuous IV infusion of regular insulin at the dose of 0,14UI/Kg/H without bolus.

Methods:

Prospective randomized study of patients aged more than 18 years with moderate to severe DKA hospitalized in the emergency department. Patients were devised into two groups: Group (A) received protocol (A) and Group (B) received protocol (B). Standardization of:1) the fluid therapy with normal saline and 5% dextrose 2)the potassium replacement. Data on glucose level, pH, serum bicarbonate, anion gap, intravenous fluid administration, and length of stay were collected. Outcomes data were: time to recovery, time to glucose control (<250mg/l), insulin dose to recovery, occurrence of complications: hypoglycemia, hypokalemia, recurrence of DKA.

Results:

We enrolled 164 consecutive DKA patients. Exclusion of 39 patients. The mean age = 39 +/- 18 years, sex ratio =0.97. DKA occurs more in type 1 diabetes n=87(47.6%) than in type 2 n= 64(39%) and was inaugural in 22 patients (13.4%). There were no differences between the two groups in clinical and biochemical data Group (A) versus Group (B) : mean age (37+/-17 vs. 37+/-17 years; p=0.95),sex ratio(0.84 vs. 0.88),Blood glucose level  (30.2+/-9.9 vs. 32.5+/-11.9 mmol/l; p=0.27), pH(7.14+/-0.13 vs. 7.15+/-0.12; p=0.7), anion gap (28.63+/ 5.74 vs. 28.9+/-7.21; p=0.8) ; also in outcomes data Group(A) vs. Group(B): time to recovery (17.6+/-13 vs. 17.4+/-21.5 hours; p=0.9), insulin dose to recovery (76.5+/-55.1 vs. 74.9+/-35.3 UI; p=0.8) length of stay in intensive care unit (28.3+/-18.2 vs. 32.4+/-20.3 hours; p=0.3),complications : hypoglycemia(n= 4 vs. 10; p=0.12 ) , hypokalemia (n= 32 vs. 31; p=0.33) , recurrence of DKA (n=1 vs. 7; p=0.31) .

Discussion:

 These two protocols of IV insulin infusion were safe and had a comparable efficiency without majoring the risk of complications.

Background     Contaminated environmental surfaces are known provide an important potential source for transmission of healthcare associated pathogens  and prehospital treatment have been associated to increased risk of infection. Nevertheless, few studies present and discuss prehospital hygiene, resulting in limited knowledge and understanding of related challenges. Our aim was to assess microbial contamination and influencing factors in order to assess the extent of the risks and illuminate eventual solutions.

Methods          A nationwide, semi-blinded, cross-sectional study was conducted in Denmark from August to November 2016. Using a combined swab/agar method, samples from environment, equipment and personnel were randomly collected from 80 ambulances and crew, in-between patient courses, after cleaning. Focus was on colony forming units (cfu) and healthcare associated pathogens. In addition, explanatory variables e.g. hours from last thorough cleaning, area of service (rural/city) and number of patient courses within the shift, were collected and used in bivariate analyses.

Results              800 sites, showed an average of 11.3 cfu/cm² (environmental sites e.g. blood pressure cuff, patient harness and defibrillator 5.01 cfu/cm², hands of the personnel 11.1 cfu/cm² and uniforms 30.6 cfu/cm²). Staphylococcus aureus, Enterococcus and Enterobacteriaceae were found on 10, 3.4 and 0.5 % of the imprints, respectively. One imprint was MRSA, two were VRE but none was ESBL. Furthermore, we found no correlation between the explanatory variables and the degree of microbial burden.

Conclusion      Our study underlines that microbial contamination and related challenges in the EMS calls for further attention. As seen in prior studies, several sites were contaminated with healthcare associated pathogens.  However, neither time from cleaning, number of patients nor area of service were of influence on the degree of contamination, hence not contributing to an explanation. Future research on hygienic challenges and routes of transmission is recommended.

Background

Patients in cardiac arrest must receive algorithm-based management such as basic life support and advanced (cardiac) life support. International guidelines dictate diagnosing and treating any factor that may have caused the arrest or may be complicating the resuscitation. Ultrasound is recognized to be of potential value in this process. Also, it is shown to be feasible in a prehospital setting. We aim to determine the impact of prehospital echocardiography during cardiopulmonary resuscitation (CPR) and its impact on treatment decisions in a Dutch physician-staffed helicopter emergency medical service (HEMS).

Methods

We conducted a prospective, observational study from February 2014 through October 2016 of patients treated by the Nijmegen HEMS. Inclusion criteria were CPR irrespective of its cause and concurrent echocardiography. Echocardiography-trained physicians performed the examinations within the same time window where chest compressions are interrupted to analyze heart rhythm. Data collection included patient demographics; type of incident; CPR details and outcome; vital signs; ultrasound findings (ventricular dimensions; global myocardial function; any pericardial fluid); physician-reported image quality and ease of procedure; impact on treatment decisions. Outcome parameters were: impact on treatment decisions; characteristics of the population; feasibility of echocardiography in this setting.

Results

Of 6694 recorded scrambles and 3229 patients treated, 425 underwent CPR. In 56 patients 102 ultrasound examinations were documented. Treatment decisions were impacted in 49 patients (88% - CI 79.5-96.5%) and in 62 (61% - CI 51.5-70.5%) ultrasound examinations. Overall, we found 78 changes. They were termination of CPR in 32 patients (57%) and continuation hereof in 21 (38%). Other changes were related to fluid management (14.3%), adjustment of drugs and doses (14.3%), and choice of receiving hospital (5.4%). The causes of cardiac arrest were trauma (48%), cardiac (21%), medical (14%), asphyxia (9%), and other (7%). The ease of the entire procedure was scored a median of 7 (numeric rating scale 1-10) and image quality per examination good (59%), moderate (29%), or poor (12%).

Discussion

Ultrasound impacts management in 88% of patients. This is in accordance to results of (peri-) resuscitation studies by Breitkreutz (78%) and Shokoohi (12%-31% in different categories, overall unknown). Ultrasound images can help explain futile care to caregivers and relatives, even if the sensible decision already is to terminate CPR. Prehospital (traumatic) CPR is often impeded by stress, time pressure, environmental factors including a restricted workspace, and an inaccessible ultrasound machine. This likely explains the limited number of inclusions. In conclusion, prehospital ultrasound during CPR in our HEMS significantly impacts patient treatment. This suggests echocardiography should be a standard tool in every prehospital resuscitation.

Background

Time based standards have been used as a key performance measure for EMS internationally, despite a lack of evidence that they actually lead to good clinical care. Achievement of standards in an environment of rising demand potentially leads to operational behaviours that may be inefficient such as dispatching multiple vehicles before the problem is known.  In England, the Ambulance Response Programme is developing new operational models of care. One strategy has been to test if additional call triage time before starting the response interval clock start can lead to better use of resources and improved dispatching.

Methods

A controlled before and after time series analysis of a new intervention – Dispatch on Disposition (DoD)– comprising a short set of pre-triage questions to identify time critical emergencies needing immediate dispatch of a resource and up to 4 minutes to triage all other calls (compared to the existing 60 seconds). DoD was implemented in 6 of the 10 regional services in England and 4 services were control sites. We measured weekly trends in average resource allocation per call and resources on scene for different call types (life-threatening, emergency, urgent) for 1 year before and 7 months after implementation, and used time series regression models to compare changes between intervention and control sites adjusted for seasonality, call volumes and hours lost at hospital handover. We also conducted a survey of dispatch and operational staff.

Results

There was a statistically significant reduction in average resources allocated per incident of -0.1 for life-threatening calls, -0.06 for emergency and -0.12 for urgent, and a reduction in resources arriving on scene per incident of -0.006 for life-threatening calls and -0.02 for urgent in the intervention groups compared to control. Scaled up the resource allocation reductions will potentially produce an additional 10243 whole resources available to respond per week in England. Dispatch staff reported they were better able to manage call queues and allocate the right rather than any response. Operational staff reported a substantial reduction in calls where they were cancelled before arriving on scene.

Conclusions

Prior to DoD ambulance services in England had to dispatch a resource within 60 seconds of receiving an emergency call and in order to achieve a response time of 8 minutes for the most serious calls multiple resources could be sent before establishing if the call was serious. Allowing additional time to properly triage calls other than those likely to be life-threatening has created efficiencies by substantially reducing multiple resource allocations and freeing up vehicles for other calls. In an environment of increasing demand and diminishing resources this allows better use of existing resources.

Background: The Out-of-Hospital Cardiac Arrest Outcomes (OHCAO) project aims to understand the epidemiology and outcomes of out-of-hospital cardiac arrests (OHCA) across the UK. Significant variation exists between ambulance services in outcomes for patients with attempted resuscitation following OHCAs. Importantly, a great deal of the variability of reported outcomes can be traced back to the quality of data that results are based on.

This study is a sub-project of OHCAO and aims to establish the feasibility of producing a registry of OHCAs by linking OHCAO data to the Office for National Statistics (ONS) mortality data, via NHS (National Health Service) Digital, to improve data quality and establish accurate 30-day survival outcomes for OHCAs.

Methods: Data were collected from 1^st January 2014 to 31^st December 2014 as part of a prospective, observational study of all OHCAs attended by ten English NHS Ambulance Services. 28,729 OHCA cases had resuscitation attempted by Emergency Medical Services and were included in the study. Of these, a randomly selected sample of 3,120 cases (10% of total) were securely transferred to the ONS. This allowed OHCAO demographic data to initially be matched to NHS patient demographic data, using the NHS Digital list cleaning service to return previously missing data. Following this, cases were linked to ONS mortality data to provide accurate death dates where applicable to calculate 30 day survival.

Results: A total of 80.5% of OHCAO cases were matched to the ONS database. OHCAO collected complete demographic datasets on 868 (27.8%) cases. Using the linkage process, missing demographic data was retrieved for 72.7% of the 2,249 cases with incomplete data. Confirmation of 30-day survival improved by 37.6% with a reduction in unknown 30-day survival status from 46.1% to 8.5%. The most important data point required for linkage was the NHS number which provides a unique patient identifier. However, it was only retrieved by the OHCAO project for 31.7% of cases. This study found that if at least 3 other demographic data points were collected, the NHS number could be retrieved using the linkage process in up to 89.9% of cases.

Discussion: Ensuring high data quality is essential as this forms the basis of decisions that ultimately impact on changes in care and healthcare resource allocation. Data linkage was shown to successfully improve the quality of OHCA demographic data and survival status 30 days after OHCA. Importantly, this process has allowed the provision of demographic details to allow patients to be followed longitudinally, potentially to assess morbidity following OHCAs. The linkage process can be used to produce a registry of OHCAs and information gained from this can be fed back to institutions providing source data to improve OHCA outcomes.

Fire Medical Response Early Indications of Clinical Value

Background

In the UK, response to serious medical emergencies has been solely provided by the National Health Service (NHS), but UK Fire and Rescue Services (FRS) are increasingly establishing a presence as ‘Fire Medical Responders.’  FRS are highly developed organisations with the potential to offer a rapid response to medical emergencies. They also operate with ‘latent capacity,’ compared to health care resources, creating an opportunity to assist in meeting urgent patient need.  The aim of this study was to investigate the impact of FRS co-responding on the delivery of emergency medical response.’

Methods

An observational study generating data from 42/50 Fire & Rescue Services, FRS, during 2016 to assess current involvement in EMS co-response.  We have a) compared response time distributions between Fire and Ambulance services b) described the types of calls attended by FRS crews c) estimated the likely survival benefit based on optimal response time curves for successful defibrillation in out of hospital cardiac arrest and d) conducted an economic evaluation.

Results

There was a statistically significant difference in response performance between the Fire and NHS Ambulance Services, with FRS arriving first in 62% of cases, NHS Ambulance Service arriving first in 23% and no record of who arrived first in 25%.   For every 10% increase in the proportion of ‘whole time duty fire stations’ [stations with 24 hrs/day staffing], there was an 8.4% improvement in response time and mean 84 second shorter response time compared to stations using retained staff utilised via an “on call” system. The top 5 clinical categories attended by FRS were: cardiac problems/chest pain (23%); breathing problems (13%); unconsciousness (13%); cardio-respiratory arrest (9%), and fitting (9%).  From the response time improvement data, we have estimated a potential survival benefit of 1.2 Quality Adjusted Life Years (QALYs) gained.  Using NICE figures of £20,000 per QALY, there is a potential benefit of around £23,000 per critical medical event but this figure should be treated with caution as it is a theoretical extrapolation and the study was not designed to measure individual patient outcome.

Conclusions

Fire Medical Responding is a new development in the UK, having previously operated a strict demarcation between fire and ambulance services although it is well-established in some European countries and in other parts of the world, such as the USA.  This study found FRS can frequently respond more rapidly to medical emergencies than the ambulance service and that they can be appropriately deployed to time critical conditions. This offers an opportunity to employ an underutilised, potentially life-saving resource more widely at low cost.  There is a potential life-saving advantage in further developing and evaluating a Fire Medical Response capability.

IMMIGRATION PROBLEM IN GREECE,

The impact for Emergency Medical System in Athens

INTRODUCTION

Over 10000 refugees have lost their lives in the Mediterranean since 2016 in their endeavor to reach the EU

In the first half of 2016 there were 2809 recorded deaths

Immigration today is one of the most important problems in the world and at the same time a purely anthropocentric challenge for all stakeholders, especially the EMS in GREECE (National Centre for Emergency care) (First responder)

SCOPE-METHOD

    In 2015 Greece became the main point of entry into the EU for refugees and immigrants from Turkey. It is estimated that 850,000 people attempted the dangerous passage of the Aegean Sea

Of these, more than 53,000 refugees remained in Greece Most of them (about 90%) come from Syria, Iraq, and Afghanistan. Among them are small children, people with severe health problems, pregnant women, and infants

Disease-related diseases (refugees) are often unexpectedly severe and complex (extreme age groups - infants, children, and the elderly)

To describe & estimate the effect on EMS/NHS

CHARACTERISTICS:

Children make up 48% of the refugees, while for adults, 30% of them are men and 22% are women

• 10% of refugees in Greece are only 2 to 4 years old
• 14% are aged between 5 and 9 years
• 11% are aged 10 and 14 years old

There are camps that sheltered all these people in ATTICA 

 DATA for the present study are from the central Department of EMS & Na.H.O.C. archives

In the year 2016 they took place: more than 5000 records of emergency transportations from the above camps to Hospital through the Na.H.O.C. & EMS  

CONCLUSIONS:

• The effect of migration problem in our country runs through every activity
• EMS/NaHOC is responsible to manage the medical problems of the immigrants
• This require resources from our country in a very difficult period to be available  
• Although our country overcomes the present problems & reacts in the best practice 

Background: Intoxications with prescription drugs are a common burden at emergency departments (EDs). Mortality associated with intoxication has been increasing. Intoxications are a common way to commit suicide, especially among women. Our study aimed to evaluate intoxication patients` psychiatric history and other clinical features.

Material and methods: We identified all patients with ICD10- code TX36 from year 2014 at Tampere University ED.  We collected the data on age, gender, arrival time and date from hospital records. We also collected patient-specific data such as psychiatric diagnoses, previous psychiatric care and suicide attempts, alcohol and/or drug consumption, difficulties in life (with relationships, money, work or with own or relatives` health) and somatic symptoms.

Results: There were a total of 372 patients with a slight female predominance (51,6 %).The median age was 38 years (1-92 years).  40% of cases arrived to ED between 6 p.m-12 p.m. The most used prescription drug was benzodiazepam (34%). 13% of patients had also used some type of illegal drug, for example cocaine, LSD and cannabis. Activated charcoal was given to 71 % of the patients. According to our data, 53% of intoxications were intentional/suicidal and in 18% of cases the feature of self-harm was not registered. 10% of patients had had one previous intoxication in the database of Tampere University Hospital during the previous two years, and 2% had had more than one intoxication during the same time. The most common difficulties in life were associated with interpersonal relationships (43%). 54% of the patients received psychiatric consultation and 66% were guided to psychiatric after-care.  The most common somatic complication of the intoxication was respiratory tract infection (7 %). Seven-day mortality was 0.8 % and one-year mortality 4.8 %.

Discussion: The results of our study were convergent with previous intoxication studies. Gastrointestinal decontamination was executed rarely but the number of complications, however, was low. This can be partly explained by effective and well-functioning treatment chains and settings. Understanding the associated features of intoxication patients is important for care guidance to these patients. As we could see in this study, many of the patients attempted suicide but only a minority of them were in danger of death.  These cases should be recognized as a cry for help. It is important for health care workers to identify high-risk patients and to guide them to psychiatric care as soon as possible, in order to prevent recurrent intoxications. This study shows that if the intoxication patient reaches the hospital, the prognosis is good. The mortality rate in this patient group is low.

Background: Alcohol is one of the most frequently abused drugs. Alcohol exposure of pediatric population is gradually increasing all over the world thus leading to acute alcohol intoxication and its consequences. 

Objective: The aim of this study was to describe presentations and analyze demographic, clinical and laboratory characteristics of pediatric patients presented to the pediatric emergency department with acute ethanol intoxication.

Methods: We conducted a retrospective review of pediatric patients, who presented to a pediatric emergency department with any complaint and had serum ethanol level determined between January 2006 and December 2016. Patients with serum ethanol level below 50 mg/dL, patients with insufficient data and patients older than 18 year-old were excluded from analyses.

Results: Serum ethanol levels were determined for 917 patients. Among these, 229 patients were tested positive for alcohol abuse having serum ethanol levels >50 mg/dL. Nine patients were excluded because of having insufficient data so a total of 220 patients (Male 128; female 92) were included in the study. 53% patients were brought to the emergency department by emergency medical services. Mean age was 16.0±1.6 years. Most frequent complaints at presentation were decreased level of consciousness (29.5%, n=65), nausea/vomiting (21.8%, n=48) and trauma (14.1%, n=31). The median Glasgow Coma Score on admission to the emergency department was 15.  Only 5 patients had GCS ≤8. Minor injuries were identified in vast majority of patients with trauma. Most common injury type was falls (5.5%, n=12). 11.8% (n=26) patients consumed alcohol as part of a suicidal attempt. Serum ethanol level ranged between 50.8-341.2 mg/dl (mean: 157.9±57.9 mg/dl).  63% (n=140) patients had blood gas analysis. Among these 68.6% (n=96) had hyperlactinemia. 207 patients had biochemical investigations, which revealed abnormal kidney functions in 20.8% (n=43). Likewise, 19% (n=39) had hypokalemia (<3.4 mEq/L) while 17.6% (n=36) had hypophosphatemia (<2.7 mEq/L). None of the patients had hypoglycemia. However, 51.9% (n=95) had mild hyperglycemia (100-200 mg/dL). Blood glucose level and pH were correlated with serum ethanol levels (p=0.007, R²=0.053 and p=0.008, R²= 0.038, respectively). Vast majority of the patients (94%) received treatment in the pediatric emergency department.

Discussion: Acute alcohol intoxication in pediatric population is a preventable emerging problem. It is important to recognize that hyperlactatemia, hypokalemia, hypophosphatemia, mild hyperglycemia and abnormal kidney functions are common biochemical findings in children with acute ethanol intoxication. 

Background: Minor head trauma is a major cause of emergency department visits. Head computed tomography (CT) is the reference standard for the emergency assessment of head trauma. A recent multicentre study of Research in European Pediatric Emergency Medicine (REPEM) network demonstrated that 30% of patients with a minor head trauma (MHT) underwent skull radiography (SR).

Objective: Describe the practice variation in the use of SR for MHT in a group of hospitals affiliated to REPEM.

Design/Methods: Subanalysis of a multicenter retrospective study, including 15 hospitals from 9 European countries. Patients up to 18 years with MHT, defined by Glasgow Coma Scale score (GCS) 14-15, evaluated in years 2012, 2013 and 2014 were included. Pediatric Emergency Care Associated Research Network (PECARN) rules were considered the standard to assess differences in management and to stratify the risk for clinically important Traumatic Brain Injury (ciTBI).

Results: In the main study 13.266 patients (GCS 13-15) were included and 10.109 (76.8%) patients had MHT. The prevalence of ciTBI was 79/10109 (0.77%). SR was performed in 2.762 (27.3%) patients. The rate of SR varied between centres from 0.42% to 92% (figure 1).
Fifty-four (1.96%) had a skull fracture in the SR. In 27 (50%) a head CT confirmed a skull fracture or documented an intracranial lesion. Thirteen (48.1%) patients with skull fracture had other intracranial findings while 14 (51.9%) children had an isolated displaced skull fractures.
Three (11.1%) patients required surgery, due to intracranial lesions.
In 27 (50%) patients, the head CT did not demonstrate any fracture or intracranial lesion.
Twenty-one (77.8%) true positive patients would be classified as intermediate or high risk for ciTBI according PECARN criteria.
Focusing on children determined as low risk for ciTBI according to PECARN rules, SR was performed in 1.933 (28.9%) patients, and demonstrated a fracture in 12 (0.62%) patients. Three (0.16%) patients had an associated intracranial lesion.
Factors associated with the use of SR were grouped as intermediate or high risk for ciTBI according to PECARN rules and isolated scalp hematoma.

Conclusion(s): Although the low diagnostic value does not justify its use, SR is frequently ordered in management of MHT patients in a representative group of pediatric emergency department of the REPEM. We demonstrated a wide variation in the use of SR. These differences are mainly to be due to local or national guidelines and consolidated practices more than lack of adherence to validated prediction rules.

Background

Cervical spine (C-spine) injuries in pediatric trauma are rare (0-2%)¹, but can cause long-term morbidity. Recent Advanced Paediatric Life Support (APLS) guidance¹ updated advice on management of suspected C-spine injury, not advocating routine use of hard collars. There are minor differences within existing national guidance on the same subject [National Institute for Health and Care Excellence (NICE)², Joint Royal Colleges Ambulance Liaison Committee (JRCALC)³, Advanced Trauma Life Support (ATLS)⁴]. This study was undertaken to formally understand observed variations in C-spine stabilisation for pediatric trauma in the East of England (EoE) Trauma Network. 

Methods

An online questionnaire was sent to members of the EoE Trauma Network.  Respondents were presented with five hypothetical scenarios, reflecting changes in guidance, summarised as follows; 1) 4 year old (yr), high-speed motor vehicle collision, intubated, 2) 9 yr, 2 metre fall, GCS 14, chest pain, 3) 7 yr, bicycle collision, C-spine tenderness, for transfer to CT, 4) 6 yr, 4 step fall, GCS 13, combative, 5) 3 yr, rollover motor vehicle collision, asymptomatic.

Respondents chose whether C-spine protection was required followed by the type of protection required [‘Manual in-line stabilisation’ (MIS), ‘Collar’ or ‘Blocks and tape’ (B&T)] and ‘No protection required’. 

Results

A total of 163 responses were received from October 2016 to February 2017, mainly from paramedics (64%, 105/163) and 13% (21/163) from doctors.

1. 77% (124/162) were unaware of recent changes in APLS guidance.
2. The majority thought stabilisation was required in scenarios 1 to 4 [97% (161/166), 77% (130/168), 96% (158/164) and 82% (136/166) respectively]. However, opinion was divided in scenario 5 with 50% (81/162) choosing to stabilise, and 50% (81/162) otherwise. 
3. The chosen method of stabilisation also varied, notably in scenario 2 [MIS; 44% (57/130), collar; 20% (26/130), B&T; 36% (47/130)]. Few chose collar for a combative child [scenario 4; 8% (11/136)].
4. Of 666 total responses, collars were chosen least often [18% (118/666)], and MIS and B&T were selected equally [41% (274/666)].

Conclusions

The results suggest varying practice in C-spine stabilisation in the region, possibly reflecting variation in national guidelines. Encouragingly, appropriate protection was used majority of the time. Despite the limitations of this survey, it provides preliminary evidence of inconsistent practice, and hence requirement for clearer guidance and education.  More studies are needed to validate these findings, and ascertain whether they are representative of a national issue.

Introduction 

Vital signs are commonly measured during the first clinical assessment at the emergency department (ED). Usually, healthcare workers judge these physiologic measurements based on existing reference ranges and values below or above the pre-specified cut-offs are interpreted as abnormal. Many different vital signs reference ranges exist for use in children, but their diagnostic value is uncertain. Therefore, the aim of this study is to determine the diagnostic value of commonly used heart rate and respiratory rate reference ranges for the recognition of serious illness in children at the ED.

Methods

We assessed commonly used paediatric reference ranges for heart rate and respiratory rate, including those from guidelines, textbooks, medical literature and those provided in triage systems or early warning scores. The analysis is based on a observational cohort of children under 16 years of age, presenting to the ED of a university hospital in the Netherlands (2009-2012). Nurses routinely recorded patient data, vital signs and patient destination in the electronic health record. Missing vital signs were imputed 10 times using a multiple imputation approach. In a descriptive analysis we explored differences between the age-classification and cut-off values of the different reference ranges. Moreover, we assessed the diagnostic accuracy of these reference ranges for serious illness in children, defined as the need for ICU admission or hospital admission immediately after the ED visit

Results 

In our cohort, 15,099 children attended the ED during the study period, of whom 314 (2.1%) were admitted to ICU and 2681 (17.8%) to hospital. We identified 11 commonly used paediatric reference ranges for heart rate, respiratory rate, or both. These showed a large variation in age classification and corresponding cut-off values. Application of the different reference ranges in our cohort classified 2.4% to 58.0% of heart rate and 1.0% to 61.4% of respiratory rate values as abnormal. None of the individual vital signs had both a high sensitivity and a high specificity to detect serious illness in children, but the trade-off was very different for each of the reference ranges. Abnormal heart rate had a sensitivity ranging from 0.13 to 0.76 and a specificity ranging from 0.42-0.98 for ICU admission. For hospital admission, sensitivity ranged from 0.06 to 0.72 and specificity from 0.45 to 0.98. The diagnostic accuracy of respiratory rate also had a wide range, with sensitivity 0.05 to 0.69 and specificity 0.39 to 0.99 for ICU admission and sensitivity 0.02 to 0.68 and specificity 0.39 to 0.99 for hospital admission.

Conclusion

Several vital sign reference ranges for children exist and differences are large. It is important to be aware whether a certain reference range is better at ruling-in or ruling-out serious illness. Future research should aim at optimizing the cut-off of individual vital signs to improve existing reference ranges for children at the ED.

Introduction

Blood pressure measurement is recommended in children at the emergency department (ED) because low blood pressure is considered a marker of serious illness. However, blood pressure measurement is time consuming and a burden for (young) children. Moreover, different reference values are available and little evidence exists about the diagnostic value of low blood pressure in children. This study aims to identify lower reference values for systolic blood pressure and to investigate the diagnostic value of routine blood pressure in addition to heart rate in children at the ED.

Methods

A systematic review was performed to define age-specific cut-off points for low blood pressure. Secondly, we used blood pressure cut-offs from two well-known international guidelines (APLS and PEWS) in a prospective cohort of children attending a university ED (2009-2013) in the Netherlands. To investigate the diagnostic value for these two blood pressure cut-offs, we performed multivariable logistic regression to assess the association of abnormal blood pressure with serious illness, adjusted for abnormal heart rate. Sensitivity and specificity for serious illness (hospital or ICU admission) according to abnormal blood pressure defined by the APLS and PEWS were calculated. To assess the additional value for heart rate, sensitivity and specificity were computed for tachycardia and for patients who had both tachycardia and abnormal blood pressure according to the two cut-offs.

Results

18 articles and 11 guidelines reported reference ranges. Only one guideline cited literature references. There was a large variation between the different age-related cut-offs for hypotension (differences ranging from 15 to 30 mmHg in age groups). In the observational study, 5467 children had complete data of blood pressure and heart rate. Frequency of ICU- and hospital admission was 5.5% and 34.7%, respectively. Abnormal blood pressure was significantly associated with hospital admission when adjusted for heart rate based on APLS (OR 1.32 95%CI 1.17–1.48) or based on PEWS (OR 1.76 95%CI 1.57–1.98). Similar associations were found with ICU admission. Abnormal blood pressure according to the APLS showed moderate sensitivity (67%; 61%) and low specificity (43%;45%) for ICU- and hospital admission. The PEWS demonstrated low sensitivity (55%;44%) and moderate specificity (64%; 68%). Tachycardia had low sensitivity (37%;30%) and high specificity (81%;85%) for ICU- and hospital admission. When combining tachycardia and abnormal blood pressure, the APLS showed high specificity (88%; 90%) and low sensitivity (21%;14%). The PEWS showed similar results.

Conclusion

Clinical references for blood pressure show large differences and are mostly not evidence based. Abnormal blood pressure showed an association with serious illness at the ED, but its diagnostic value is uncertain. However, the combination of tachycardia and abnormal blood pressure appears to be good at ruling-in serious illness. 

Background: neonates with fever without source (FWS) present a higher prevalence of invasive bacterial infection (IBI) than older infants. For this reason, it has been universally recommended performing a lumbar puncture and the admission with antibiotic treatment for any febrile neonate, even for those who are well-appearing. The “Step-by-Step” approach uses the 21-days-old cut-off point to identify high-risk patients. Our objective was to analyze the performance of the procalcitonin (PCT), C-reactive protein (CRP) and absolute neutrophil count (ANC) to identify IBIs among well-appearing neonates ≤21 days old with FWS.

Methods: a prospective registry-based cohort study including all the infants ≤90 days old attended in the Pediatric Emergency Department of a tertiary teaching hospital between September 2008 and August 2016 with FWS. We compared the prevalence of IBI (isolation of a pathogen bacterium in blood or cerebrospinal fluid) between those well-appearing patients ≤21 days old and >21 days old without leukocyturia in two groups: those with altered blood tests (PCT ≥0.5 ng/mL, CRP >20 mg/L or ANC >10000/mcL) and those with normal blood tests. We excluded those patients in whom the value of any of the three blood tests, the urine dipstick result or the blood culture result was not available.

Results: we included 1,762 of the 1,970 infants ≤90 days old with FWS attended (89.4%). Of them, 1,358 (77.0%) infants were well-appearing and had no leukocyturia in the urine dipstick. PCT, CRP and ANC values were normal in 126 of the 178 infants ≤21 days old (76.7%) and in 956 of the 1,180 infants > 21 days old (81.0%).

Prevalence of IBI in infants ≤21 days old was 3.2% among those with normal blood tests (vs 0.1% in infants >21 days old; OR 31.31 [IC 95%: 3.28-741.52]) and 5.7% among those with any of the three blood tests altered (vs 4.9% in infants >21 days old; OR 1.19 [IC 95%: 0.25-4.83]). Two of the four well-appearing infants ≤21 days old with normal blood tests who had an IBI were diagnosed with a bacterial meningitis.

Sensitivity and specificity of the three blood tests for identifying IBIs were 42.9% (15.8-75.0%) and 71.3% (64.2-77.6%), respectively in infants ≤21 days old and 91.7% (64.6-98.5%) and 81.8% (79.4-83.9%), respectively in infants >21 days old.

Discussion: PCT, CRP and ANC do not have a good performance to identify febrile infants less than 21 days old at low risk for IBI. In contrast to older infants, these tests cannot be used to identify patients suitable for a less aggressive management. Accordingly, neonates under 21 days old with FWS must be admitted with empiric antibiotic treatment after performing a lumbar puncture, regardless the general appearance and the results of the blood tests.

Introduction

Fever is the main presenting symptom of children presenting at paediatric emergency departments (EDs) in Europe, with a majority related to respiratory tract infections (RTI’s). Despite a low rate of bacterial infections (5 – 10% of febrile children), we observe antibiotic prescription rates of 40-56% in children with RTI’s, with high variability among European EDs. This study aims to evaluate the association between clinical characteristics and antibiotic prescription rates in children under five with suspected lower respiratory tract infections at 6 Dutch EDs.

Methods

Prospective collected data of a multicentre study in 6 paediatric EDs in The Netherlands, both teaching and non-teaching. The population consisted of children aged 1 month to 5 years presenting at the ED with fever and cough or dyspnoea. We computed a risk profile for bacterial infection based on clinical characteristics, using a clinical prediction rule (Feverkidstool). Variation in risk profile and antibiotic prescription rate were assessed and associations tested.

Results

Results are based on 206 patients, 63% male, median age 16 months (IQR 7 – 32m). Median predicted risk of a bacterial infection according to the Feverkidstool was 9% (IQR 9 – 17%), ranging between centres from 5 – 15%. Overall antibiotic prescription rate was 42% (range 24 – 60% between centres). Children with a higher risk profile had a significantly higher prescription rate (Nagelkerke’s R²=23%). There was no association between prescription rate and centre. When stratified by clinical profile, antibiotic prescription rate was 13% in low-risk patients (predicted risk 0-5%), 34% in medium-risk patients (predicted risk 5-10%) and 65% in high-risk patients (predicted risk >10%).

Discussion

Antibiotic prescription rates are high among children with respiratory tract infections with variable rates among 6 Dutch EDs. Variability among centres is mainly explained by risk profile. Given the nature of lower respiratory tract infections, watchful waiting and follow-up in low and medium-risk patients could add to a reduction in antibiotic prescriptions.

Introduction: Increased Intracranial Pressure (IIP) is a highly clinical mortality condition, which can be caused by various causes. It should be diagnosed rapidly, and its treatment should be done timely and correctly in emergency units. The procedures performed for the purpose of diagnosing and determination of etiology in patients with IIP are either invasive or cause radiation exposure. In this study, we aimed to determine the benefit of measurement of the optic nerve sheath diameter (ONSD) by ultrasonography (US) and power of the test in the evaluation of IIP.

Materials and Methods: After the primary treatment of the patient who was brought to our pediatric emergency department, transorbital ultrasonography was applied in the supine and neutral position as his/her eyes closed. Sonographic ONSD evaluation was performed using a SonoSite Edge ultrasound device with 6–13 MHz linear probe. The diameter of the optic nerve, which appeared as a hypoechoic two-sided line at a depth of 3 mm of globes which is determined as more sensitive to IIP alteration, was measured and recorded in both longitudinal and transverse sections.

Findings: Fifty-seven cases with IIP suspicion brought to our unit (31 males; 138 ±56 months old) and 35 controls (17 males; 151± 45 moths old)  were included in the study between June 2015 and December 2016. Thirty-one cases (54%) were trauma cases with the high probability of clinical signs of IIP. Others had headache, vomiting, altered consciousness, seizures. Eight of our patients had GCS <= 8. One patient died and 16 children were admitted to our intensive care unit. 19 of our patients were treated with anti-edema treatment. The ONSD value of the 38 patients without brain edema on CT scan was 4.8 ± 0.05 mm (Processing time: 2.8 ± 1 min). The ONSD of those with brain edema was 5.5 ± 0.07 mm (Processing time: 2.0 ± 1 min). The mean ONSD of all patients (5.0 ± 0,07 mm) showed significantly increased compared with the controls (3,9 ± 0,02 mm) (p<0.01). The ideal cut-off value of ONSD was found to be 4.9 mm when the cerebral edema detected in the CT scan was accepted as a reference (Sensitivity 84.2% and specificity 63.2%). Six patients had optic disc elevation (The median ONSD was 6 mm). The CT scan of all of these patients was compatible with brain edema.

Conclusions: CT examination and fundoscopy for diagnosing IIP are useful methods for middle/late stages of the IIP syndrome. As ONSD begins to expand within minutes when intracranial pressure begins to increase, ONSD measurement may be more sensitive in the acute stage and guide patient management in case of clinical suspicion of IIP.

Background and aims: Italian pediatric antimicrobial prescription rates are among the highest in Europe. It is essential to identify efficient measures to improve antimicrobial stewardship (AS) programs. Since Clinical Pathways (CPs) have proven a promising tool to reduce antibiotic prescriptions in primary care and in-hospital settings, we hypothesized that their implementation in the Padua University Hospital Pediatric Emergency Department (PED) would decrease overall prescription of antibiotics, especially broad-spectrum (BS), for common infectious diseases such as Community-acquired pneumonia (CAP). 

Materials and methods: CP was implemented at the Department for Woman and Child Health of Padua on 01/10/2015. This is a pre-post quasi-experimental study comparing the 6-month period prior to CP implementation (baseline period: 15/10/2014-15/04/2015) and during the 6 months after intervention (post intervention: 15/10/2015-15/04/2016). We collected data from children aged 3 months -15 years diagnosed with CAP. We assessed differences in various measures of antibiotic prescription between pre and post periods including rates, breadth of spectrum, duration of therapy and, for inpatients, length of hospital stay. Chi-square, Fisher’s exact test and Wilcoxon rank sum test were used as appropriate. 

Results: 120 pre and 86 post-intervention clinic visits were associated with CAP. In regards to outpatients, we observed a decrease of BS regimens (50% vs. 26.8%, p=0.0215), in particular macrolides, and an increase of narrow-spectrum ones (amoxicillin). Children received less antibiotics (median DOT from 10 to 8, p=0.0001) for fewer days (median LOT from 10 to 8, p=0.0001). Physicians prescribed a narrow- spectrum monotherapy more frequently than BS combination therapy (DOT/LOT ratio 1.157 vs. 1.065). No difference in treatment failure incidence was reported before and after the implementation (2.3% vs. 11.8%, p=0.2862). Among inpatients we also noted a decrease in BS regimens (100% vs. 66.7%, p=0.0238) and the introduction of narrow-spectrum regimens (0% vs. 33.3%, p=0.0238). Admitted patients received less antibiotics (median DOT from 18.5 to 10, p=0.004), while there was no statistical difference in LOT (median LOT from 11 to 10, p=0.0629). In particular, children received a notably lower amount of BS days of therapy (median bsDOT from 17 to 4.5, p <0.001). No difference in treatment failure was reported before and after CP implementation (16.7% vs. 15.4%, p >0.999).

Discussion: Our study showed sustained changes in physicians' prescribing behaviors for CAP after implementation of a clinical pathway. Prescribing changes for CAP included an immediate increase in amoxicillin prescriptions with a concomitant reduction of BS antibiotic prescriptions, use of combination therapy and duration of treatment for CAP indicates effectiveness of CP for AS in this setting. 

Computed tomography is the golden standard for evaluating haemodynamically stable blunt trauma patients. As a consequence, medical radiation induced cancers have been increasing exponentially. Ultrasound evaluation has been actively investigated as an alternative, but currently an ultrasound based algorithm for the investigation of blunt trauma patients cannot be supported. In this study instead of correlating ultrasound imaging findings to computed tomography imaging findings we correlate ultrasound imaging findings with patient clinical outcomes. This prospective clinical study took place in Nikaia general hospital Greece, between 6/2014 and 12/2014. We studied 60 (sixty) consecutive, haemodynamically stable, adult, blunt trauma patients, without injuries requiring immediate surgical intervention or hospitalisation, who met abdominal CT investigation criteria. All patients were initially investigated and treated according to current ATLS recommendations. An abdominal ultrasound evaluation was performed in all patients prior to the CT scanning. The ultrasound examination was performed by an ultrasound expert radiologist. We assessed the presence of free fluid as well as all solid abdominal organs for evidence of injury. Patients then underwent a formal trauma protocol abdominal CT. All patients were admitted to the surgical ward for a minimum of 48-hour observation, where they were closely monitored and investigated.  If there was no deterioration patients were discharged. In any other case patients were treated accordingly. In 21/60 patients, the ultrasound was negative for fluid and solid organ injury. A further 21/60 despite the presence of free fluid no solid organ injury was detected. Of these patients 7/42 were found to have some degree of solid organ injury at CT.  All 42 patients (100%) had an uneventful clinical course and were safely discharged. In 12/60 patients, ultrasound was positive for fluid and solid organ injury, all 12 patients had some degree of solid organ injury in the subsequent CT scan. Of these patients 25% were eventually treated surgically. In 6/60 patients, ultrasound was negative for free fluid but positive for solid organ injury. 5/6 of these patients had some degree of solid organ injury on CT. From this patient group 1/6 needed surgical intervention. This study provides evidence that abdominal ultrasonographic evaluation of trauma patients, when performed by an experienced professional can safely be used as a screening tool, since negative ultrasound findings correlate with 100% non-surgical clinical outcome and positive ultrasound findings correlate with a 100% positive CT findings and a 25% need for surgical intervention. Taking these results a step forward, it could be implied that blunt trauma patients can be safely discharged and followed up based on an ultrasound based algorithm.

Introduction

Chest X ray (CXR) is the most frequent imaging exam in the Emergency Department (ED). However, its interpretation is frequently difficult; thus, CXR based clinical decisions might be harmful. We investigate discordance (DI) between EP and RA on CXR interpretation prescribed for a non-traumatic indication.

Patients and methods

inclusion criteria

patients older than 15 years old

non traumatic indication

exclusion

pregnancy,

method

This observational study was conducted in the ED of a teaching hospital with an annual census of 75000 patients. During a 3-month period, EP and ED residents completed a reporting form for a convenience sample of patients with CXR. It included clinical features, a CXR systematic analysis and a conclusion. A similar form without clinical features was completed by RA being blind to the EP interpretation.

The main objective was DI between EP and RA. Secondary objectives were performance of EP interpretation, therapeutic and orientation impact induced by all CXR, therapeutic errors induced by a wrong interpretation and potentially serious illnesses missed by EP. RA interpretation was considered as the gold standard.

Statistics

For a DI 0.1, with alpha risk 0.05 and beta 0.1, the required number of patients was 196. Qualitative data were expressed as percentage and 95% confidence intervals. They were compared by a Chi2 test, p < 0.05 being significant.

Results

417 patients were included, 15 were excluded for lack of clinical informations. Age was 61 + 15 years old, sex ratio 1.09. CXR quality was assessed as good for 266 for EP and 136 for RA (p < 0.0001). DI was 23% [19-27%]. Sensitivity, specificity, negative predictive value, positive predictive value were 87% [82-86%], 64% [54-71%], 80% [72-85%] and 75% [69-80%] respectively. Therapeutic and orientation impacts of all CXR were 44% [39-49%] and 19% [15-23%], respectively. Due to wrong EP interpretation, 42 antibiotics and 31 diuretics were mistakenly prescribed. Eight suspect opacities were missed.

Discussion

DI in our ED was comparable to other published studies. The main limit was a selection bias since only 20% of all prescribed CXR during the study period were included. Despite a rather bad quality and a DI affecting 23% of CXR, therapeutic impact of CXR remains high. Training of EP and ED residents to CXR interpretation has to be organized. However, intrinsic diagnosis qualities of this exam remain poor even with good realization and interpretation.

Implementation of lung Point-of-Care Ultrasound which has demonstrated far better performance might be interesting. This diagnosis procedure is inexpensive, radiation free, performed on the patient’s bedside and with immediate result.

Introduction

Transthoracic echocardiographic examination (TTE) that is performed at the patient’s bedside in emergency departments has several recognized important indications.

Objective

 The purpose of our study is to evaluate the agreement of the estimates of left ventricular ejection fraction (LVEF) obtained by emergency physicians with the findings obtained by cardiologists in patients admitted to emergency departments.

Material and methods

This randomized prospective study was carried out in the emergency department of the military hospital of Tunis (Tunisia) over a period of 6 months going from September 2015 through February 2016, and involving patients aged > 16 years whose condition required an emergency TTE.

The patients included in the study had to undergo a double echocardiographic examination:

1-An initial investigation that was performed in the emergency department by an emergency physician who had previously received a three-month training in Doppler echocardiography.

2-A subsequent echocardiographic examination that was performed by an echo-Doppler proficient cardiologist.

Left ventricular ejection fraction was evaluated by both readers using the following methods:

1-the global visual estimation (GVE) method,

 2-Teicholtz’s method in time movement mode (TM)

3-and Simpson Biplan method (SB).

We excluded from the study patients with:

1-segmental kinetic disorders

2-or with hearts out of alignment.

The findings thus obtained were compared using the inter-class concordance coefficient of Cronbach’s alpha.

Results 

Fifty-two patients were involved in the study. Mean age was 55 + 11 years; sex-ratio was 7 males/4 females.

-For the GVE method, the findings obtained by the emergency physician were similar to those obtained by the cardiologist: alpha = 0.72 (IC 95% = [0.68-0.78]; p<10-3).

-The findings obtained by both operators by Teicholtz’s method were as follows: alpha = 0.94 (IC 95% = [0.80-0.95]; p<10-3).

-The concordance of the findings obtained by the emergency physician and of those obtained by the cardiologist for their assessment of LVEF by SB method was shown by alpha=0.91 (IC95% = [0.80 – 0.98]; p<10-3).

Conclusion

Global visual estimation of LVEF can be performed similarly by an emergency physician or by a cardiologist provided they are sufficiently experienced. The results yielded by both other methods (Teicholtz’s method and SB method) were very similar indicating an excellent concordance independently of the degree of deterioration of the left ventricle contractility. Biplan Simpson’s method is, however, a time-consuming procedure.

Objective: Most studies of children with isolated skull fractures have been relatively small, and rare adverse outcomes may have been missed. Our aim was to evaluate the short-term clinical outcomes of children with isolated skull fractures.

Methods: We performed a systematic review and meta-analysis of studies indexed in EMBASE, MEDLINE and Cochrane Library databases through August 2016 reporting on short-term outcomes of children ≤18 years with linear, non-displaced, isolated skull fractures (i.e. without intracranial injury on neuroimaging). Two reviewers independently reviewed identified articles for inclusion, assessed quality and extracted relevant data. Our primary outcome was emergent neurosurgery or death. Secondary outcomes were hospitalization and new intracranial hemorrhage on repeat neuroimaging.  We calculated a pooled estimate of each outcome by fitting a random-effects model and then tested for heterogeneity across studies.

Results: Of the 385 studies screened, the 21 that met our inclusion criteria, included 6646 children with isolated skull fractures. One child needed emergent neurosurgery and no children died [pooled estimate: 0.0%, 95% confidence interval [0.0-0.0%]; I² =0%]. Of the 6280 children with known emergency department disposition, 4914 (87%, 95% CI 78-95%; I² = 98%) were hospitalized.  Of the 644[SB1]  children that underwent repeat neuroimaging, six had a non-operative intracranial hemorrhage (0.0%, 95% CI 0.0-0.1%; I² = 79%).

Conclusion:Children with isolated skull fractures were at extremely low risk for emergent neurosurgery or death, but were frequently hospitalized. After careful consideration of non-accidental trauma, clinically stable children with an isolated skull fracture could safely be managed outpatient.

Background and Objectives

Italian pediatric antimicrobial prescription rates are among the highest in Europe. To date it has not been paid adequate attention on how to implement and improve the antimicrobial prescriptions. As a first step for antimicrobial stewardship (AS) implementation, clinical pathways (CP) outlining standard of care for acute otitis media (AOM), and group A streptococcus (GAS) pharyngitis were developed and implemented on 1 October 2015 at the Pediatric Emergency Department in collaboration with Children’s Hospital of Philadelphia.

The primary aim of this study was to assess changes in antibiotic prescription before and after CP implementation for AOM e GAS pharyngitis; secondary aims were to compare treatment failure and to assess the change in the total antibiotics costs before and after CP implementation.

Methods

Pre-post quasi-experimental study comparing the 6-month period prior to CP implementation (baseline period: 15/10/2014-15/04/2015) and during the 6 months after intervention (post intervention: 15/10/2015-15/04/2016).

We assessed differences in various measures of antibiotic prescription appropriateness, including type and breadth of spectrum prescribed, using chi-square and t-tests as appropriate. We also assessed the total cost and the cost for each class of antibiotics comparing the two groups and relating it to 1000PD.

Results

295 pre- and 278 post-intervention clinic visits were associated with AOM. After CP implementation there was an increase in “wait and see” (21.7%vs.33.1%,p<0.01) and a decrease from 53.2% to 32.4% (p<0.01) in overall prescription of broad-spectrum (BS) antibiotics. The total cost was significantly reduced (8.033,08€/1000PDvs. 5.878,30€/1000PD), with a decrease especially in BS antibiotics, above all cephalosporines, and a slight increase in the cost for amoxicillina. 151 pre- and 166 post-implementation clinic visits were associated with GAS pharyngitis, with decrease in BS prescriptions (46.4%vs.6.6%,p<0.01). The total cost was reduced (9.337,68€/1000PDvs. 6.247,23€/1000PD), with a sharp decline in the cost for BS antibiotics and an increase in the cost for narrow spectrum antibiotic contextually to the increase in its use.

Discussion

Our study showed sustained changes in physicians' prescribing behaviors for AOM and GAS pharyngitis after implementation of a clinical pathway. Prescribing changes for AOM included an immediate increase in “observation with close follow-up” approach and amoxicillin prescriptions with a concomitant decrease in BS antibiotic prescriptions. Complying with the CP, a dramatic increase of amoxicillin prescriptions for GAS pharyngitis was documented with a concomitant decrease in BS antibiotic use. In summary, our data show that clinical pathways for AOM and GAS pharyngitis are associated with reduced rates of antimicrobial prescription and cost for antibiotics purchase with no significant change in treatment failure rates.

Background:

Trauma patients are particularly vulnerable to negative experiences of healthcare. The psychological effects of trauma and restricted movement from cervical spine immobilisation combine to heighten a patient’s fear and anxiety. One factor identified to reduce anxiety amongst spinal immobilised patients is eye contact, however this has been neglected from communication tools used within emergency medicine. One explanation for this is the relative challenge of objectively assessing eye contact between doctor and patient using traditional methods.

New wearable technologies offer a way of addressing this blind spot in assessing doctor-patient communication. We subsequently set out to examine the use of point of view glasses as a method of objectively assessing the frequency and location of eye contact between a spinal immobilised patient and doctors in high fidelity trauma simulation.

Methods:

This study was integrated into an emergency medicine module for clinical medical students. High fidelity trauma simulations requiring cervical spinal immobilisation were recorded using covert point of view glasses and ceiling mounted cameras. The simulation footage was analysed, examining the frequency of paired verbal communication and eye contact at five predefined locations around the patient (the foot of the bed, bellow the waist, above the waist, above the shoulders and at the head of the bed).

Results:

110 communication events and 29 eye contact events were observed during six high fidelity simulations. There was a significant difference in the number verbal communication events and eye contact events below the waist, above the waist and above the shoulders (p=0.0312, 0.0156 and 0.0312 respectively). Verbal communication at the head of the bed achieved the greatest eye contact on 95% of occasions (p=0.500).

Conclusion:

Whilst methods for assessing communication skills have been validated for emergency medicine, they have neglected non-verbal communication that can only be assessed from the patient’s perspective.

Using new point of view technologies this study demonstrates an objective method for the identification of non-verbal doctor-patient communication and highlights the poor attainment of eye contact amongst medical students when managing trauma patients. Although this cannot be extrapolated to clinician’s, greater awareness of body position when communicating with spinal immobilised patients, especially in the absence of an anaesthetist, will help to improve eye contact with patients.

Whilst the priority for trauma patients will always be managing their medical condition it is important to give consideration to the patient’s experience, especially for vulnerable groups such as trauma patients. With this awareness and incorporation into current communication tools we aim to provide further feedback for learners during simulation, improving communication and thereby the patient’s experience of trauma care.

Background: Bronchiolitis is a lower respiratory tract infection affecting principally the small airways. The disease is the most common cause of infant hospitalization during the winter months. High flow nasal cannula therapy is recommended in patients with severe disease. The aim of the study was to determine the parameters associated with high flow nasal cannula therapy failure in children with bronchiolitis in pediatric emergency department.

Methods: The patients were aged between 6 weeks and 24 months presenting to the pediatric emergency department of the Health Sciences University, Tepecik Teaching and Research Hospital with acute bronchiolitis between 01.01.2014 and 31.12.2015 were evaluated retrospectively. Vital signs and clinical findings were determined before interventions such as suctioning, antipyretic medication, oxygen support, and I.V. fluid. We included the patients with bronchiolitis treated with high flow nasal cannula therapy. Patients were divided in two groups: High flow nasal cannula therapy responders and non-responders. High flow nasal cannula therapy failure (non-responders) was defined as the need for escalation to another ventilation support: non-invasive ventilation or invasive mechanical ventilation. 

Results: A total of 84 infants (median age: 5 month; 25-75 percentile: 2-10 month; minimum: 6 weeks – maximum: 19 months; female/male: 25/59) with bronchiolitis were treated with high flow nasal cannula therapy. 23 of them (27.4%) were in non-responders group; 19 of them were intubated and mechanically ventilated. Underlying chronic disease, prior hospitalization due to bronchiolitis, prior admission to the pediatric intensive care unit, significant tachycardia (0-12 months> 160 / min, 12-24 months: 150 / min), physical examination findings of significant dehydration (5% or more), pH <7.30 and high pCO2 level (>45 mm Hg) were found more frequently in non-responders group (p <0.05). In the logistic regression analysis, underlying chronic disease (p: 0.031; OR: 4.677; 95%CI: 1.148-19.062), significant tachycardia (p: 0.015; OR: 5.088; 95%CI: 1.369-18.910), and significant dehydration (p: 0.038; OR: 3.811; 95%CI: 1.079-13.459) were the most significant parameters.

Conclusion: The presence of underlying chronic disease, significant tachycardia, and significant dehydration were the most powerful predictors of high flow nasal cannula therapy failure in children with bronchiolitis.

Background and Objectives: High-flow nasal cannula (HFNC) is a reliable method of respiratory support that has demonstrated large utility in the pediatric population. HFNC may be able to avoid intubations in patients with respiratory distress. There is limited data about its use in the pediatric emergency department (PED). The aim of this study was to evaluate whether the use of HFNC therapy is associated with reduced respiratory distress and a decreased need for intubation in patients presenting to the PED.

Methods: This was a single –center prospective observational study conducted over six months  (October 2016 - March 2017) on children with severe respiratory distress (SRD) who commenced HFNC therapy in our PED. Baseline demographic and clinical data, as well as respiratory variables at baseline and various times after HFNC initiation during 24 h, were recorded. Therapy failure was defined as clinical deterioration in respiratory status after that requiring another form of non-invasive ventilation (nasal positive airway pressure (CPAP) or bilevel positive airway pressure (BiPAP)) or invasive ventilation (intubation) within 24 hours from the time of HFNC initiation. The rate of intubation, admisson to pediatric intensive care unit, therapy failure and predictors of therapy failure were also recorded.

Results: A total of 115 children commenced on HFNC therapy in PED during the study period. The median age was 12 months and 70% patients were male. The most common diagnosis was acute bronchiolitis (n=73, 63.5%) followed by pneumonia (n = 23, 20%) and asthma (n = 18, 15.7%). Seven-teen   patients (14.8%) failed HFNC therapy;10 required secondary invasive mechanical ventilation and 7 BiPAP.  Children who had higher initial respiratory score (RS)  and comorbodity were more likely to fail the HFNC therapy (p=0.001, p=0.039). HFNC significantly reduced the respiratory rate, heart rate and (RS) at 2 hours of admission (p<0.05). These improvements were observed as early as 15 min after the beginning of HFNC for respiratory rate and heart rate.

Conclusion: HFNC has a beneficial effect on clinical signs and respiratory score in PED patients with acute severe respiratory distress. It also significantly reduced the respiratory rate and heart rate at the beginning.

OBJECTIVES: In a group of patients with type I and type II acute respiratory failure (ARF), treated with noninvasive ventilation (NIV), we tested if an early evaluation through a validated scale, using variables easily obtained at the bedside, can identify patients at high risk of adverse outcome.

METHODS: This was a retrospective study including all patients with ARF requiring NIV over a two-year period (January, 2014-July, 2016), admitted in an Emergency Department High-Dependency Observation Unit (ED-HDU). Clinical data were collected at baseline, 1 hour, and 24 hours; HACOR score (previously employed only in patients with hypoxemic respiratory failure) was calculated before NIV and after 1 hour and 24 hours of treatment. For prognostic analysis, the score was evaluated as continuous value and as dichotomized value (≤5 or >5, as suggested in the validation study). The primary outcome was in-hospital mortality, need of ICU admission and NIV weaning in a 48-hour time interval.

RESULTS: The study population includes 348 patients, mean age 77±15 years, 53% male gender. Most frequent admission diagnosis were pneumonia in 59% of patients, congestive heart failure in 34% and sepsis in 20%, which overlapped in some patients. Ninety-eight patients presented a Type I ARF and 250 a Type II ARF. In-hospital mortality was 33% in Type I ARF patients and 19% in Type II patients (p=0.012).  Compared with survivors, Type I non-survivors showed  comparable HACOR score before NIV (7.3±5.5 vs 5.9±2.9, p=NS), but higher score after 1-hour (6.8±6.4 vs 3.4±3.3, p=0.025) and 24-hour (6.2±5.0 vs 3.0±2.6, p=0.022) NIV treatment; moreover, HACOR score reduction during the first hour of NIV treatment was significantly higher in survivors compared with non survivors (-2.7±3.2 vs -0.7±3.7, p=0.016). Analysis for repeated measures showed a significantly more marked score reduction  in survivors compared with non-survivors (p=0.001). Compared with survivors, Type II non-survivors showed  higher HACOR score before NIV (8.5±5.0 vs 6.3±4.2, p=0.005) and after 1-hour (6.7±4.7 vs 3.9±3.3, p<0.001) and 24-hour (3.8±4.3 vs 1.7±2.0, p=0.002) NIV treatment. Analysis for repeated measures showed a significantly more marked score reduction  in survivors compared with non-survivors (p=0.001). Compared with patients with HACOR ≤5, Type II patients with HACOR score >5 showed a significantly higher mortality rate at every evaluation point (before NIV: 68 vs 48%, p=0.026; 1-hour: 55 vs 26%, p<0.001; 24-hour: 18 vs 5%, p=0.009); analysis with dichotomized values did not show significant differences among patients with Type I ARF.     

CONCLUSIONS: among patients with Type II respiratory failure , a Hacor score value >5 was significantly associated with an increased mortality rate; among Type I ARF patients, patients with adverse outcome showed significantly worst score value compared with patients with a good prognosis. 

Objectives

On November 8, 2013, the central Philippines islands were struck by typhoon Haiyan (Yolanda), damaging many local hospitals and disrupting acute and regular healthcare. The Belgian First Aid and Support Team erected a type 1 field hospital and water purification unit in Palo (south of Tacloban) to temporarily replace the damaged local healthcare. This study aims to describe the diagnoses encountered, and treatment provided by the Belgian team. Hypothesis is that besides disaster-related trauma, medical problems emerge soon, emphasising the importance of an appropriate composition of Emergency Medical Teams and their supplies.

Methods

Using a descriptive observational study design, all patients presenting to the Belgian field hospital were retrospectively reviewed. Each patient had an individual paper medical record that included gender, age, triage chief complaint, all secondary complaints, diagnosis, management and possible referral information. These were compared with a control group of patients from the same area and season, but from another year.

Results

All 1267 patients triaged, examined and treated in the field hospital, were included and analysed. Almost 28% suffered from injury, but the most important part of patients rather presented with medical diseases (64%), particularly respiratory (31%), dermatological (11%) and digestive (8%). The remaining 8% presented for follow-up, mental problems or pregnancy related issues. Most (53%) of the patients did not present with direct disaster-related acute pathology. More than 59% showed signs of infection within two weeks after the event. Most frequently used treatments were wound care (47%), pain relief (33%), vitamins and minerals (31%) and antibiotics (29% of all patients). Procedures were needed for 9%, fluid therapy for 8%, and psychological support for 5%. Logistic regression analysis indicates that children younger than 5 are more at risk for infectious diseases (OR 18.8 CI 10.6-33.3), and for injury (OR 10.3 CI 6.3-16.8), and males are twice more prone to be injured than females (OR 2.1 CI 1.6-2.6).

Conclusions

These results reveal that one week after the acute phase of a typhoon, respiratory, dermatological, and digestive problems emerge to the prejudice of trauma. Most patients did not present with direct disaster-related but with less acute diseases. Young children are more at risk for injury and infectious diseases. These facts should be anticipated when composing Emergency Medical Teams and supplies to be sent to the disaster site. 

Purpose: To examine what roles bystanders and volunteers can play during a disaster and to propose an action card for the (medical) incident commander to manage this often unutilised resource of manpower and knowledge.

Relevance: During the terrorist attacks in Brussels on March 22^nd 2016 the advantages as well as the problems with bystanders and volunteers during disasters were encountered. Many lives were saved by bystanders by using simple bleeding control techniques and the application of tourniquets. Among the bystanders there were soldiers patrolling the national airport. Their knowledge of lifesaving first aid (e.g. the use of tourniquets), extrication and triage proved pivotal in the first aid rendered to the victims.

However there were also unsolicited medical volunteers arriving with unauthorised private ambulances who became more a liability than a fruitful resource. Drawn to the scene out of curiosity and desire to help, not only were they unfamiliar with the Incident Command System (ICS), but because of the fact they were not registered in the regular EMS system they didn’t have the appropriate means of communication and started freelancing in deciding what to do on site and where to take victims in unregistered ambulances.  

An updated version of a national medical emergency plan was recently published. Being updated with the terrorist attacks in mind, it still fails to recognize bystanders as first aid providers.

Methods: Review of existing literature and unstructured interviews with medical personnel who volunteered during the aftermath of the Brussels Airport attack and the national EMS director.

Results: It is a certainty that (untrained) bystanders are the first providers of help during disasters. During the Brussels attacks we have seen the benefits of bystanders performing lifesaving bleeding control techniques, extrication of entrapped victims, transportation of casualties away from the blast sites and comforting victims.

All too often bystanders and volunteers at a disaster site remain unused and kept outside a perimeter upon arrival of EMS services and police and a clear guideline on how to integrate bystanders and volunteers was and still is lacking in any existing medical emergency plan. However, coordination of highly motivated bystanders and volunteers by professional EMS personnel had a positive effect on reducing morbidity and mortality in the Brussels attacks.  

Conclusion:  Because regular EMS is initially disrupted, bystanders and volunteers can aid in immediate search and rescue and first aid to victims. Instead of discouraging, their desire to help should be encouraged and planned for in emergency plans. To reduce the risk of untrained bystanders and volunteers disrupting the official organised response an action card to aid onsite incident managers is proposed.

Introduction

When working with infectious diseases of high consequences the simple act of donning/doffing Personal Protective Equipment (PPE) becomes a lifesaving procedure not only for the medical staff but for the thousands of people who depend on them. Previous studies have shown that donning and doffing skills may be insufficient even after training. That raises a question on how to properly train staff in PPE donning and doffing to maintain the long term competence of PPE donning and doffing?

This study assessed if there is a performance difference one month after training between a control group that received instructor training and a study group that had access to a video over the month to compare the traditional tutorial based teaching method in teaching PPE donning and doffing.

Methods

This was a randomized controlled trial pilot study of video versus instructor training. 21 medical students and junior doctors where divided into 2 groups by simple randomization. The control group received training with an instructor. This training involved a demonstration of donning and doffing of PPE. Students were then observed donning and doffing of PPE until satisfactory performance had been achieved. There was no other intervention until the day of the assessment. The video group watched a training video demonstrating proper donning and doffing. The video group was given the video to watch at home as many times as they wanted. The training video was made by the same nurse who conducted the instructor training, using the same equipment. After a 1month period, a doctor performed a blind evaluation of all control and study groups using an adjusted Donning and Doffing PPE Competency Validation Checklist 2014.

Results

19 participants attended assessment session after a month with 9 in a control group and 10 in the study group. The mean donning score was 86,5/100.  Mean score was 84,8 for the instructor group, and 88.0 for the video group.There was no statistically significant difference in the donning score between the video and instructor group (95% confidence interval for the effect: -7.7 to 9.5; p-value: 0.54). The mean doffing score was 76,4/100. The mean score for instructor group was 79.1 and 73,9 for the video group. There was no significant difference in the doffing score between the video and instructor group (95% confidence interval for effect: -7,6 to 18,0; p-value: 0.54).

Conclusion

The study results suggests that donning and doffing competencies are similar for those who were trained with an instructor to those who were trained with the video method.

Objective: To find out if a table-top team training program would positively affect perception towards teamwork, and their ability to recognize the presence and quality of team skills in disaster events.

Background: Disaster training involves coordination and communication between various units, which necessitates involvement of the whole chain of response simultaneously. Due to the strict hierarchy culture in Korea, it is not easy to train healthcare providers in teamwork. Hospitals in Korea performs large-scale disaster exercise which takes a lot of preparation and resources. It is also difficult to assess teamwork and communication during a large-scale disaster exercise. Table-top simulation enables you to look at the whole process and the advantage of reflective, repetitive, and safe learning environment, where effective feedback can be provided. Since table-top simulation allows people to go through the thinking process, this could be a good module to train people improving on teamwork competency.

Methods: The educational intervention consisted of a half-day workshop for selected 24 emergency residents and 24 emergency nurses. Participants were given lectures on incident command system and surge capacity. They were randomly mixed into 6 groups (8 participants per group, 4 residents and 4 nurses). Participants were assessed as a group before and after the intervention, which consisted of debriefing session, focusing on the 5 components of teamwork (Team Structure, Communication, Leadership, Situation Monitoring, and Mutual Support). Discussions were focused on breaking the barrier of hierarchy in a crisis situation. The correct number of triage, treatment, and teamwork aspects were assessed. Assessment of teamwork was done in two parts. One was self-assessment of perception of teamwork, the Teamwork Perceptions Questionnaire (TPQ), and the other was assessment of the team performances, the Team Performance Observation Tool (TPOT). Both tools were derived from the TeamSTEPPS^® Project (5-point Likert scale). They were modified to fit our culture and translated to Korean. Content validity index was performed (0.94). All pre-to-post differences within subjects were analyzed with paired t tests. The statistical level of significance was set at 0.05.

Results: Correct triaging and treatment improved after training with table-top simulation. Under triaging improved as well, but over triaging seemed to increase. Pre- and post-intervention differences for the 5 sections of the TPQ and TPOT improved. All results were statistically significant (p<0.05).

Conclusion: Teamwork and communication has cultural different aspects. Therefore, carefully planned curriculum tailored to the trainees, and debriefing session including discussion on cultural aspect is important when training for teamwork. Table-top exercise can positively affect perception toward teamwork, and their ability to recognize the presence and quality of team skills in disaster events.

INTRODUCTION

Well-prepared humanitarian workers are now more necessary than ever; simulation-based training and evaluation are essential for the preparation process. This study aimed to assess the efficacy of a training program developed by the Research Center in Emergency and Disaster Medicine (CRIMEDIM) in collaboration with the international organization Médecins Sans Frontières (MSF) to professionalize senior residents in humanitarian assitance.

METHODS

The first three levels of the Kirkpatrick’s evaluation model (Reaction-Learning-Behaviour) were assessed.The 8 participants of the third course edition (4 residents in emergency medicine, 3 in anaesthesia and 1 in pediatrics) were enrolled in the study; the median age was 31. Residents participated in an introductory phase, completed a 3 month e-learning course, attended a residential week and were deployed with MSF in Pakistan (1), Afghanistan (2), Democratic Republic of Congo (1), South Sudan (1), Central African Republic (2) and Yemen (1). Reaction was assessed through a Likert scale questionnaire. The three dimensions of Learning were evaluated separately through a pre and post test as follows: a multiple-choice test was used to assess knowledge, a Likert scale questionnaire was used to evaluate attitudes, and simulation-based performance tests (using the Ottawa Global Rating Scale-GRS) were used to assess skills. Total multiple-choice scores and GRS overall performance scores were considered as primary and secondary outcomes, respectively. Differences were assessed using paired t-­tests. P ­values of less than 0.05 were considered significant. Behaviour was assessed qualitatively at the end of students’ missions by their field supervisors through the MSF standard evaluation module. Supervisors were blind to the students’ participation in the training program.

RESULTS

Reaction: the delivery modality and residential course were highly appreciated. The average median score for the overall course was 5 (excellent). Knowledge: there was a significant improvement in the post-test multiple choice scores when compared to the pre-test scores (p = 0.0011). Skills: there was a significant improvement in the overall performance score (P = 0.000001). No differences were detected in attitudes scores. Behaviour: for most participants the following strengths were highlighted: compliance with MSF standards and principles, flexibility, good team working skills and cross-cultural sensitivity.Their professional competence was never questioned. All residents were recommended for future MSF missions.

CONCLUSIONS

Residents were highly satisfied with this training program and their knowledge and skills in low-resource simulated humanitarian environments improved after participation in the course. The implementation of this project represents a model of how academia can successfully partner with humanitarian aid organizations to increase professionalization within the humanitarian health sector.

Study/Objective:

To quantify the frequency and intention with which “stampede” is used to describe types of mass gathering disasters

Background:

Hazard vulnerability analysis would identify “human stampedes” as high probability events at mass gatherings. Over 200 “stampedes” have occurred in the past 30 years. At the 2015 Hajj, at least 2000 pilgrims died in one of the deadliest mass gathering disasters in recent history. News and literature referenced the event as the “Hajj Stampede”, implying abruptly increased speed and mass panic.  At the crux of many of these events, however, is a dense, immobile crowd – not the uncontrolled mindless mass implied.

Methods:

The authors performed a systematic search of peer reviewed literature indexed in PubMed, EMBASE, and Web of Science. Studies were limited to human studies using the keyword stampede. Gray literature using “stampede” in the title or abstract in reference to mass gathering disasters were also reviewed.

Results:

Search strategy using the term “stampede” yielded 649 articles. After excluding those using the term 1) apropos computing, 2) as an acronym, or 3) colloquially, 61 remained which used the term in reference to mass gathering disasters. 29 of 61 articles describe a slow-moving, highly dense crowd. 12 articles cite sudden mass movement as the main trigger for the referenced disaster. The remaining 18 described both slow-moving crowds and sudden mass movement. Only three articles distinguished between crowd disasters caused by sudden movement and high density. Overall, “Stampede” was used in the same context as “crowd disaster”, “turbulence”, “quake”, “mass panic”, “crush”, and “trampling”.

Discussion:

It is important to distinguish between stampede and non-stampede events for the benefit of survivors and for mitigation strategies. Few articles describing stampedes actually involve speed anywhere in the description. The generic “stampede”, through suggesting a fast moving, irrational and culpable crowd, focuses on herding the masses rather than improving safe access or egress routes to the gathering site. We must stem the notion that these disasters are a whim of the crowd and work towards evidence-based engineered solutions. 

Introduction: Nurses can be confronted with disasters and the care for the patients resulting from these incidents, be it in - or prehospital or in their own environment. Are they educated to do so? Following worrisome results in a survey on this subject amongst medical students our study hypothesis is that nursing students aren’t educated at all.

Material and methods: To evaluate disaster education in Flemish nursing students an online survey on Disaster Medicine, training and knowledge and willingness to report was sent to students in the last year of basic nursing training. This reported knowledge was tested by a mixed set of 10 theoretical/practical questions. A similar survey was sent to students in the Bachelor-after-Bachelor (BAB) specialisation year of Emergency Medicine and Intensive Care.

Results: Preliminary results from the first responding educational basic training centers reveal a M/F ratio of 13/87 with 29% that state that they have had any disaster training what so ever. 25% state to have some knowledge on CBRN incidents. 42% are convinced that a basic training on disaster management should absolutely be included in the basic nursing curriculum. None of the respondents found it useless. Estimated knowledge on several disaster scenarios varied from 1.92/10 (dirty bomb) over 2.24/10 (nuclear) and 3.49/10 (Ebola) to 4.05/10 in highly contagious influenza pandemic. Self-estimated capability to deal with these incidents varied from 2.22/10 (dirty bomb) over 2.59/10 (nuclear) and 3.54/10 (Ebola) to 4.14/10 in mass shooting incidents. Willingness to report to work in these incidents was much higher and varied from 7.19/10 (dirty bomb) over 7.41/10 (Ebola) and 7.56/10 (nuclear) to 7.94/10 in mass shooting incidents. Some topics of the theoretical / practical case mix raise some concern. 64% directs potentially contaminated patients direct into the emergency department. 75% believes that iodine tablets protect against external radiation and 37% would use them as the first step in nuclear decontamination. 37% believes that chemical decontamination consists of a total body antidote spray in a civil defence cabin and only 20% would use a shower with water and soap. Up to 47% would rush unprotected into a traffic accident scene with active leakage from a tanker truck. Self-reported knowledge and capability in the BAB group did not differ from the basic students and also the willingness to report was similar. They had however a better score on the theoretical / practical case mix.

Conclusion: Our data support the hypothesis that Flemish nursing students are ill-educated in disaster management. Despite low estimated knowledge and capability there is a high willingness to report.

In the context of the current European refugee crisis, at the Geneva University Hospitals (HUG) the languages which caused most problems were in 2016 Tigrinya, Arabic and Farsi. Several researchers pointed out serious problems of quality, security and equitability when no communication is possible between the doctor and his patient. BabelDr (http://babeldr.unige.ch/) is a common project of Geneva University's Faculty of Translation and Interpreting (FTI) and HUG which aims at facing this problem. The BabelDr application is a flexible speech-enabled phrase-book. The linguistic coverage is organised into domains, centered around body parts (abdomen, chest, head, kidney). Each domains has a limited semantic coverage consisting of 2000-2500 canonical sentences, but users can use a wide variety of surface forms when speaking to the system. The translation is not automatic; the canonical forms are translated into the target languages by translation experts from the FTI, which guarantees the quality of translation. At runtime, the system matches the spoken doctor’s utterance to a canonical sentence and echoes it back to the source-language user, only producing a translation if the source-language user approves. 

We compared BabelDr with the statistical MT system Google Translate (GT) for the anamnesis in emergency settings. French speaking doctors were asked to use both systems to diagnose Arabic speaking patients with abdominal pain, based on two scenarios. For each scenario (appendicitis and cholecystitis), a patient was standardized by the HUG.  Participants were four medical students and five doctors from HUG, who each performed two diagnoses, one with BabelDr and one with GT. 

The translation quality was evaluated in terms of adequacy and comprehensibility by three Arabic advanced translation students. Adequacy was judged on a four point scale (nonsense/mistranslation/ambiguous/correct) and comprehensibility on a four point scale (incomprehensible/syntax errors/non idiomatic/fluent). For the BabelDr translations, 93% of doctor's interactions sent to translation were correct and 94% fluent at the majority judgements. For GT, we respectively obtained 38% and 38%. Inter-annotator agreement for both evaluations was moderate (Light's Kappa for adequacy: 0.483; for comprehensibility: 0.44). With Google Translate 5/9 doctors found the correct diagnosis, against 8/9 with BabelDr. The satisfaction of doctors was also higher with BabelDr than with Google Translate: doctors were more confident in the translation to the target language with BabelDr than Google Translate (1/9 negative opinion with BabelDr vs 8/9 in GT). They also think they could integrate BabelDr in their everyday practice in the emergency room, contrary to GT (1/9 negative opinion with BabelDr vs 5/9 with GT).These results tend to show that BabelDr is a promising tool for the task and that GT translations are insufficiently adequate, accurate and comprehensible for emergency settings. 

Background

Emergency department (ED) crowding and increasing ED utilization are well-recognized problems for patient care in the United States.  To what degree predictive analytic techniques can improve wait times and patient outcomes when employed early in the ED stay--specifically during the triage process--is not well described.  We created predictive models to compare logistic regression (LR) and multilayer neural network (MLNN) techniques to predict hospital admission/transfer or discharge following initial presentation to ED triage with and without the addition of natural language processing (NLP) to analyze patient-reported free-text information.

Methods

Using data from the National Hospital Ambulatory Medical Care Survey (NHAMCS), a cross-sectional probability sample of United States EDs from 2012 and 2013 survey years, we developed several models to predict patients’ disposition: hospital admission or transfer vs. discharge. We included patient characteristics which are immediately available after the patient presents to the ED following ED triage. We used this information to construct logistic regression and multilayer neural network models which included NLP and principal component analysis to incorporate the patient-reported reason for visit.  Ten-fold cross validation was used to test the predictive capacity of each model and c-statistics / receiver operating curves (AUC) were calculated to compare these predictive models.

Results

Of the 47,200 ED visits from 642 hospitals, 6,335 (13.4%) resulted in hospital admission (or transfer). A total of 48 principal components were extracted by NLP from patient’s reason for visit, which explained 75% of the overall variance for hospitalization. In the model excluding patient’s free-text reason for visit, the AUC was 0.824 (95% CI 0.818-0.830) for LR and 0.823 (95% CI 0.817-0.829) for MLNN. When patients’ free text reasons for visit were included, the AUC increased to 0.846 (95% CI 0.839-0.853) for LR and 0.844 (95% CI 0.836-0.852) for MLNN.

Conclusions

The predictive accuracy of hospital admission/transfer or discharge for patients who presented to ED triage improved with the inclusion of free text data from patients’ reason for visit regardless of modeling approach.  The predictive ability of these models was generally quite good at predicting disposition with the limited information immediately available during the triage process. Natural language processing and multilayer neural networks provide ways to incorporate patient-reported free-text information when predicting various outcomes that are important in providers’ clinical decision-making. 

Background:

Coronary Artery Disease (CAD) is one of the major causes of death. Evidence suggests that some preventive measures by patients in emergency phase can reduce the rate and risk of mortality. Thus, understanding the signs and risk factors of CAD from the patients’ perspective and their ways of dealing with this disease is of vital importance.

Objectives: This qualitative study aimed to explore the Iranian patients’ experiences about CAD and how they manage it in their first encounter.

Patients and Methods: This study was a grounded theory study conducted on 18 patients with CAD. The data were collected through semi-structured interviews. Initially, purposeful sampling was performed followed by maximum variety. Sampling continued until data saturation. Then, all the interviews were recorded and transcribed verbatim. After all, the data were analyzed by constant comparative analysis using MAXQUDA2010 software.

Results:

The themes manifested in this phase of disease included 1- "Invasion of Disease"  with subthemes of "warning signs" and "risk factors", 2- "Patients’ Primary Challenges" with subthemes of "doubting primary diagnosis and treatment", and feeling of being different from others", 3- "Psychological Issues" with subthemes of "mental preoccupation", "fear of death and surgical intervention", "stress due to recurrence",  and "anxiety and depression", 4- "Management Strategies" with subthemes of "seeking for information", "follow-up' , and "control measures".

Conclusions: Based on the results, physicians and nurses should focus on empowerment of patients by facilitating this process as well as by educating them with regards to dealing with CAD. Further, it is also essential for the mass media to educate the public on how to treat patients with CAD.

Key words: Management, Coronary Artery Disease, Emergency Phase, Grounded theory, Iran 

Background: Our Emergency department at New Queen Elizabeth Hospital in Birmingham is one of the busiest Major Trauma Centres in UK with around 250 staff with annual attendances of more than 100,000 patients. The issues of work load, space, staffing level and performance makes it a challenging environment. A multidisciplinary teams are more likely to perform better if they are made to feel appreciated and their good practices are acknowledged. 

Learning from excellence is a new concept. The positive feedbacks are widely used in education for children and by corporate non- medical institutions. We wanted to explore its role in adult education and behaviour in the medical background. It’s very rare we recognise, celebrate and learn from good practice in the current working climate. The feedback had to timely, genuine, reflective, meaningful, specific and consistent. A template of a positive feedback along with specific local human factors framework (modification of NOTECHS II system) was developed based on the previous work on Favourable Event Reporting Form (FERF). The Human factors were analysed under the following topics- Leadership & management, Team building & cooperation, Problem solving and decision making, Situational awareness and environment.