691 - Probability of Arterial or Venous Source of Same-Source Umbilical Cord Blood Gases

Approximately one of seven concomitant arterial-venous umbilical cord blood gases (CAV-UBGs) are unreliable because the sample source was apparently transposed or both samples were drawn from the same type of blood vessel. It has been argued that same-source samples most likely come from the umbilical vein because the vein is larger and more easily sampled than the artery but there has been no previous objective study of this issue.


 

Objective:

We aimed to determine the probability of an arterial or venous source of same-source samples using the generalized additive binomial regression model (GAM) that we developed in an earlier study of CAV-UBGs using rigorously defined arterial-venous sources (Clin Chem Lab Med, 2022).


 

Design/Methods:

We evaluated 56,703 CAV-UBGs collected over 8.8 years from neonates born in 3 hospitals within the Hospices Civils de Lyon health system in Lyon, France. We excluded samples with preanalytic or analytic issues and those samples with a pCO₂ < 2.9 kPa (~22 mmHg), and we selected CAV-UBGs in which the absolute arterial-venous differences were < 0.02 units for pH, < 0.7 kPa (~5 mmHg) for pCO₂, and < 0.4 kPa (3 mmHg) for pO₂. From the resulting 1,878 CAV-UBGs considered to be sampled from the same source, we selected the 241 CAV-UBGs with a pH of both specimens of 6.70-7.25 to focus on this clinically and medical-legally important range. We averaged the hydrogen ion concentration, the pCO₂ and pO₂ values for each sample pair and used the GAM model to calculate the probability of an arterial (ProbAS) or venous source (ProbVS=1-ProbAS) with their respective 95% confidence intervals (CIs).


 

Results:

The distribution of the ProbAS values is shown in Figure 1. Overall, the ProbAS was >0.5 in 42% and ≤0.5 in 58% of the CAV-UBG pairs. The ProbAS for the CAV-UBG pairs as a function of pH is shown in Figure 2 and there was a steep change in the likelihood of an arterial source with a pH cutoff of 7.19. Considering CAV-UBG pairs with a pH ≤7.19, the ProbAS and CIs were >0.5 in 80%, ProbVS and CIs >0.5 in 16%, and the source indeterminant in 4% (Table 1). CAV-UBG pairs with a pO₂ < 2.0 kPa (15 mmHg) or a pCO₂ >8.2 kPa (~62 mmHg) were similarly much more likely to be arterial than venous.



 

Conclusion(s):

Our findings support the opinion that same-source CAV-UBGs are most likely to be venous for the majority of cord blood gases that have results within the normal ranges. However, same-source CAV-UBGs with pH ≤7.19 and/or pCO₂ >8.2 kPa and/or pO₂ < 2.0 kPa are much more likely to be arterial than venous. The arterial or venous source of any UBG may be determined in most cases using our GAM model.


 

Figure 1.jpeg

Figure 2.jpeg