To the Editor:
I agree with the premise for routine screening that Drs. George A. Macones and Alison Cahill set forth in their November 2003 article, “Is routine sampling of fetal fibronectin justified?” However, I disagree with their discussion as it applies to fetal fibronectin.
One must be able to properly diagnose a disease before effective treatment options can be considered. For example, the initial screening test for cervical cancer was first described by Dr. Papanicolaou in 1943, but it was more than 40 years before the Bethesda System uniformly defined the disease, and only recently did we come to understand the association with human papillomavirus and develop the ability to test for high-risk subtypes. A similar comparison can be made with HIV testing: The screening was developed ahead of our ability to provide proper medical care.
In 2002, 12% of deliveries in the United States were premature.1 This certainly qualifies as a significant burden—one of the 5 requirements for routine screening given in the article. Second, the fetal fibronectin test has been shown to be highly sensitive, with a negative predictive value of 99%.2 Third, it is easy to perform, at a relatively inexpensive cost (roughly $125 per test). Fourth, the test is safe and acceptable to patients. Fifth, treatment is available for patients who test positive—namely, tocolysis, which can delay delivery for at least 48 hours, sometimes even a week. This allows time for treatment with antenatal steroids to reduce infant morbidity and mortality and, if necessary, for transfer of the mother to a center able to care for the preterm infant.
Finally, proper identification of women at higher risk of preterm delivery allows for continued study of more effective diagnostic tools and treatment options.
THOMAS A. RASKAUSKAS, MD
Drs. Macones and Cahill Respond:
Dr. Raskauskas argues that treatment (tocolysis) is available for women who are positive for fetal fibronectin. It seems that he is confusing a screening test (used in asymptomatic subjects) with a diagnostic test (used in subjects with symptoms).
At this point, the use of fetal fibronectin (or ultrasound measures of the cervix) for screening asymptomatic women in clinical practice is not justified. There is debate, however, on the merits of fetal fibronectin testing in women with symptoms of preterm labor. In this setting, most argue that a negative test is more clinically useful than a positive test. Still, some believe that fetal fibronectin may have a role in such cases. As stated in our article, we do not use this test in our center.
Dr. Raskauskas believes there is intrinsic benefit to developing screening tests for preterm birth, and we agree. However, we feel strongly that the implementation of new tests in clinical practice should not occur until their benefit is demonstrated. Obstetrics is full of examples of how our zeal for the “newest” test/device has resulted in widespread use before clinical advantage is proven. Two classic examples are electronic fetal heart rate monitoring (responsible for tripling the cesarean rate without demonstrable benefit) and home uterine activity monitoring (shown to be of no benefit in rigorously done studies).
We should learn from the mistakes of the past rather than repeat them, and introduce new tests/therapies only when clear evidence shows that their implementation will lead to improved clinical outcomes. It is with this approach that we will provide the safest and most effective care.