Team resuscitation for paediatrics (TRAP); application and validation of a paediatric resuscitation quality instrument in non-simulated resuscitations.

Abstract

Background: Resuscitation of paediatric cardiac and respiratory arrest is a high-stakes and low frequency event in the paediatric emergency department. Resuscitation team performance assessment tools have been developed and validated for use in the simulation environment, but no tool currently exists to evaluate clinical performance in non-simulated, live paediatric resuscitations.

Methods: This is a validation study assessing inter-rater reliability of a novel assessment tool of clinical performance of non-simulated resuscitations, the Team Resuscitation for Paediatrics tool. Videos of medical resuscitations at a tertiary care paediatric emergency department were collected and analysed over a 6-month period. Four paediatric emergency medicine attending physicians reviewed the videos and scored team performance based on the tool. Percent agreement and Fleiss' Kappa were calculated in 3 subcategories: team communication, cardiac arrest and respiratory arrest. Percent agreement ranges were established a priori as > 80 % considered good and < 60 % poor.

Results: Of 51 resuscitations occurring during the study period, 24 met inclusion criteria. All subcategories demonstrated overall moderate agreement however individual items showed a wide range of agreement. Kappa scores were low on both individual items and overall. Three of four items on the team communication tool met criteria for good agreement, 12/34 items on the cardiac arrest tool met good agreement and 9/27 items on the respiratory arrest tool met good agreement.

Conclusion: This study demonstrated that development, application and testing of clinical tools to assess resuscitation team performance of non-simulated, video-recorded resuscitations is feasible, however, the Team Resuscitation for Paediatrics tool did not demonstrate adequate inter-rater reliability suggesting that further tool development may be necessary to better evaluate clinical resuscitation performance.