Does the plethysmographic variability index predict fluid responsiveness in mechanically ventilated children? A meta-analysis

Abstract

Editor—The plethysmographic variability index (PVI), a dynamic index of preload dependence based on the respiratory variations in the pulse oximetry plethysmographic waveform amplitude, is a reliable and non-invasive predictor of fluid responsiveness in mechanically ventilated adults.1Chu H Wang Y Sun Y Wang G Accuracy of pleth variability index to predict fluid responsiveness in mechanically ventilated patients: a systematic review and meta-analysis.J Clin Monit Comput. 2016; 30: 265-274Crossref PubMed Scopus (29) Google Scholar However, few studies in the literature have investigated the ability of the PVI to predict the response to volume expansion in the paediatric population, with conflicting results.2Gan H Cannesson M Chandler JR Ansermino JM Predicting fluid responsiveness in children: a systematic review.Anesth Analg. 2013; 117: 1380-1392Crossref PubMed Scopus (144) Google Scholar The goal of this meta-analysis was to summarize available evidence about the diagnostic accuracy of the PVI for the prediction of fluid responsiveness in children under mechanical ventilation. We searched the Medline, Google Scholar, and Cochrane databases, from inception to April 2016, to identify studies published as full-text articles in indexed journals that investigated the diagnostic accuracy of the PVI in predicting fluid responsiveness in mechanically ventilated children. The search terms used were as follows: ‘pleth’ or ‘plethysmography’; ‘child’ or ‘adolescent’ or ‘infant’; ‘variation’ or ‘variability’; and ‘index’ or ‘indices’. All statistical analyses were performed using Meta-Disc software version 1.4 (Ramon y Cajal Hospital, Madrid, Spain) for Windows. We calculated pooled values of diagnostic odds ratio, sensitivity, and specificity of PVI to predict the response to fluid challenge, using a random-effects model. A summary receiver operating characteristic curve was drawn to define the ability of PVI to discriminate between responders and non-responders to fluid challenge. Heterogeneity between studies was assessed using the Cochran's Q and I2 tests. All values were reported as the point estimate with 95% confidence interval (CI). Four studies, with a total of 144 patients and 187 fluid boluses, were included in the meta-analysis.3Pereira de Souza Neto E Grousson S Duflo F et al.Predicting fluid responsiveness in mechanically ventilated children under general anaesthesia using dynamic parameters and transthoracic echocardiography.Br J Anaesth. 2011; 106: 856-864Abstract Full Text Full Text PDF PubMed Scopus (114) Google Scholar, 4Renner J Broch O Gruenewald M et al.Non-invasive prediction of fluid responsiveness in infants using pleth variability index.Anaesthesia. 2011; 66: 582-589Crossref PubMed Scopus (85) Google Scholar, 5Byon HJ Lim CW Lee JH et al.Prediction of fluid responsiveness in mechanically ventilated children undergoing neurosurgery.Br J Anaesth. 2013; 110: 586-591Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar, 6Julien F Hilly J Sallah TB et al.Plethysmographic variability index (PVI) accuracy in predicting fluid responsiveness in anesthetized children.Paediatr Anaesth. 2013; 23: 536-546Crossref PubMed Scopus (28) Google Scholar The mean responder rate was 48%. All studies were carried out in the operating theatre. The mean threshold value for the identification of responders to volume expansion was 14% (sd 3). The area under the summary receiver operating characteristic curve of PVI to predict fluid responsiveness was 0.86 (Fig. 1). The pooled sensitivity, specificity, and diagnostic odds ratio of PVI for the overall population were 72% (95% CI 62–81), 81% (95% CI 71–88), and 14 (95% CI 7–31), respectively. No significant statistical heterogeneity between the studies was found for specificity or diagnostic odds ratio (P>0.1; I2<50%). Conversely, a significant heterogeneity was found for sensitivity (P=0.02; I2=69%). Given the low number of studies included in this work, meta-regression analysis to explore the possible sources of clinical or methodological heterogeneity between studies was not performed. In summary, this meta-analysis suggests that the PVI could be an accurate predictor of fluid responsiveness in children under mechanical ventilation in the operating theatre. In the future, this non-invasive haemodynamic monitoring tool could be incorporated into an intraoperative fluid management algorithm in the paediatric population. However, given the low number of studies and participants and the heterogeneity among studies in terms of sensitivity, additional studies are required to confirm our findings before recommending the PVI for routine assessment of fluid responsiveness in children. None declared.