The preceding papers1-7 in this symposium discuss criteria for evaluating screening tests and screening programs and demonstrate how these criteria can be applied to several screening techniques of particular current interest. This paper will briefly review how a number of procedures measure up to the suggested criteria for screening tests and suggest some future directions which must be pursued if screening is to be based on science rather than on polemic.
WHERE ARE WE NOW?
Table I lists a number of diseases, functions and tests which have been considered to be of use in screening. For each of these it also gives my own estimate of how well each disease meets the criteria listed by Frankenburg2 and how well the available tests meet his criteria for effectiveness, direct costs, costs of false positives, and costs of false negatives. In the following paragraphs some of the considerations which led to these judgments are discussed.
Phenyketonuria
Despite its low frequency, the cost effectiveness of screening for PKU is well demonstrated, tests are adequately sensitive and specific, and with expert management the potentially high costs of false-positive tests can largely be eliminated.3
Galactosemia
Though even rarer than PKU, current methods make routine neonatal screening both feasible and cost effective, provided that the test is combined with routine PKU screening.8
Other Inborn Metabolic Errors
With the kind of regionalized facilities and planning discussed by Scriver,3 testing for these diseases can be added to routine neonatal screening at very little additional cost. Some of them lend themselves to effective early treatment; with others, genetic counsulling is the only currently available useful intervention.