POSTNATAL UMBILICAL CORD BLOOD TESTING: PROLIFERATION OF CONFOUNDING NON-VALUE ADDED LABORATORY TESTING

Background:   The measurement of umbilical arterial and venous blood pH is a minimum medico legal standard used to rule out prolonged asphyxia which can result in neonatal brain damage.  There is a continuous stream of postnatal investigations that generally support the measurement of more cord blood analytes including lactate, bicarbonate, base excess, and even troponin.  Our attempt to limit cord blood gas testing to only pH was met with recommendations to engage our obstetrical colleagues.  As there is reluctance amongst most clinicians to limit test ordering, we conducted a literature review to better understand the weaknesses of cord blood testing.

Methods:   We performed a non-systematic literature review to determine the salient factors associated with abnormal cord blood tests and their clinical interpretation.  Of 637 articles (cord blood gas analysis at delivery), 618 (umbilical cord blood gas analysis at delivery), 18 (cord blood troponin at delivery) and 80 [(cord blood troponin or (umbilical cord troponin) or (cord troponin)], we reviewed a clinical subset (43 papers) to gain insight into the complexities of using postnatal measurements to diagnose neonatal asphyxia or improve obstetrician feedback.

Results:   Based on these papers, we assembled some of the factors that confound test results and medical decision-making.

1. Preanalytical conditions that alter test results including placental size, placental pathology, cord length, gestational age, time of labor, time elapsed between birth and cord clamping, time between cord ligation and blood sampling, sampling from the umbilical artery (20 to 40% error rate), storage of blood specimen in the case of prolonged analysis delay and time elapsed between blood sampling and analysis.   
2. Analytical Issues including analytical differences between different blood gas instrument and the fact that some results are calculated and not measured.
3.  Decision-Making Issues
   1. Multiple tests – no change in normal limits to compensate for decreased specificity
   2. Inadequate reference limits for normal, abnormal, profoundly abnormal
   3. Recommendations to use these test results for asphyxia diagnosis and physician feedback with little knowledge of positive and negative predictive values
   4. While some professional groups advocate testing only higher risk patients, many of these measurements are performed in the general population, thus decreasing the predictive value of a positive test.

Conclusion: Based on these multiple confounders, it seems that postnatal cord analysis should be classified as a research test until its utility is proven or disproven.