Predictive value of central venous-to-arterial carbon dioxide partial pressure difference for fluid responsiveness in septic shock patients: a prospective clinical study / 中华危重病急救医学

ABSTRACT

Objective To evaluate the accuracy of central venous-to-arterial carbon dioxide partial pressure difference (Pcv-aCO2) before and after rapid rehydration test (fluid challenge) in predicting the fluid responsiveness in patients with septic shock. Methods A prospective observation was conducted. Forty septic shock patients admitted to medical intensive care unit (ICU) of Peking Union Medical College Hospital from October 2015 to June 2017 were enrolled. All of the patients received fluid challenge in the presence of invasive hemodynamic monitoring. Heart rate (HR), blood pressure, cardiac index (CI), Pcv-aCO2 and other physiological variables were recorded at 10 minutes before and immediately after fluid challenge. Fluid responsiveness was defined as an increase in CI greater than 10% after fluid challenge, whereas fluid non-responsiveness was defined as no increase or increase in CI less than 10%. The correlation between Pcv-aCO2 and CI was explored by Pearson correlation analysis. Receiver operating characteristic (ROC) curves were established to evaluate the discriminatory abilities of baseline and the changes after fluid challenge in Pcv-aCO2 and other physiological variables to define the fluid responsiveness. The patients were separated into two groups according to the initial value of Pcv-aCO2. The cut-off value of 6 mmHg (1 mmHg = 0.133 kPa) was chosen according to previous studies. The discriminatory abilities of baseline and the change in Pcv-aCO2(&Delta;Pcv-aCO2) were assessed in each group. Results A total of 40 patients were finally included in this study. Twenty-two patients responded to the fluid challenge (responders). Eighteen patients were fluid non-responders. There was no significant difference in baseline physiological variable between the two groups. Fluid challenge could increase CI and blood pressure significantly, decrease HR notably and had no effect on Pcv-aCO2 in fluid responders. In non-responders, blood pressure was increased significantly and CI, HR, Pcv-aCO2 showed no change after fluid challenge. Pcv-aCO2 was comparable in responders and non-responders. In 40 patients, CI and Pcv-aCO2 was inversely correlated before fluid challenge (r = -0.391, P = 0.012) and the correlation between them weakened after fluid challenge (r = -0.301, P = 0.059). There was no significant correlation between the changes in CI and Pcv-aCO2 after fluid challenge (r = -0.164, P = 0.312). The baseline Pcv-aCO2 and &Delta;Pcv-aCO2 could not discriminate between responders and non-responders, with the area under ROC curve (AUC) of 0.50 [95% confidence interval (95%CI) =0.32-0.69] and 0.51 (95%CI = 0.33-0.70), respectively. HR and blood pressure before fluid challenge and their changes after fluid challenge showed very poor discriminative performances. Before fluid challenge, 16 patients had a Pcv-aCO2 > 6 mmHg. Their mean CI was significantly lower and Pcv-aCO2 was significantly higher than that in 24 patients whose Pcv-aCO2 ≤6 mmHg [n = 24; CI (mL·s-1·m-2) 48.3±11.7 vs. 65.0±18.3, P < 0.01; Pcv-aCO2 (mmHg) 8.4±1.9 vs. 2.9±2.8, P < 0.01]. Pcv-aCO2was decreased significantly after fluid challenge in patients with an initial Pcv-aCO2 > 6 mmHg and their &Delta;Pcv-aCO2 was notably different as compared with the patients whose baseline Pcv-aCO2≤6 mmHg (mmHg -3.8±3.4 vs. 0.9±2.9, P < 0.01). 68.8% (11/16) patients responded to the fluid challenge in patients with an initial Pcv-aCO2 > 6 mmHg. The AUC of the baseline Pcv-aCO2 and &Delta;Pcv-aCO2 to define fluid responsiveness was 0.85 (95%CI = 0.66-1.00) and 0.84 (95%CI = 0.63-1.00), respectively, and the positive predictive value was 1 when the cut-off value was 8.0 mmHg and -4.2 mmHg, respectively. 45.8% (11/24) patients responded to the fluid challenge in patients whose baseline Pcv-aCO2≤6 mmHg. There was no predictive value of baseline Pcv-aCO2 and &Delta;Pcv-aCO2 on fluid responsiveness. Conclusion Pcv-aCO2 and its change cannot serve as a surrogate of the change in cardiac output to define the response to fluid challenge in septic shock patients whose baseline Pcv-aCO2≤6 mmHg, while the predictive values of baseline Pcv-aCO2and the change in Pcv-aCO2 are presented in patients with the initial value of Pcv-aCO2 > 6 mmHg. Clinical Trial Registration Clinical Trials, NCT01941472.