RESUMO
OBJECTIVES: To perform a systematic review of the accuracy of carotid ultrasound measures in determining volume responsiveness in adults. METHODS: We conducted a systematic review of Ovid MEDLINE and Scopus from conception until January 1, 2019. Two independent reviewers used an iterative process to identify relevant articles and abstract information from them. The quality and risk of bias were assessed with the Quality Assessment of Diagnostic Accuracy Studies version 2 tool. RESULTS: We identified 17 relevant articles with 956 patients. The 2 most frequently cited carotid measures of fluid responsiveness were corrected flow time and peak velocity or change in peak velocity with respiration (ΔCDPV). Accordingly, the diagnostic characteristics of corrected flow time in these studies varied widely, with sensitivities from 60% to 73%, specificities from 82% to 92%, and areas under the receiver operating characteristic curves from 0.75 to 0.88. Optimal cutoff values for ΔCDPV ranged from 9.1% to 14%, with areas under the receiver operating characteristic curves from 0.81 to 0.91, sensitivities from 73% to 86%, and specificities from 78% to 86%. Other measures, such as carotid blood flow and carotid diameter, had limited data to support their use. Heterogeneity of the studies prohibited a meta-analysis. Most studies had a moderate risk of bias and high applicability. CONCLUSIONS: Preliminary research suggests that carotid ultrasound measures may be useful adjunct measures of fluid status; however, they should not be interpreted as absolute and should be placed in a clinical context. The most well-defined and supported measure currently is ΔCDPV, with cutoffs from 9% to 14%. Corrected flow time shows promise, because of heterogeneity of how this value is measured, an optimal cutoff has not been established.