Continuous central venous oxygenation measurement by CeVOX in patients undergoing off-pump coronary artery bypass grafting

Introduction

Less invasive measurement of central venous O₂ saturation(ScvO₂) has shown to be a valuable alternative to the determination of mixed venous O₂ saturation for monitoring of O₂ supply/demand [1, 2]. The aim of this study was to compare ScvO₂ measured continuously by the new CeVOX (Pulsion Medical System, Munich, Germany) device (CScvO₂) with ScvO₂ determined by blood gas co-oximetry (BScvO₂).

Methods

Ten ASA III patients undergoing elective off-pump coronary artery bypass grafting were studied during the operation (OP) and during their ICU stay. In addition to the standard hemodynamic monitoring according to institutional policy, a CeVOX fiber-optic probe was introduced into a standard central venous catheter placed via internal jugular vein access. OP and ICU measurement started after in-vivo calibration of CeVOX. BScvO₂ and CScvO₂ readings were recorded at intervals of 30 min during OP and of 120 min during ICU. Data were statistically analyzed using Bland-Altman analysis, Pearson correlation and t test for the periods during OP, ICU, a set of three consecutive measurements during OP immediately after calibration (OP_cal) and 4 hours later (OP_4h) as well as immediately after calibration on the ICU (ICU_cal) and 14 hours later (ICU_14h). Trend analysis was performed, calculating differences (Δ) between consecutive measurements. P < 0.05 was considered significant.

Results

One hundred and twenty-nine matched sets of data were obtained (OP:n = 78, ICU: n = 51) with a wide range of ScvO₂ values (BScvO₂ = 48.0–91.0%, CScvO₂ = 49.0–94.0%). The OP observation time was 4.0–6.5 hours and the ICU measurement sequence was 14.0–20.0 hours. Bland-Altman analysis revealed an overall mean bias ± 2SD (limits of agreement) of-0.7 ± 7.8% for CScvO₂ – BScvO₂ during OP and -1.1± 11.6% during ICU (Fig. 1). There was no significant difference between CScvO₂ and BScvO₂ (OP: P = 0.120, ICU: P = 0.167). The correlation coefficient(r²) for CScvO₂ vs BScvO₂ was 0.885 (OP) and0.592 (ICU). Statistics for OP_cal' OP_4hand ICU_cal were comparable, whereas for ICU_14hthe bias ± 2SD increased and r² decreased (Table 1). Trend analysis showed no significant difference (OP: ΔBScvO₂ = -1.3 ± 9.0%, ΔCScvO₂ = -1.1 ± 8.6%, P = 0.663; ICU: ΔBScvO₂ = -0.4 ± 6.8%, ΔBScvO₂ = -0.4 ± 5.8%, P = 0.828).

Figure 1

Table 1

Conclusion

These preliminary results indicate that ScvO₂ can be reliably assessed by CeVOX. Scheduled recalibration at intervals <14 hours may be mandatory.