Optimization of established and assessment of novel newborn screening strategies in the Netherlands

Newborn screening (NBS) is an important method for prevention of long-term health consequences or premature death in newborns with inherited disorders. A challenge in NBS is to find a balance between disease detection and false-positive referrals by choosing appropriate screening markers and cut-off values (COVs). In our studies, we performed systematic reviews and prospective and retrospective studies to collect Dutch NBS data for evaluation and optimization of NBS strategies.
We proposed to perform the new GALK deficiency NBS with total galactose as a primary marker with a COV of ≥ 2500 µmol/L blood, to ensure detection of GALK deficiency patients and minimize false-positive referrals. Furthermore, for maple syrup urine disease NBS (positive predictive value (PPV) 3.3%), we recommend a strategy with the current leucine and valine COVs of ≥ 340 µmol/L blood as a first-tier test and the leucine/phenylalanine ratio with a COV of ≥ 5.0 as a second-tier to improve the PPV to 19%. In congenital hypothyroidism NBS (PPV 21%), 55% of false-positive referrals were caused by the thyroxine/thyroxine-binding globulin (TBG) ratio. To prevent false-positives due to (partial) TBG deficiency, we recommended against referral when TBG concentrations were above 105 nmol/L blood. Alternatively, we showed that random forest models have the potential to improve the PPV to 26%. In NBS for congenital adrenal hyperplasia, we recommended the addition of the 21-deoxycortisol second-tier test to all inconclusive first heel puncture results in order to reduce false-positives by 67% and remove the need for a second heel puncture.