Variability of peripheral arterial peak velocity predicts fluid responsiveness in patients with septic shock / 中华危重病急救医学

ABSTRACT

Objective To explore the accuracy of fluid responsiveness assessment by variability of peripheral arterial peak velocity and variability of inferior vena cava diameter (&Delta;IVC) in patients with septic shock. Methods A prospective study was conducted. The patients with septic shock undergoing mechanical ventilation (MV) admitted to intensive care unit (ICU) of Beijing Electric Power Hospital from January 2016 to December 2017 were enrolled. According to sepsis bundles of septic shock, volume expansion (VE) was conducted. The increase in cardiac index (&Delta;CI) after VE ≥ 10% was defined as liquid reaction positive (responsive group), &Delta;CI < 10% was defined as the liquid reaction negative (non-responsive group). The hemodynamic parameters [central venous pressure (CVP), intrathoracic blood volume index (ITBVI), stroke volume variation (SVV), &Delta;IVC, variability of carotid Doppler peak velocity (&Delta;CDPV), and variability of brachial artery peak velocity (&Delta;Vpeak-BA)] before and after VE were monitored. The correlations between the hemodynamic parameters and &Delta;CI were explored by Pearson correlation analysis. Receiver operating characteristic (ROC) curve was plotted to analyze the predictive value of all hemodynamic parameters on fluid responsiveness. Results During the study, 74 patients with septic shock were included, of whom 9 were excluded because of peripheral artery stenosis, recurrent arrhythmia or abdominal distension influencing the ultrasound examination, and 65 patients were finally enrolled in the analysis. There were 31 patients in the responsive group and 34 in the non-responsive group. SVV, &Delta;IVC, &Delta;CDPV and &Delta;Vpeak-BA before VE in responsive group were significantly higher than those of the non-responsive group [SVV (12.3±2.4)% vs. (9.2±2.1)%, &Delta;IVC (22.3±5.3)% vs. (15.5±3.7)%, &Delta;CDPV (15.3±3.3)% vs. (10.3±2.4)%, &Delta;Vpeak-BA (14.5±3.3)% vs. (9.6±2.3)%, all P < 0.05]. There was no significant difference in CVP [mmHg (1 mmHg = 0.133 kPa) 7.5±2.5 vs. 8.2±2.6] or ITBVI (mL/m2 875.2±173.2 vs. 853.2±192.0) between the responsive group and non-responsive group (both P > 0.05). There was no significant difference in hemodynamic parameter after VE between the two groups. Correlation analysis showed that SVV, &Delta;IVC, &Delta;CDPV, and &Delta;Vpeak-BA before VE showed significant linearity correlation with &Delta;CI (r value was 0.832, 0.813, 0.854, and 0.814, respectively, all P < 0.05), but no correlation was found between CVP and &Delta;CI (r = -0.342, P > 0.05) as well as ITBVI and &Delta;CI (r = -0.338, P > 0.05). ROC curve analysis showed that the area under ROC curve (AUC) of SVV, &Delta;IVC, &Delta;CDPV, and &Delta;Vpeak-BA before VE for predicting fluid responsiveness was 0.857, 0.826, 0.906, and 0.866, respectively, which was significantly higher than that of CVP (AUC = 0.611) and ITBVI (AUC = 0.679). When the optimal cut-off value of SVV for predicting fluid responsiveness was 11.5%, the sensitivity was 70.4%, and the specificity was 94.7%. When the optimal cut-off value of &Delta;IVC was 20.5%, the sensitivity was 60.3%, and the specificity was 89.7%. When the optimal cut-off value of &Delta;CDPV was 13.0%, the sensitivity was 75.2%, and the specificity was 94.9%. When the optimal cut-off value of &Delta;Vpeak-BA was 12.7%, the sensitivity was 64.8%, and the specificity was 89.7%. Conclusions Ultrasound assessment of &Delta;IVC, &Delta;CDPV, and &Delta;Vpeak-BA could predict fluid responsiveness in patients with septic shock receiving mechanical ventilation. &Delta;CDPV had the highest predictive value among these parameters.