BackgroundTo assess the validity of central and pulmonary veno-arterial CO2 gradients to predict fluid responsiveness and to guide fluid management during liver transplantation.MethodsIn adult recipients (ASA III to IV) scheduled for liver transplantation, intraoperative fluid management was guided by pulse pressure variations (PPV). PPV of ≥15% (Fluid Responding Status-FRS) indicated fluid resuscitation with 250 ml albumin 5% boluses repeated as required to restore PPV to < 15% (Fluid non-Responding Status-FnRS). Simultaneous blood samples from central venous and pulmonary artery catheters (PAC) were sent to calculate central venous to arterial CO2 gap [C(v-a) CO2 gap] and pulmonary venous to arterial CO2 gap [Pulm(p-a) CO2 gap]. CO and lactate were also measured.ResultsSixty seven data points were recorded (20 FRS and 47 FnRS). The discriminative ability of central and pulmonary CO2 gaps between the two states (FRS and FnRS) was poor with AUC of ROC of 0.698 and 0.570 respectively. Central CO2 gap was significantly higher in FRS than FnRS (P = 0.016), with no difference in the pulmonary CO2 gap between both states. The central and Pulmonary CO2 gaps are weakly correlated to PPV [r = 0.291, (P = 0.017) and r = 0.367, (P = 0.002) respectively]. There was no correlation between both CO2 gaps and both CO and lactate.ConclusionCentral and the Pulmonary CO2 gaps cannot be used as valid tools to predict fluid responsiveness or to guide fluid management during liver transplantation. CO2 gaps also do not correlate well with the changes in PPV or CO.Trial registrationClinicaltrials.gov Identifier: NCT03123172. Registered on 31-march-2017.