Abstract
The recent guidelines of the American Thyroid Association (ATA) (1), in agreement with the position of the Endocrine Society (2), reaffirm that universal screening of pregnant women for thyroid dysfunction is not necessary. At the same time, similar to the previous recommendations (2), the ATA defined a high-risk group that should be investigated. This group consists of pregnant women with current disease or a history of thyroid dysfunction, partial thyroidectomy, or neck radiotherapy, and pregnant women with signs or symptoms of hypothyroidism, goiter, autoimmune disease (including type 1 diabetes), or a family history of thyroid disease (1,2). It should be noted that these conditions are indications for the assessment of thyroid function even in the general population (3). A history of preterm delivery or abortion is also included as a risk factor (1,2). Adopting these criteria, 22% and 13% of pregnant women were classified as high risk in an Italian study involving 4657 patients (4) and in a Chinese study involving 2899 patients (5), respectively. Similarly, 220 (20.8%) of 1058 pregnant women evaluated by our group were excluded from a study (6) because they were classified as high risk. Adoption of the previous criteria (2) would result in the investigation of only a minority of pregnant women.
In contrast to the Endocrine Society (2), the ATA included age ≥30 years as a risk criterion (1). This inclusion deserves discussion. Considering that low-risk pregnant women (according to the definition of the Endocrine Society) present a mean age of ∼28 years (4 –6), it can be estimated that about 40% of these women are 30 years or older. This group (low risk) corresponds to 75%–85% of pregnant women (4 –6). Consequently, age alone [in the absence of other risk factors (2)] would result in the investigation of ∼30% of pregnant women, a percentage that exceeds all other indications together (4 –6). Or even, the inclusion of age as an indication for the assessment of thyroid function would be more than double the number of pregnant women investigated.
The argument to include age ≥30 years as a risk factor is that ∼7% of women in this age group would present thyrotropin (TSH) >5 mIU/L according to two recognized studies (1,7,8). In one of these studies (7), this rate refers to all women studied, not only those without any risk factor for thyroid dysfunction or those who already knew to have this condition; in these women, the frequency of elevated TSH certainly would be much lower. In the other study (8), TSH >5 mIU/L was observed in only 2% of women aged 30–39 years without known thyroid disease or goiter.
The concern that classical risk factors (2,3) fail to identify a good part of pregnant women with elevated TSH (4,5,9,10) is understandable, but the inclusion of age in the definition of high risk does not seem to minimize this problem. In the study of Vaidya et al. (9) and in our study (6), age was not a predictor of thyroid dysfunction in the low-risk group. Wang et al. (5) also did not report age to be a factor associated with altered TSH. In the series of Negro et al. (4), the chance of abnormal thyroid function (considering high- and low-risk women, not only the latter) only discretely increased after 29 years (2.8% at this age vs. 4.2% at 39 years), an increase that seems to be insufficient to include age ≥30 years as a risk factor.
What needs to be determined is whether the number of cases of thyroid dysfunction that are additionally diagnosed with the inclusion of age as a risk criterion justifies the relevant increase in the number of pregnant women investigated, that is, the cost effectiveness of this measure. We question whether this cost effectiveness would not be the same as that of universal screening, which reaches a sensitivity of 100%.
We conclude that among the extreme proposals of investigating only pregnant women with classical risk factors for thyroid dysfunction (2,3)—which shows low sensitivity—and of universal screening—which presents 100% sensitivity—the inclusion of age as a risk criterion does not seem to minimize the problem or to be justified.