Diagnostic accuracy of stroke volume variation for predicting fluid responsiveness in children undergoing cardiac surgery: a systematic review and meta-analysis.

BACKGROUND Stroke volume variation (SVV) appears to be reliable for predicting fluid responsiveness (FR) in adults and its predictive value in paediatric patients has been recently reported. However, its predictive value in children undergoing cardiac surgery is unclear. METHODS A review and meta-analysis were performed on the diagnostic utility of SVV for predicting FR in children undergoing cardiac surgery. All relevant articles for prospective research assessing the value of SVV were searched in the Embase, MEDLINE (PubMed), and Cochrane databases through March 2020. The primary outcome was the accuracy of SVV for predicting FR in children. The combined data were analysed by a meta-analysis. Publication quality was assessed using the QUADAS (quality assessment for studies of diagnostic accuracy, maximum score) standard guidelines. RESULTS Six articles were included in the meta-analysis, following the search strategy. A total of 251 children were included from 6 prospective studies. Fluid therapy for all patients used crystalloids or colloids. The results of the analysis revealed a pooled diagnostic odds ratio of 8.23 (95% CI: 3.07-22.11), pooled sensitivity of 0.73 (95% CI: 0.64-0.80), and pooled specificity of 0.66 (95% CI: 0.58-0.74). Additionally, the overall area of the summary receiver operating characteristic curve (SROC) was 0.78. There was significant moderate heterogeneity in these studies (p<0.05, I2 = 42.1%) due to thresholds. CONCLUSIONS There was some heterogeneity due to thresholds in the included studies. An evaluation of SVV may represent a reliable predictor of FR in children undergoing cardiac surgery. After operative cardiac output optimization, the possible impact of goal-directed fluid treatment depending on SVV on the perioperative outcome in the children population should subsequently be assessed.