Modest change in bilirubin calibration has major effect in neonatal care

Phototherapy, the standard treatment for neonatal jaundice, is guided by age-related serum bilirubin concentration; the older the baby, the higher is the recommended triggering serum bilirubin concentration.

So that for example, UK guidelines state that for a full-term baby who is 12 hours old, phototherapy should be started if serum bilirubin exceeds just 150 µmol/L (8.8 mg/dL), for those who are 72 hours old, the triggering serum bilirubin concentration is 300 µmol/L (17.5 mg/dL). 

Delivery of treatment for neonatal jaundice obviously depends crucially on the accuracy of serum bilirubin measurement and therefore the accuracy of the bilirubin value assigned to the calibration material used.

As this recently published study nicely demonstrates, quite modest change in assigned calibration value can have a remarkable clinical effect in terms of the proportion of babies diagnosed with significant hyperbilirubinemia, requiring phototherapy.

The study was conceived after the manufacturer of a bilirubin assay reduced the assigned value of their assay calibrator.

This adjustment was made after extensive investigation of customer claims that the method had a positive bias when compared with other methods in external quality control programs. 

The study was conducted at Kaiser Permanente Northern California, a healthcare group that includes 21 affiliated hospitals where the bilirubin assay in question is used. The neonatal units of 12 hospitals within the group share a common policy of submitting all neonates for serum bilirubin estimation prior to discharge. 

Investigators retrieved serum bilirubin results relating to all live births >35 weeks gestation at the 12 hospitals for 2 years (Jan 1st 2010 – April 30th 2012) prior to recalibration and 2 years post recalibration (July 1st 2012 – Dec 31st 2013).

This amounted to 61,677 bilirubin results before recalibration and 42,571 after recalibration. Prior to recalibration, mean (SD) serum bilirubin was 10.1 (4.9) mg/dL and post recalibration it was 8.8 (4.5) mg/dL.

The absolute difference was 1.25 mg/dL (21.3 µmol/L), which is in line with manufacturers’ prediction of the effect their reduction in calibrator value would have. 

In order to assess the clinical effect of the recalibration, investigators determined four outcome measures: the % of neonates with serum bilirubin >15 mg/dL (256 µmol/L); the % of neonates with serum bilirubin above their age-related threshold for phototherapy; the % of neonates who actually received phototherapy; and finally the % of neonates who had to be readmitted to hospital for phototherapy. 

Prior to recalibration, 20.4 % of neonates had serum bilirubin >15 mg/dL (256 µmol/L); this reduced to 12.4 % after recalibration.

Prior to recalibration, 9.3 % of neonates had a serum bilirubin concentration above their age-related threshold for initiation of phototherapy; this reduced to 4.5 % after recalibration.

Prior to recalibration, 9.4 % of neonates received phototherapy; this reduced to 3.9 % after recalibration. And finally, prior to recalibration 3.8 % of neonates were readmitted to hospital for phototherapy; this reduced to 1.8 % after recalibration.

In summary the quite modest reduction in calibrator value had major clinical consequence, effectively halving the number of babies requiring phototherapy. The authors estimate the cost to Kaiser Permanente of overestimating serum bilirubin during the 2 years prior to recalibration, to be a staggering $8 million.

The same assay is used at 380 other hospitals across the US, so this $8 million figure is a gross underestimate of the total cost to US healthcare. The authors of the study also highlight the non-financial costs, including the unnecessary separation of mother and baby in the days after birth that is inevitably associated with photothe.