Abstract 2646: CPR Quality and the Effect of Audiovisual Feedback in a U.S. Emergency Department
Background: The quality of chest compressions (CC) during cardiac arrest (CA) resuscitation has been shown to be suboptimal during both in-hospital and out-of-hospital CA. However, neither the quality of CCs delivered in US Emergency Departments (EDs) nor the impact of real-time audiovisual feedback on CC quality in the ED has been studied.
Objective: We sought to examine CC quality in an academic ED, and hypothesized that audiovisual feedback would improve CC quality over the initial minute of resuscitation care.
Methods: CC data were collected from Sept 2008 to May 2009 from an urban, tertiary care hospital ED. CPR-sensing monitor/defibrillators enabled to provide real-time audiovisual feedback were used during CA events, with feedback prompts given to correct CC depth, rate, and leaning. Rescuer response to feedback was measured within the first “block” of uninterrupted CCs (≥100 sec) from each CA, thus representing single rescuers. These blocks of CCs were subdivided into two 30 sec periods to compare CPR quality “before” and “after” a period of CPR with feedback.
Results: CC data were analyzed from 37 resuscitation episodes in 36 patients. Mean patient age was 59.5±15.1; 19/36 (53%) were male. The overall ROSC rate was 43%, and survival to discharge was 14%. The predominant initial rhythms were asystole (46%) and PEA (35%). Quality analysis of all recorded CCs from the first five min of resuscitation (n=14,129) revealed a mean rate of 107±9/min, with 21% of CCs too shallow (<38 mm) and leaning in 11% of CCs (≥2.5 kg of force). When the first uninterrupted blocks of CCs were analyzed to examine rescuer performance before and after a period of time with feedback, mean CC rate and variance remained statistically unchanged from 107±11 to 107±8 (p=NS), as did the percentage of CCs too shallow (19% both before and after, p=NS), however the percentage of CCs with leaning fell significantly (14% to 7%, p<0.001).
Conclusions: The overall quality of CCs delivered in this ED was high, and little improvement was seen within initial CPR blocks associated with the delivery of audiovisual feedback. The one parameter that did improve significantly was leaning during CC. Further work will be required to assess feedback effects later during CA events or during discontinuous CPR.