Background and aim Femoral vein catheterization is very commonly performed while managing children with septic shock. However, the agreement between superior venal caval (ScvO₂) and femoral venous oxygen saturation (SfvO₂) has not been reported in children till date. Our aim was to evaluate the agreement between ScvO₂ and SfvO₂ values.

Methods Children with septic shock in whom femoral vein catheter was inserted before shifting to the ICU were eligible. Paired blood samples from subclavian/Internal jugular vein and femoral catheters were drawn simultaneously after initial fluid resuscitation in these children and analysed. Agreement was assessed by Bland and Altman (BA) analysis. The study was approved by IRB.

Results A total of 32 patients were enrolled. The mean ScvO₂ and SfvO₂ were 72.3% and 76% respectively (p < 0.001). BA analysis showed relatively poor agreement between ScvO₂ and SfvO₂ with mean bias of 3.6 and 95% limits of agreement of -1.95 to 9.25 (r=-0.28, p = 0.14) (Figure 1). In general, SfvO₂ values were higher than ScvO₂ values with 14 of the 32 children having values ≥5%. The SfvO₂ had poor sensitivity (45.5%) but 100% specificity for detecting low saturations. The PPV was 100% and the NPV was 77.8%. As SfvO2 values were generally higher than ScvO₂ values, we tried a higher cut-off for SfvO₂ to detect low mixed venous saturations. With a SfvO₂ cut off of 72%, the sensitivity and specificity were both 100%.

Conclusions A higher cut off of 72% for SfvO₂ may be used in place of ScvO₂ in situations where ScvO₂ estimation has failed or is impractical. Our study findings however, need further validation.

• Download figure
• Open in new tab
• Download powerpoint

Abstract O-117 Figure 1

Agreement between ScvO₂ and SfvO₂ values at 1 h