Is venous blood drawn from femoral access adequate to estimate the central venous oxygen saturation and arterial lactate levels in critically ill patients?

ABSTRACT

OBJECTIVES: The purpose of this study was to test if venous blood drawn from femoral access can be used to estimate the central venous oxygen saturation and arterial lactate levels in critically ill patients. METHODS: Bland-Altman analysis and Spearman correlations were used to compare the femoral venous oxygen saturation and central venous oxygen saturation as well as arterial lactate levels and femoral lactate. A pre-specified subgroup analysis was conducted in patients with signs of hypoperfusion. In addition, the clinical agreement was also investigated. RESULTS: Blood samples were obtained in 26 patients. In 107 paired samples, there was a moderate correlation (r = 0.686, p < 0.0001) between the central venous oxygen saturation and femoral venous oxygen saturation with a bias of 8.24 ± 10.44 (95% limits of agreement -12.23 to 28.70). In 102 paired samples, there was a strong correlation between the arterial lactate levels and femoral lactate levels (r = 0.972, p < 0.001) with a bias of -2.71 ± 9.86 (95% limits of agreement -22.03 to 16.61). The presence of hypoperfusion did not significantly change these results. The clinical agreement for venous saturation was inadequate, with different therapeutic decisions in 22.4% of the situation; for lactate, this was the case only in 5.2% of the situations. CONCLUSION: Femoral venous oxygen saturation should not be used as a surrogate of central venous oxygen saturation. However, femoral lactate levels can be used in clinical practice, albeit with caution.