Fluid responsiveness after cardiac surgery (FRACAS): a prospective observational study using peripheral near-infrared spectroscopy

Objectives: To describe tissue oxygen saturation (StO₂) in response to a vascular occlusion test using thenar eminence and forearm near-infrared spectroscopy (NIRS) and the association with volume responsiveness after cardiac surgery. Design: Single-center, prospective, observational cohort study. Setting: Cardiothoracic intensive care unit. Participants: Seventy-six post-cardiac surgical adults. Interventions: Immediately before and 10 minutes after a 250-to-500 mL fluid bolus, StO₂ was measured in response to a vascular occlusion test to calculate tissue deoxygenation (R_des) and reoxygenation (R_res) rates. Concurrently, systemic hemodynamic, metabolic, and blood gas variables were collected. Measurements and Main Results: A total of 203 boluses were captured using thenar NIRS and 141 boluses using forearm NIRS. Approximately 25% of boluses increased cardiac output by ≥15% (volume responders). Thenar and forearm R_des decreased in responders, but increased (thenar) or remained unchanged (forearm) in nonresponders. A logistic regression model of the association among StO₂, R_des and R_res, and volume responsiveness was significant for thenar measurements (p = 0.001) with an area under the receiver operating characteristic of 0.69 (95% confidence interval: 0.62-0.75). It also was significant (p = 0.02) for forearm measurements, with an area under the receiver operating characteristic of 0.71 (0.62-0.79). R_des was an independent variable in both instances (odds ratio 0.31 [0.14-0.69], thenar; odds ratio 0.60 [0.45-0.80], forearm). Thenar and forearm NIRS variables were correlated poorly with cardiac output, stroke volume, systemic oxygen delivery and consumption index, mixed venous, and central venous oxygen saturation (Spearman's coefficients, r = 0.17-0.46, p < 0.002). Conclusion: In post-cardiac surgical patients, thenar and forearm NIRS variables were associated with volume responsiveness although not achieving precision necessary for clinical management.