Modeling instructor preferences for CPR and AED competence estimation

Cardiopulmonary resuscitation (CPR) and automated external defibrillator (AED) skills competency can be tested using a checklist of component skills, individually graded "pass" or "fail." Scores are typically calculated as the percentage of skills passed, but may differ from an instructor's overall subjective assessment of simulated CPR or AED adequacy.

To identify and evaluate composite measures (methods for scoring checklists) that reflect instructors' subjective assessments of CPR or AED skills performance best.

Associations between instructor assessment and lay-volunteer skill performance were made using 6380 CPR and 3313 AED skill retention tests collected in the Public Access Defibrillation Trial. Checklists included CPR skills (e.g., calling 911, administering compressions) and AED skills (e.g., positioning electrodes, shocking within 90 s of AED arrival). The instructor's subjective overall assessment (adequate/inadequate) of CPR performance (perfusion) or AED competence (effective shock) was compared to composite measures. We evaluated the traditional composite measure (assigning equal weights to individual skills) and several nontraditional composite measures (assigning variable weights). Skills performed out of sequence were further weighted from 0% (no credit) to 100% (full credit).

Composite measures providing full credit for skills performed out of sequence and down-weighting process skills (e.g., calling 911, clearing oneself from the AED) had the strongest association with the instructor's subjective assessment; the traditional CPR composite measure had the weakest association.

Our findings suggest that instructors in public CPR and AED classes may tend to down-weight process skills and to excuse step sequencing errors when evaluating CPR and AED skills subjectively for overall proficiency. Testing methods that relate classroom performance to actual performance in the field and to clinical outcomes require further research.