Objectives: To review the overall performance of a neonatal screening program for cystic fibrosis (CF) from 1981 to 1994, and to compare two strategies of case detection.
Program design: Initially, immunoreactive trypsin (IRT) was measured in dried blood spots, and because of the low sensitivity of this test at days 3 to 5, a second sample was needed from babies with positive test results. Since 1993 a positive IRT assay result has been followed by direct gene analysis for the common CF mutation, delta F508, with the use of the same sample. Cases with false-negative results were actively sought throughout the period.
Results: With IRT alone, 1,015,000 babies were tested. Of 389 babies with CF, 30 had a clinical diagnosis of CF made after a negative screening test result or an administrative error. Early diagnosis was achieved in 92%. With the IRT/DNA protocol, 59 of 62 infants had a positive screening test result (44 were homozygous for delta F508) among 189,000 babies tested. Three babies with CF had no copy of this mutation, but two were identified early because of meconium ileus. The false-positive rate was much greater for IRT alone than for the IRT/DNA test (0.69% vs 0.054%). All false-positive cases in the IRT/DNA protocol were, of necessity, CF carriers.
Conclusion: The percentage of babies with CF who had an early diagnosis was similar with the two protocols, but we concluded that the advantages of the IRT/DNA test for screening, particularly in the avoidance of the need for second IRT samples, outweighed the drawback of unwanted carrier detection.