Traditional Poster Round
The American Heart Association recommends starting cardiopulmonary resuscitation (CPR) within 1 minute of cardiopulmonary arrest (CPA) because we know CPR saves lives. Little research exists characterizing the recognition phase of cardiac arrest or the time it takes to start CPR. However, in a recent pilot study approximately 50% of cases did not initiate CPR within 1 minute, with a wide variability in the time to initiation of CPR.
The primary aim is to understand the cognitive and decision making process by which a bedside provider recognizes simulated CPA and decides to initiate CPR. The secondary aim is to evaluate the impact of the amount of information on the bedside monitoring on the timeliness of the decision and action to initiate CPR. The hypothesis is it will take longer to initiate CPR in a complexly monitored patient when compared to a simply monitored patient who experiences pulseless electrical activity (PEA) resulting in CPA.
Bedside providers will participate in a simulation on medical decision making in the acutely decompensating pediatric patient, however they will not be informed of the goal to perform CPR. Two groups will experience the same case; a pediatric patient recently admitted to the PICU after experiencing a head trauma requiring a decompressing craniotomy. At the start of the case the patient will be intubated, sedated with multiple other injuries and low blood pressure concerning for developing shock. At 90 seconds into the case, the patient will experience PEA requiring CPR. One group will have the patient on a simple monitor, with only electrocardiography (ECG), end-tidal carbon dioxide, arterial line and pulse oximetry wave forms and respiratory rate. The other group will have the same patient on a complex monitor, with ECG, central venous pressure, intracranial pressure, end-tidal carbon dioxide, arterial line and pulse oximetry waveforms and respiratory rate. We will record how long the provider took from CPA to initiate CPR.
All cases will be video and audio recorded for analysis of actions taken. Errors of omission and commission will be recorded. A subset of participants will be asked to perform a cognitive task analysis where they will walk through their thought process during the case while walking through the room. This will be a group of approximately 5 per profession (nurse vs physician) and level of experience (novice vs expert) for a total of 20 analyses.
Finding the right study design and measurements to answer the question of does more data on the bedside monitor complicate the medical decision making process and delay the initiation of CPR.
Questions for discussion:
We believe a mixed methods approach will be appropriate for this pilot study, using a combination of human factors analysis and recording the time to initiation of CPR. Given the high variability in time observed, the ability to discern statistical significance from a simulation-based study with limited study subject is difficult. We chose to have one type of CPA (PEA vs bradycardia vs ventricular fibrillation) and will analysis based on experience and profession. However, the monitor complexity may have different influence based on the type of CPA and experience of the provider. Overall, we would like to collect data that will be actionable to improving patient care, but acknowledge the complexity of the question. Is there a different study design that could better answer the question?