Efficacy of statewide neonatal screening for cystic fibrosis by assay of trypsinogen concentrations.

To evaluate the feasibility and efficacy of measuring immunoreactive trypsinogen in blood to screen for cystic fibrosis, we performed this test in 279,399 newborns in Colorado from 1982 to 1987. Immunoreactive trypsinogen was measured in dried blood spots when the infants were 1 to 4 days old; if the level was elevated (greater than or equal to 140 micrograms per liter), the measurement was repeated (mean age, 38 days); if the level was again elevated, sweat testing was performed (mean age, 49 days). For the second test, two cutoff levels (120 and 80 micrograms per liter) were evaluated. We found an incidence of cystic fibrosis of 1 in 3827 (0.26 per 1000), with 3.2 newborns per 1000 requiring repeat measurement. When adjusted for race and compliance with testing, the incidence among the white infants (1 in 2521) was close to the expected incidence. The false positive rate with the initial cutoff level (92.2 percent) was similar to the rate found in neonatal screening programs for other diseases. False negative results occurred because of laboratory error or changes in procedure (three infants) and trypsinogen concentrations lower than the initial cutoff level (three infants) or lower than the remeasurement cutoff level of 120 micrograms per liter (one infant). Sweat tests were negative in 168 infants with an elevated initial trypsinogen level but a level below 80 micrograms per liter on remeasurement, confirming the value of 80 micrograms per liter as an appropriate cutoff for repeat-test results. Overall, 95.2 percent of the infants with cystic fibrosis (95 percent confidence interval, 85 to 99 percent) who did not have meconium ileus could be identified with the use of a trypsinogen cutoff level of 140 micrograms per liter on initial testing and 80 micrograms per liter on repeat testing. Statewide screening for cystic fibrosis based on measurements of immunoreactive trypsinogen in dried blood spots is feasible and can be implemented with acceptable rates of repeat testing and false positive and false negative results.