Objectives: A systematic review of the literature to determine the ability of dynamic changes in arterial waveform-derived variables to predict fluid responsiveness and compare these with static indices of fluid responsiveness. The assessment of a patient's intravascular volume is one of the most difficult tasks in critical care medicine. Conventional static hemodynamic variables have proven unreliable as predictors of volume responsiveness. Dynamic changes in systolic pressure, pulse pressure, and stroke volume in patients undergoing mechanical ventilation have emerged as useful techniques to assess volume responsiveness.
Data Sources: MEDLINE, EMBASE, Cochrane Register of Controlled Trials and citation review of relevant primary and review articles.
Study Selection: Clinical studies that evaluated the association between stroke volume variation, pulse pressure variation, and/or stroke volume variation and the change in stroke volume/cardiac index after a fluid or positive end-expiratory pressure challenge.
Data Extraction and Synthesis: Data were abstracted on study design, study size, study setting, patient population, and the correlation coefficient and/or receiver operating characteristic between the baseline systolic pressure variation, stroke volume variation, and/or pulse pressure variation and the change in stroke index/cardiac index after a fluid challenge. When reported, the receiver operating characteristic of the central venous pressure, global end-diastolic volume index, and left ventricular end-diastolic area index were also recorded. Meta-analytic techniques were used to summarize the data. Twenty-nine studies (which enrolled 685 patients) met our inclusion criteria. Overall, 56% of patients responded to a fluid challenge. The pooled correlation coefficients between the baseline pulse pressure variation, stroke volume variation, systolic pressure variation, and the change in stroke/cardiac index were 0.78, 0.72, and 0.72, respectively. The area under the receiver operating characteristic curves were 0.94, 0.84, and 0.86, respectively, compared with 0.55 for the central venous pressure, 0.56 for the global end-diastolic volume index, and 0.64 for the left ventricular end-diastolic area index. The mean threshold values were 12.5 ± 1.6% for the pulse pressure variation and 11.6 ± 1.9% for the stroke volume variation. The sensitivity, specificity, and diagnostic odds ratio were 0.89, 0.88, and 59.86 for the pulse pressure variation and 0.82, 0.86, and 27.34 for the stroke volume variation, respectively.
Conclusions: Dynamic changes of arterial waveform-derived variables during mechanical ventilation are highly accurate in predicting volume responsiveness in critically ill patients with an accuracy greater than that of traditional static indices of volume responsiveness. This technique, however, is limited to patients who receive controlled ventilation and who are not breathing spontaneously.