Impact of mechanical power and positive end expiratory pressure on central vs. mixed oxygen and carbon dioxide related variables in a population of female piglets

Abstract

Introduction: The use of the pulmonary artery catheter has decreased overtime; central venous blood gases are generally used in place of mixed venous samples. We want to evaluate the accuracy of oxygen and carbon dioxide related parameters from a central versus a mixed venous sample, and whether this difference is influenced by mechanical ventilation.

Materials and methods: We analyzed 78 healthy female piglets ventilated with different mechanical power.

Results: There was a significant difference in oxygen-derived parameters between samples taken from the central venous and mixed venous blood (S $\overline{v}$ O₂ = 74.6%, ScvO2 = 83%, p < 0.0001). Conversely, CO2-related parameters were similar, with strong correlation. Ventilation with higher mechanical power and PEEP increased the difference between oxygen saturations, (Δ[ScvO2-S $\overline{v}$ O₂ ] = 7.22% vs. 10.0% respectively in the low and high MP groups, p = 0.020); carbon dioxide-related parameters remained unchanged (p = 0.344).

Conclusions: The venous oxygen saturation (central or mixed) may be influenced by the effects of mechanical ventilation. Therefore, central venous data should be interpreted with more caution when using higher mechanical power. On the contrary, carbon dioxide-derived parameters are more stable and similar between the two sampling sites, independently of mechanical power or positive end expiratory pressures.

Keywords: ARDS; ScvO2; SvO2; carbon dioxide; central versus mixed gap; intrathoracic pressures; mechanical power; oxygen saturation; septic shock.

FIGURE 1

Linear regressions between oxygen and carbon‐dioxide related variables in a mixed and in a central venous blood sample. P$\overline{\mathrm{v}}$O₂ and PcvO₂ showed a moderate correlation (CCC = 0.51), as well as S$\overline{\mathrm{v}}$O₂ and ScvO₂ (CCC = 0.43). P$\overline{\mathrm{v}}$CO₂ and PcvCO₂ showed a very strong correlation (CCC 0.98). CCC, concordance correlation coefficient; mmHg, millimeters of mercury; PvO₂, mixed venous blood oxygen partial tension; P$\overline{\mathrm{v}}$CO₂, mixed venous blood carbon dioxide partial tension; PcvO₂, central venous blood oxygen partial tension; PcvCO₂, central venous blood carbon dioxide partial tension; S$\overline{\mathrm{v}}$O₂, mixed venous blood oxygen saturation; ScvO₂, central venous blood oxygen saturation.

FIGURE 2

Bland–Altman analysis between oxygen and carbon‐dioxide related variables in a mixed and in a central venous blood sample. Panel a: mean bias = 7.82 [7.30–8.35]; upper LOA = 21.19 [20.29–22.09]; lower LOA = −5.53 [−6.43 to‐4.64]. Panel b: mean bias = 8.47 [7.92–9.01]; upper LOA = 22.30 [21.37–23.23]; lower LOA = −5.36 [−6.29 to‐4.44]. Panel c: mean bias = 0.54 [0.39–0.69]; upper LOA = 4.38 [4.12–4.64]; lower LOA = −3.30 [−3.56 to‐3.04]. LOA, Level of Agreement; P$\overline{v}$O₂, mixed venous blood oxygen partial tension; P$\overline{\mathrm{v}}$CO₂,mixed venous blood carbon dioxide partial tension; PcvCO₂, central venous blood carbon dioxide partial tension; PcvO₂, central venous blood oxygen partial tension; S$\overline{\mathrm{v}}$O_{2 =} mixed venous blood oxygen saturation; ScvO₂, central venous blood oxygen saturation. 95% confidence intervals are shown in square brackets.

FIGURE 3

Time course of the Δ(S_cvO₂ – S$\overline{\mathrm{v}}$O₂) in the two groups of mechanical power during the whole experimental phase. Mean +/− SD of piglets in the lower MP group are represented in blue, while means +/− SD of piglets in the higher MP group are shown in red. Excluding the baseline timepoint, from the beginning of the experimental phase through the end of the experiment, the Δ(S_cvO₂−S$\overline{\mathrm{v}}$O₂) in the two groups of mechanical power was constant, without any effect of the time (β = 0.006, CI = [−0.07–0.06], p = 0.849). It was constantly different after setting the mechanical power (from timepoint 0.5 h towards 48 h). X‐axis: experimental timepoints, expressed as hours from the beginning of the experiment. Y‐axis: Δ(S_cvO₂−S$\overline{\mathrm{v}}$O₂), difference between central and mixed venous sample oxygen saturation.

FIGURE 4

Δ(S_cvO₂−S$\overline{\mathrm{v}}$O₂) in the two groups of MP, PEEP, TV and RR. Blue and red boxplots refer respectively to piglets ventilated with low and high mechanical power. MP, mechanical power; PEEP, positive end‐expiratory pressure; TV, tidal volume; RR, respiratory rate; Δ(S_cvO₂−S$\overline{\mathrm{v}}$O₂), difference between central and mixed venous blood oxygen saturation. For numerical details, see also Table S3 in the supplementary materials.