Purpose
A perceived simple task as providing basic life support (BLS) can be difficult and studies have shown room for improvement. Therefore, we performed a study of whether a simplified algorithm for BLS could reduce time from recognition of arrest till delivery of the first shock. 

Materials and methods
In this before and after study, we performed unannounced in-situ simulations in the emergency department of Odense University Hospital, Denmark. The simulations were conducted in all areas of the department (e.g. toilets, patient and treatment rooms) primarily during day-time and evenings using regular staff and equipment (Laerdal Resusci-Anne mannequin, Zoll R-series and Lifepak 20 defibrillators) from May 2015 till May 2019. Data were collected from the mannequin and two cameras. Primary outcome was time from recognition of cardiac arrest to initial shock. The intervention was a simplified BLS algorithm (figure 1) as supplement to standard ERC algorithm (before). 

Results
We conducted 36 simulations before and 22 simulations after. Time from recognition to shock was 5:16 (95% CI 4.5-6.0 minutes) and 3:46 (95% CI 3.0-4.5 minutes), respectively, see figure 2. We found a reduction of 28%, p<0.02. Furthermore, we observed a series of system errors such as difficulty operating especially one of the two types of the defibrillators. Technical difficulties when calling both the cardiac arrest response team and alerting nearby staff. Other difficulties such as participants administrating medications outside ERC-guidelines and incorrect use of the bag valve mask. Participants did not optimize conditions for administering BLS (e.g. failing to adjust height of bed for compressions or remove pillow before ventilating the patient).

Conclusion
Our data show, that by introducing a simplified algorithm for BLS in an ED, we could reduce time to first shock by 28%. Also, a series of system errors were identified and sought addressed during the study.

2 - A simplified algorithm – Is simpler better?