Study design and setting
This is a multicentre retrospective study. Consecutive patients with electrical injuries admitted to the Emergency Departments (ED) (adult and paediatric) of five French Hospitals were included. From these five centres, two EDs are University Hospital Centres (average ED admissions in 2019: 100,000 (70,000 in the adult department and 30,000 in the paediatric department), one is a High-Capacity Hospital Centre (average admissions in 2019 > 60,000) and two are Low-Capacity Hospital Centres (average admissions in 2019 < 60,000). These 5 hospitals cover a defined population in the west of France. In these ED, patients older than 16 years are referred to the adult ED managed by emergency physicians, and younger patients are referred to the paediatric department managed by paediatricians. None of the 5 centres had a well-established protocol for the management of patients suffering from electrical injuries. The STARD recommendations were followed for the reporting of diagnostic studies [21].
Firstly, we will describe the outcome of the entire cohort. Secondly, we will assess the predictive value of troponin in patients who had a troponin assay. Thirdly, we will perform a subgroup analysis according to risk factors to position troponin testing in a relevant as possible wat in the management of patients.
Participants
We extracted all discharges records of ED patients from a 15-year period (2005–2019) with the corresponding ICD-10 codes as the principal diagnosis (“Appendix”). The most affected victims (i.e., out of hospital cardiac arrest, extensive burns, severe rhythmic disorders, comatose patients or early MI) were treated outside the hospital by an Emergency Medical Service (EMS) and were not included in this study because they were referred directly to an intensive care unit (ICU) or an operating room without being admitted to the ED. Clinical data were obtained from the hospital information system and patient records. Baseline demographics, medical history, and antiarrhythmic medication were registered along with the location, time, and circumstances of the electrical injury. Furthermore, all clinical parameters which are deemed to be risk factors for cardiac arrhythmias based on the ERC criteria were summarised. We also recorded presenting symptoms, severity of burns, and other injuries.
Measurements
The threshold for troponin elevation is based on the European Society of Cardiology guidelines for patients presenting without persistent ST segment elevation as there are no specific prospective data available on this specific population affected by electrical accidents [22]. From 2005 to 2011, the laboratories of the five centres measured troponin T and the threshold was set at 0.03 μg l−1 according to the manufacturer’s guidelines; from 2011, high-sensitivity troponin T was used. We used the manufacturer’s recommended 99th percentile upper reference limit (URL) to reduce site-to-site variability when determining the cut-off point. A significant increase in troponin was defined as an increase of at least 10 ng/l within 6 h or 6 ng/l within 3 h as per the manufacturer’s guidelines.
Outcomes
The primary endpoint was the rate of Major Adverse Cardiac Events (MACEs) from admission to the ED until discharge or during hospitalisation or upon re-presentation to the ED or in another department of the referral hospital within 30 days. MACE was a composite measure defined as (i) acute myocardial infarction according to the Fourth Universal Definition of Myocardial Infarction, Myocardial Infarction with No Obstructive Coronary Arteries (MINOCA), myocarditis or Tako-Tsubo Syndrome [23], (ii) sustained Ventricular Tachycardia (VT) or Ventricular Fibrillation (VF) according to the definition provided by the American Heart Association [24], (iii) in-hospital cardiac arrest with return of spontaneous circulation (ROSC), (iv) death from any cause. Patients diagnosed with unstable angina or myocardial injuries not meeting the previously stated criteria were not considered as having suffered a MACE.
The secondary endpoint was the rate of cardiac events during in-hospital monitoring or upon re-presentation to the ED or in another department of the referral hospital within 30 days. A cardiac event includes MACE and the need for cardiac treatments (revascularisation and/or cardiac medication), non-sustained VT, arrythmias including sinus tachycardia (more than 30 min), atrial tachyarrhythmia or bradyarrhythmia requiring monitoring and/or treatment.
Ethics
Due to the non-interventional retrospective nature of the current study no informed consent was required (Deliberation no. 2016-262, 2016-263, CNIL MR-003). Ethical approval for this study was obtained from the Nantes Research Ethics Committee (Groupe Nantais d’Ethique dans le Domaine de la Santé, GNEDS).
Statistical analysis
Continuous variables are presented with their median, first and third quartiles (Q1–Q3). Categorical variables are summarised with the number of patients and percentage with a 95% confidence interval (95% CI). The Chi-squared test, Fisher’s exact test, Student’s t-test and the Mann–Whitney U test were used when appropriate (two-tailed; level of significance p < 0.05).
To assess the performances of the initial troponin aim 7 say and the second troponin assay for predicting the primary endpoint, we assessed diagnostic performances (sensitivity, specificity, negative and positive predictive values, negative and positive likelihood ratio and accuracy).
Performances of the 4 high-risk clinical items combined (with prior known heart disease and/or exposure to a high voltage of ≥ 1000 Volts and/or with initial loss of consciousness or/and an abnormal initial ECG) were assessed to predict MACE. This item is considered as positive if one or more item was present. The aim is to identify simple clinical criteria to isolate a group of patients who will not have MACE (i.e., < 1%) without the need for a bioassay, for a safe ED rule out.
A subgroup analysis was performed in low-risk patients and in high-risk patients using the same endpoint. Patients without ECG at baseline were excluded from this analysis.
All statistical analyses were 2-tailed, and a p value less than 0.05 was required for statistical significance. No imputation of missing data was performed. Since occurrences of electrical injuries are rare, we did not determine a necessary sample size. All statistical analysis was performed using R Statistical Software (version 4.0.3; R Foundation for Statistical Computing, Vienna, Austria) (URL https://www.R-project.org/).