Screening for phenylketonuria in New York City. Threshold values reconsidered.

In New York City from 1966 to 1970, almost all 736,469 newborns were screened for phenylketonuria (PKU). Among 1,094 infants with presumptive positive test results, 763 were followed up and 46 of them were judged to require preventive treatment. The considerable annual variation observed in the frequency of values of 4 to 6 mg/100 ml bood phenylalanine suggested low reliability at this level. Screening test results of 4 mg/100 ml, with no cases detected among them, represented 53% of all false positive results; newborns with 6 mg/100 ml results yielded 1 infant in need of treatment and accounted for 40% of the false positive results. The large volume of presumptive positive results generated by these levels presumably contributed to incomplete followup. If the threshold value for followup were raised, the effectiveness and the efficiency of the screening program could be improved. The experiences of other large PKU programs in the United States support these observations.