Predictive value of partial pressure of end-tidal carbon dioxide on the effect of active abdominal compression-decompression cardiopulmonary resuscitation and serum S100B protein on cerebral function

Abstract

To explore the predictive value of partial pressure of end-tidal carbon dioxide (PETCO2) on the effect of active abdominal compression-decompression cardiopulmonary resuscitation (AACD-CPR) and serum S100B protein on cerebral function. 142 adult patients with in-hospital cardiac arrest (IHCA) AACD-CPR in Zhengzhou People's Hospital, Affiliated Southern Medical University from September 2014 to December 2017 were enrolled. Patients were divided into successful group and failure group according to restoration of spontaneous circulation (ROSC) or not; and then according to Glasgow-Pittsburgh cerebral performance categories (CPC) one month after ROSC, the successful group was divided into good prognosis group (CPC 1-2) and poor prognosis group (CPC 3-5) further. The variations of hemodynamic, arterial blood gas index, PETCO2 and serum S100B protein level (25 healthy subjects as normal S100B protein level reference value) during the recovery were analyzed. The predictive value of PETCO2 on the effect of AACD-CPR and serum S100B protein on cerebral function of successful resuscitation patients were analyzed by receiver operating characteristic curve (ROC). (1) According to the traditional qualitative indexes, such as pulsation of the large artery, redness of lips and extremities, spontaneous fluctuation of chest, narrowing of pupil, existence of shallow reflex, etc, 54 in 142 patients with IHCA were successfully resuscitated; 57 cases were successfully resuscitated through the guidance of PETCO2, there was no significant difference between the two groups (χ2 = 0.133, P = 0.715). With the AACD-CPR, 142 CA patients' arterial partial pressure of oxygen (PaO2), arterial blood carbon dioxide partial pressure (PaCO2) were all improved with different degrees; heart rate (HR), mean arterial pressure (MAP), PaO2 and PaCO2 were further improved at 20 minutes after ROSC. At beginning of AACD-CPR, PETCO2 of both groups were about 10 mmHg (1 mmHg = 0.133 kPa). PETCO2 was gradually rising to above 20 mmHg in successful group during AACD-CPR process; the failed group increased slightly within 2-5 minutes, then gradually decreased to below 20 mmHg, there was a significant difference in PETCO2 between the two groups at each time. The area under the ROC (AUC) of PETCO2 at CPR 20 minutes in predicting the outcome of the resuscitation was 0.969, 95% confidence interval (95%CI) was 0.943-0.995 (P = 0.000), when the cut-off value of PETCO2 was 24.25 mmHg, the sensitivity was 90.7%, and the specificity was 96.6%. (2) The level of serum S100B protein at 0.5 hour after ROSC in the good prognosis group and the poor prognosis group were significant higher than that of the normal control group; there was no significant difference between poor prognosis group and good prognosis group. S100B protein concentration of the poor prognosis group reached the peak within 3-6 hours, then gradually decreased, and was higher than that of the normal control group at ROSC 72 hours; the good prognosis was gradually decreased and recovered to normal control group within ROSC 72 hours. The AUC of S100B at 3 hours after ROSC on cerebral function prognosis prediction was 0.925, 95%CI was 0.867-0.984 (P = 0.000), when the cut-off value of S100B protein was 1.215 μg/L, the sensitivity was 85.2%, and the specificity was 85.5%. The variation of PETCO2 can be used as an objective index to predict the success of AACD-CPR, and serum S100B protein can be used as an objective clinical index to predict cerebral function after AACD-CPR, both of which have some reference and guiding significance for clinical treatment.