Increasing arterial oxygen partial pressure during cardiopulmonary resuscitation is associated with improved rates of hospital admission.

AIM As recent clinical data suggest a harmful effect of arterial hyperoxia on patients after resuscitation from cardiac arrest (CA), we aimed to investigate this association during cardiopulmonary resuscitation (CPR), the earliest and one of the most crucial phases of recirculation. METHODS We analysed 1015 patients who from 2003 to 2010 underwent out-of-hospital CPR administered by emergency medical services serving 300,000 inhabitants. Inclusion criteria for further analysis were nontraumatic background of CA and patients >18 years of age. One hundred and forty-five arterial blood gas analyses including oxygen partial pressure (paO2) measurement were obtained during CPR. RESULTS We observed a highly significant increase in hospital admission rates associated with increases in paO2 in steps of 100 mmHg (13.3 kPa). Subsequently, data were clustered according to previously described cutoffs (≤ 60 mmHg [8 kPa]], 61-300 mmHg [8.1-40 kPa], >300 mmHg [>40 kPa]). Baseline variables (age, sex, initial rhythm, rate of bystander CPR and collapse-to-CPR time) of the three compared groups did not differ significantly. Rates of hospital admission after CA were 18.8%, 50.6% and 83.3%, respectively. In a multivariate analysis, logistic regression revealed significant prognostic value for paO2 and the duration of CPR. CONCLUSION This study presents novel human data on the arterial paO2 during CPR in conjunction with the rate of hospital admission. We describe a significantly increased rate of hospital admission associated with increasing paO2. We found that the previously described potentially harmful effects of hyperoxia after return of spontaneous circulation were not reproduced for paO2 measured during CPR. CLINICAL TRIAL REGISTRATION n/a.