Is Routine Antenatal Toxoplasmosis Screening Justified in the United States? Statistical Considerations in the Application of Medical Screening Tests
Toxoplasmosis when acquired by the immunologically naïve woman during early pregnancy, although a rare event in the United States, can infect the fetus and cause devastating consequences.1 Recent advances in both immunologic and molecular biologic technologies have provided clinicians the ability to identify early pregnancy-associated maternal infection as well as the potential to reliably diagnose in utero infection.2,3 Because of the potentially severe fetal effects of toxoplasmosis and because there are therapies known to modify these effects, it is desirable that pregnancy-associated Toxoplasma infections be accurately identified. Not only is it important to avail treatment for the genuinely infected fetus identified by a "true positive" test, but it is also essential that unneccessary maternal anxiety, administration of potentially toxic antibiotics, or perhaps proceeding with interruption of pregnancy for "false positive" diagnoses be avoided.1,2
Because most acute Toxoplasma infections are relatively asymptomatic, the problem that obstetricians face is how to identify affected pregnancies. One strategy that has been discussed frequently, but never widely adopted in the United States, is that of routine population screening. Although the adoption of such a strategy intuitively seems the obvious solution to the problem of congenital toxoplasmosis, certain caveats must first be recognized and rigorously considered. These caveats include problems related to the diagnostic accuracy of available testing technologies and the attendant critical statistical considerations for such screening tests applied to large populations having variable disease prevalence.4,5
The purpose of this article is to provide an understanding of the mathematical underpinnings of screening tests used in medical practice. The focus of our discussion is the problem of congenital toxoplasmosis, a rare but often silent and devastating disease, and the effort to identify cases utilizing a pregnancy maternal serum screening strategy. We also provide hypothetical examples and a concrete example from the Papanicolaou (Pap) Test, a more familiar and highly regarded screening test used in clinical obstetrics and gynecology. To begin, we outline some basic notation and formulas that may simplify our later commentary.