Carboxyhemoglobin Levels As a Predictor of Risk for Significant Hyperbilirubinemia in African-American DAT(+) Infants

Objectives: To compare the degree of hemolysis in a group of direct antiglobulin test (DAT) positive (pos) African-American (AA) infants as measured by carboxyhemoglobin corrected (COHbc) for carbon monoxide in ambient air to a similar group of DAT negative (neg) ABO incompatible infants and a group without blood group incompatibility. To determine if COHbc is a better predictor of significant hyperbilirubinemia than DAT status.

Study Design: A prospective study of 180 AA infants from the Well-Baby Nursery of an inner city community hospital, all of whose mothers were type O pos. Infants (60) were ABO incompatible DAT pos, 60 were ABO incompatible DAT neg and 60 were type O(+). Blood for COHbc was drawn at the time of the infants' initial bilirubin and the infants' precise percentile on the Bhutani nomogram was calculated.

Result: Mean COHbc of type O(+) infants was 0.76±0.21 and 0.78±0.24% for ABO incompatible DAT neg infants (P=0.63). Mean CoHbc for the ABO incompatible DAT pos infants was 1.03±0.41% (P<0.0001 compared with both type O and DAT neg infants). Optimal cutoff on the receiver operating characteristic curve for COHbc to determine the risk for being in the Bhutani curve high risk zone was COHbc >0.90% (area under the curve(AUC) 0.8113). This was similar to the AUC of the receiver operating characteristic curve using any titer strength of DAT pos as a cutoff (0.7960).

Conclusion: Although not greatly superior to the titer strength of DAT pos, COHbc is useful in determining if the etiology of severe hyperbilirubinemia is a hemolytic process.