Determinants Of Venous-to-arterial CO₂ Difference And The Diagnostic Value Of CO₂ Production For Fluid Challenge In Shock Patients

Background:About one-third of the patients developed shock during their intensive care unit(ICU)stay.Early recognition and restoration of hypoperfusion are paramount important.Mixed venous-to-arterial carbon dioxide pressure difference(P(v-a)CO2)is a sensitive indicator of organ hypoperfusion.The value of P(v-a)CO2 is determined by carbon dioxide production(VCO2),cardiac output(CO),and the Haldane effect.Purpose:The purpose of the study was to investigate:(i)the diagnostic value of changes of P(v-a)CO2(ΔP(v-a)CO2)in the judgement of fluid responsiveness;(ⅱ)the contribution of VCO2,CO,the Haldane effect and their change(Δ VCO2,ΔCO,andΔ Haldane effect)to P(v-a)CO2 or ΔP(v-a)CO2 in patients with septic shock;(ⅲ)the contribution of Δ VCO2,ΔCO,and Δ Haldane effect to the change of central venous to arterial CO2 pressure difference(Δ P(cv-a)CO2)in shock patients;(ⅳ)the diagnostic value of the change of VCO2(ΔVCO2)in shock patients receiving fluid challenge(FC).Methods:The study was conducted in two parts.The first part used a retrospective approach.From October 2016 to October 2020,patients with septic shock during the medical ICU(MICU)stay in our hospital were screened.Adult patients were included when they equipped with a pulmonary artery catheter(PAC),received a fluid challenge(FC)and had simultaneous sampling from arterial and mixed venous blood gases before and after the fluid challenge.Fluid responsiveness was defined as cardiac output increased by more than 10%immediately after a FC.The CO2 content calculation model was used to quantify the Haldane effect.A receiver operating characteristic(ROC)curve was drawn to evaluate the diagnostic value of ΔP(v-a)CO2 in the judgement of fluid responsiveness.The contribution of VCO2,CO and the Haldane effect and their change to P(v-a)CO2 or ΔP(v-a)CO2 were accounted for by multiple linear regression.The contribution rate of the variable was obtained by dividing the standardized coefficient of the variable by the sum of absolute values of all standardized coefficients which is significant.The second part used a prospective approach.From October 2022 to February 2023,all shock patients requiring FCs and mechanical ventilation during their MICU stay were included.Blood gases were analyzed by simultaneous sampling from arterial and central venous before and after the FC.We investigated the contribution of Δ VCO2 calculated by Fick equation(ΔVCO2.calc),Δ CO,and Δ Haldane effect to Δ P(cv-a)CO2 by multiple linear regression.In addition,VCO2 was measured(VCO2.meas)by indirect calorimetry before and after the FC.The Spearman test was used for the correlation between the change of VCO2.meas(Δ VCO2meas)and the change of hemodynamic parameters.A ROC curve was drawn to evaluate the diagnostic value of ΔVCO2.meas in the judgement of fluid responsiveness.Results:In the first part,we included 95 septic shock patients.The baseline VCO2,CO and P(v-a)CO2 were 182±80 ml/min,6.7±2.3 L/min and 5(3,7)mmHg,respectively.Patients were divided into fluid responsive(FR)(n=56)and fluid non-responsive(FNR)groups(n=39).There was no significant difference in baseline VCO2 and P(v-a)CO2 between the two groups.The Baseline CO was lower in the FR group(6.1Δ2.2 vs.7.4±2.3 L/min,p=0.007).The contribution rates of VCO2,CO and the Haldane effect to P(v-a)CO2 were 43%,28%,and 29%before FCs and 47%,23%,and 30%after FCs,respectively.The ΔVCO2 had the largest influence on theΔP(v-a)CO2(standardized beta coefficient=0.97,p<0.001),followed by ΔCO(standardized beta coefficient=-0.29,p<0.001).The contributions of ΔVCO2 andΔCO to ΔP(v-a)CO2 were 77%and 23%,respectively.The contributions of theΔHaldane effect to ΔP(v-a)CO2 had no statistical significance(standardized beta coefficient=0.06,p=0.151).In the second part,we included 26 patients.A total of 28 FCs were conducted.The twenty-eight conductions were divided into FR(n=19)and FNR groups(n=9).There was no significant difference in baseline VCO2,CO and P(cv-a)CO2 between the two groups.The ΔVCO2.calc had the largest influence on theΔP(cv-a)CO2(standardized beta coefficient=0.84,p<0.001).The standardized beta coefficient of ΔCO was-0.35(P=0.06).After the FC,VCO2.meas was increased in FR group(190±75 to 201±73ml/min,P<0.05)and decreased in FNR group(159 ±61 to 153 ± 63ml/min,P<0.05).ΔVCO2.meas was moderately correleted with the change of mean arterial pressure(AMAP)(Spearman r=0.40,P=0.036).ΔVCO2.meas was also moderately correleted with ΔCO(Spearman r=0.43,P=0.024).The area under the ROC curve of ΔVCO2.meas for the judgement of fluid responsiveness in shock patients was 0.90(95%confidence interval,0.78-1.00,P<0.001).The threshold was 1.5 ml/min.The sensitivity and the specificity of this threshold were 79%and 100%,respectively.Conclusion:Among patients with septic shock,ΔP(v-a)CO2 was not able to diagnose the fluid responsiveness of volume expansion.The reason was that the contribution rate of ΔCO to ΔPv-aCO2 was only about 20%.ΔVCO2 showed a larger contribution to Δ P(v-a)CO2 than Δ CO did.Similar results were found for ΔP(cv-a)CO2 in patients with other types of shock.There is a significant correlation between the change of VCO2 and hemodynamic parameters,and ΔVCO2 measured by indirect calorimetry before and after the FC could determine the fluid responsiveness of patients with shock.