Background

Prehospital termination of resuscitation (pTOR) criteria including unwitnessed, no shock administered, and no prehospital return of spontaneous circulation (ROSC) in out-of-hospital cardiac arrest (OHCA) patients was widely accepted. However, patients who met the pTOR criteria had small but non-ignorable chance to achieve ROSC in the emergency department. We hypothesized that test values that could be obtained from blood gas analysis could improve the negative prediction of ROSC after the hospital arrival in OHCA patients fulfilling the pTOR criteria.

Methods

A retrospective observational study used a multicenter (73 centers) prospective registry of OHCA patients between June 2014 and December 2015 in Japan. Patients fulfilling the pTOR criteria were included. The study index tests were potassium level (mEq/L), pH, and partial oxygen pressure (mmHg) obtained in the first test after hospital arrival. Receiver operating characteristic (ROC) analysis for not being able to achieve ROSC in the emergency department used to assess the prognostic indices of the study tests consisted of area under the ROC curve (AUROC),  threshold at the given sensitivity of 0.99 and specificity at the estimated threshold. 

Results

Of the 13,491 patients registered in the registry, 1,736 OHCA patients who met the pTOR criteria were included. Of these, 256 (14.7%) patients achieved ROSC in the emergency department. AUROC for the prediction of not being able to achieve ROSC was 0.788, 0.633, and 0.623 with potassium level, pH, and partial oxygen pressure, respectively. Thresholds at the given sensitivity of 0.99 were >10.6 mEq/L potassium level, <6.38 pH, and <5.5 mmHg partial oxygen pressure. Specificities at the estimated thresholds were 0.19, 0.01, and 0.03, respectively.

Conclusion

Blood gas analyses did not sufficiently rule out subsequent ROSC in OHCA patients fulfilling the pTOR criteria.

4 - Blood gas analyses to predict subsequent recovery of spontaneous circulation in out-of-hospital cardiac arrest patients fulfilling prehospital termination of resuscitation criteria: A diagnostic accuracy study