Evaluation of Out-of-Hospital Cardiac Arrest Recognition by Non-physicians Dispatchers in a French Emergency Dispatch Call Center

1. . Justification The performance of phone-guided chest compressions is recommended to minimize the No Flow risk (1). Pre-arrival instructions and dispatch-assisted chest compressions are the first link in the chain of survival, determining the effectiveness of subsequent links. This first link is only possible if a prior detection of cardiac arrest was made. The Dispatch Center of the Fire Brigade of Paris receives more than 7,000 calls per day, including 8-10 calls for cardiac arrest. The establishment of specific training and experience feedback with emergency dispatchers should improve the detection of cardiac arrest (CA) by the same dispatchers.
2. . Objectives The main objective is to measure an improvement in the detection rate of CA in the Dispatch Center as a result of debriefings and repeated trainings for non-medical operators who receive emergency calls. The secondary objective is to identify risk factors for CA non-detection by those emergency dispatchers.
3. . Materials and Methods Experimental design: Observational prospective non-randomized controlled study. Type "before and after" study. Three observation periods are defined over a total period of 20 months. Different actions before and during each period: Period 1 Period 2 Period 3 Algorithm available to emergency dispatchers + + + New algorithm - - + Basic training for new operators - + + Daily briefing +/- +/- + Systematic Debreafing - - + Specific course before period - - + Table leaders training - - + Heads of room training - - + Period 1 is preceded and accompanied by the diffusion of a decision algorithm available to emergency dispatchers Period 2 is accompanied by:

• Short briefings, twice daily, in the presence of operators on duty that day
• Debriefing immediately after the call with some operators (but not systematically, depending on their availability). Period 3 is preceded by information to the Dispatch Center managers (heads of room and table leaders) and a short course about detection of CA (1 hour) for all operators. During this same period, a daily briefing for operators is established. The study is based on the three CA call periods' audio recordings For each period, all calls related to CA on the field were listened to and evaluated. Each call had a double audition:
• one audition by an instructor, responsible for managing calls in the Dispatch Center, and
• one audition by two emergency physicians experienced in the management of emergency calls. The second listening physician helped to verify, correct, and complete the variables collected during the first listening The variables collected during listening were:
• "detectability of the CA"
• detection or not of the CA
• the causes of non-detection
• the occurrence of dispatch-assisted chest compression
• the time between the start of the call and the time the operator pronounces the words "cardiopulmonary Resuscitation" or "CPR" or "chest compressions"
• the time between the start of the call and the first real chest compression
• causes for non-realization of chest compressions
• the presence or absence of a public defibrillator Other variables collected - Variables corresponding to the Utstein style:
• Epidemiological variables (age, sex, location, background)
• Specific backup system variables: time alert - use of an AED
• Specific first aid and medical management variables:
• type of rhythm detected by AED on the arrival of rescuers
• Success of first AED's shock (loss of ventricular fibrillaton (VF)5 seconds after shock)
• Number of shocks administered by the AED in the first 10 minutes of support by rescuers
• Recovery of a palpable pulse in the presence of rescuers, before the arrival of the medical team
• Resuscitation undertaken by the medical team YES/NO
• Injection of adrenaline: YES/NO
• Transport to the hospital with beating heart: YES/NO
• Released resuscitation YES/NO
• Survival to one year after the CA
• Quality of survival (CPC score) at 6 months, 1 year Main judgment criteria:
• Detection of CA by the emergency dispatcher:
• Detection is considered positive if the dispatcher formalized the presence of a CA , either on the software, or orally with the words "CPR" or "chest compression" at any time during the recorded call.
• It is considered negative if the dispatcher never suggested, neither orally nor in writing, the possibility of the presence of a CA.
• Secondary endpoints:
• Median time between the start of the emergency call and the pronunciation of the words "chest compressions," by the dispatcher
• Median time elapsed between the start of the emergency call and the actual start of chest compressions by caller
• Evaluation of the ventilation mode (hand on abdomen YES/NO, count the interval between two ventilatory movements YES/NO)

1. . Statistics

Sample size We need 100 patients in each period group to observe an absolute 20% difference in the rate of CA detection, compared with the detection rate of the control group (period 1) estimated at 60%, with an alpha risk = 0.05 and an 80% power in bilateral situations.

For each period:

• Description of patient characteristics and outcome

Comparison of these periods:

• Comparison of detection rate between the reference period (P1) and each of the other periods
• The number of subjects included in each period allows the realization of a χ² test with an alpha risk retained at 5%
• Adjust the detection rate on the significantly different variables between the groups (age, sex, no flow, time of rescuer arrival)
• The period 1 and 2 will be grouped and compared with period 3.

In all periods:

• Search variables associated with successful detection
• Multivariate logistic regression analysis type
• Survival analysis.
• Description of the evolution of the detection rate over time