Comparison of pulse pressure variation, stroke volume variation, and plethysmographic variability index in pediatric patients undergoing craniotomy / 北京大学学报(医学版)

OBJECTIVE@#To compare well-known preload dynamic parameters intraoperatively including stroke volume variation (SVV), pulse pressure variation (PPV), and plethysmographic variability index (PVI) in children who underwent craniotomy for epileptogenic lesion excision.@*METHODS@#A total of 30 children aged 0 to 14 years undergoing craniotomy for intracranial epileptogenic lesion excision were enrolled. During surgery, we measured PPV, SVV (measured by the Flotrac/Vigileo device), and PVI (measured by the Masimo Radical-7 monitor) simultaneously and continuously. Preload dynamic parameter measurements were collected at predefined steps: after induction of anesthesia, during opening the skull, intraoperative electroencephalogram monitoring, excision of epileptogenic lesion, skull closure, at the end of the operation. After exclusion of outliers, agreement among SVV, PPV, and PVI was assessed using repeated measures of Bland-Altman approach. The 4-quadrant and polar plot techniques were used to assess the trending ability among the changes in the three parameters.@*RESULTS@#The mean SVV, PPV, and PVI were 8%±2%, 10%±3%, and 15%±7%, respectively during surgery. We analyzed a total of 834 paired measurements (3 to 8 data sets for each phase per patient). Repeated measures Bland-Altman analysis identified a bias of -2.3 and 95% confidence intervals between -1.9 and -2.7 (95% limits of agreement between -6.0 and 1.5) between PPV and SVV, showing significant correlation at all periods. The bias between PPV and PVI was -5.0 with 95% limits of agreement between -20.5 and 10.5, and that between SVV and PVI was -7.5 with 95% limits of agreement between -22.7 and 7.8, both not showing significant correlation. Reflected by 4-quadrant plots, the con-cordance rates showing the trending ability between the changes in PPV and SVV, PPV and PVI, SVV and PVI were 88.6%, 50.4%, and 50.1%, respectively. The concordance rate between PPV and SVV was higher (92.7%) in children aged less than 3 years compared with those aged 3 and more than 3 years. The mean angular bias, radial limits of agreement, and angular concordance rate in the polar analysis were not clinically acceptable in the changes between arterial pressure waveform-based parameters and volume-based PVI (PPV vs. PVI: angular mean bias 8.4°, angular concordance rate 29.9%; SVV vs. PVI: angular mean bias 2.4°, angular concordance rate 29.1%). There was a high concordance between the two arterial pressure waveform-based parameters reflected by the polar plot (angular mean bias -0.22°, angular concordance rate 86.6%).@*CONCLUSION@#PPV can be viewed as a surrogate for SVV, especially in children aged less than 3 years. The agreement between arterial pressure waveform-based preload parameters (PPV and SVV) and PVI is poor and these two should not be considered interchangeable. Attempt to combine PVI and PPV for improving the anesthesiologist's ability to monitor cardiac preload in major pediatric surgery is warranted.

Plethysmographic variability index as a tool to assess fluid responsiveness in critically ill patients: a correlation study with inferior vena cava distensibility index

Background: In critically ill patients in the intensive care unit (ICU), early aggressive fluid replacement is the cornerstone of resuscitation. Traditionally employed static measures of fluid responsiveness have a poor predictive value. It is therefore imperative to employ dynamic measures of fluid responsiveness that take into account the heart lung interactions in the mechanically ventilated patients. The main objective of this study was to evaluate the reliability of one such non-invasive dynamic index: Plethysmographic variability index (PVI) compared to the widely employed Inferior vena cava distensibility index (dIVC).Methods: Seventy-six adult patients admitted at a tertiary care mixed ICU, who developed hypotension (MAP<65mmHg), were included in the study. PVI was recorded using the MASIMO-7 monitor and dIVC measurements done using Terason ultrasound. Based on the dIVC measurement threshold of 18%, the patients were classified into volume responders and non-responders. The hemodynamic, PVI and dIVC measurements were recorded at pre specified time points following a fluid challenge of 20 ml/kg crystalloid infusion.Results: Baseline PVI values were significantly higher in the responders (22.3±8.2) compared to non-responders (10.1±2.9) (p<0.001) and showed a declining trend at all time points in the responders. Similar declining trend was observed in the dIVC measurements. Overall, the Pearson correlation graph showed strong correlation between dIVC and PVI values at all time points (r=0.678, p=0.001). The ROC curve between the dIVC and PVI values revealed that Baseline PVI (Pre PVI) >15.5% discriminated between responders and non-responders with a 90.2% sensitivity and 75% specificity with an AUC of 0.84 (0.72-0.96) (p<0.001).Conclusions: There is good correlation between PVI values and measured dIVC values at baseline and following a fluid challenge. Thus, PVI may be an acceptable, real time, continuous, surrogate measure of fluid responsiveness in critically ill patients.