[Dispatchers' impressions and actual quality of cardiopulmonary resuscitation during telephone-assisted bystander cardiopulmonary resuscitation: a pooled analysis of 94 simulated, manikin-based scenarios].

Objectives: The quality of telephone-assisted cardiopulmonary resuscitation (CPR) needs improvement. This study investigates whether a dispatchers' perception is an adequate measure of the actual quality of CPR provided by laypersons.
Material and Methods: Individual participant data from 3 randomized simulation trials, with identical methodology but different interventions, were combined for this analysis. Professional dispatchers gave telephone assistance to laypersons, who each provided 10 minutes of CPR on a manikin. Dispatchers were requested to classify the quality of providers' CPR as adequate or inadequate. Based on actual readings from manikins we classified providers' performance as adequate at 5-6 cm for depth and 100-120 compressions per minute (cpm) for rate. We calculated metrics of dispatcher accuracy.
Results: Six dispatchers rated the performance of 94 laypersons (38 women [42%]) with a mean (SD) age of 37 (14) years. In 905 analyzed minutes of telephone-assisted CPR, the mean compression depth and rate was 41 (13) mm and 98 (24) cpm, respectively. Analysis of dispatchers' diagnostic test accuracy for adequate compression depth yielded a sensitivity of 65% (95 CI 36%-95%) and specificity of 42% (95% CI, 32%-53%). Analysis of their assessment of adequate compression rate yielded a sensitivity of 75% (95% CI, 64%-86%) and specificity of 42% (95% CI, 32%-52%). Although dispatchers always underestimated the actual values of CPR parameters, the female dispatchers evaluations were less inaccurate than the evaluations of make dispatchers; the dispatchers overall (males and females together) underestimated the adequacy of female laypersons' CPR performance to a greater degree than female dispatchers did.
Conclusion: The ability of dispatchers to estimate the quality of telephone-assisted CPR is limited. Dispatchers estimates of CPR adequacy needs to be studied further in order to find ways that telephone-assisted CPR might be improved.