Practice Variation in the Care of Subclinical Hypothyroidism During Pregnancy

A National Survey of Physicians in the United States

Freddy J.K. Toloza; Naykky M. Singh Ospina; Rene Rodriguez-Gutierrez; Derek T. O'Keeffe; Juan P. Brito; Victor M. Montori; Spyridoula Maraka


J Endo Soc. 2019;3(10):1892-1906. 

In This Article


Demographics of Respondents

Of the 5914 survey invitations sent by e-mail, 5911 were successfully delivered and 1562 (26%) were opened. We received a total of 162 responses (10%), of which 147 (91%) came from physicians who have participated in the care of pregnant women with SCH (screening question). The demographic characteristics of the respondents are summarized in Table 1. Respondents had practiced for an average of 18 years (IQR, 9 to 28 years) and had evaluated ~6 (IQR, 3 to 10) pregnant women with SCH over the past 6 months.

ATA Guideline Adherence

About 76%, 70%, and 18% of respondents had reviewed guidelines by the ATA, the Endocrine Society, and the ACOG, respectively. Only 53% of the respondents who had reviewed the ATA guidelines thought the recommendations had changed their practice.

The concordance between survey respondents' current clinical practices and the ATA recommendations (ATA guideline adherence) is summarized in Table 2. Overall, we did not find a correlation between guideline adherence and the guideline-reported strength of the recommendations [P trend = 0.66] or quality of the evidence supporting the recommendations [P trend = 0.31]. However, when analyzing by recommendation topics (SCH diagnosis/treatment/follow-up), we found a significant correlation between guideline adherence and the guideline-reported strength of the recommendations related to treatment [P trend = 0.01], but not for those related to diagnostic evaluation or therapy follow-up.

In a multivariate analysis, the number of years in clinical practice was the only significant predictor of guideline adherence [nβ = −0.23; P = 0.008] after adjustments for geographic location, specialty, the number of pregnant women with SCH treated over the past 6 months, previous reading of ATA guidelines, and type of clinical practice.

Screening and Diagnostic Evaluation

Most respondents recommended screening for thyroid dysfunction for every woman at the beginning of her pregnancy (54%), whereas 30% recommended targeted screening and 16% recommended no screening for SCH in pregnancy (Fig. 1). Survey findings regarding diagnostic evaluation of pregnant women with SCH are summarized in Table 3.

Figure 1.

Screening approaches for thyroid dysfunction during pregnancy according to survey respondents.

For the diagnosis of SCH, most respondents endorsed a TSH level >2.5 mIU/L as the cutoff (52%), whereas only 5% endorsed a population-based cutoff as recommended by the ATA. Others required different thresholds depending on the presence of clinical features (5%) or TPO-Ab status (5%). The most frequent indication for measurements of FT4 and TPO-Ab during the initial diagnostic workup was a TSH level higher than the pregnancy-specific cutoff used in the responders' clinical practice (49% and 44%, respectively).

Clinical Scenarios

Table 4 summarizes the treatment decisions among respondents for each clinical scenario and the expected risk reductions in adverse pregnancy outcomes and adverse health/cognitive outcomes of the offspring if treatment was provided.

When one-variable changes in a patient's clinical characteristics were assessed in clinical scenarios, changes in TPO-Ab status [87% for positive vs 50% for negative; P < 0.001] and thyroid physical examination results [62% for goiter vs 50% for normal; P < 0.001] significantly increased LT4 prescription rates for a first-trimester pregnant woman with TSH level = 4.4 mIU/L. In the case of first-trimester pregnant women with TSH level = 3.2 mIU/L, a change in TPO-Ab status increased LT4 prescription rates as well [57% for positive vs 18% for negative; P < 0.001]. The clinical scenarios regarding a second-trimester pregnant woman showed similar results.

In a multivariate analysis, TSH level (2.5 to 4.0 mIU/L vs >4.0 mIU/L), TPO-Ab status (positive vs negative), physical examination findings (normal vs presence of small goiter), and pregnancy trimester (first vs second trimester) were all significant predictors for starting LT4 therapy throughout the clinical scenarios. The strongest predictor was TPO-Ab status [nβ = 0.35; P < 0.001], followed by TSH level [nβ = 0.31; P < 0.001], physical examination findings [nβ = 0.1; P = 0.004], and trimester of diagnosis [nβ = 0.07; P = 0.003].

More than 70% of the clinicians who would start LT4 thought that the treatment would have a small effect (10% to 20% reduction) or very small effect (<10% reduction) on maternofetal complications regardless of the clinical scenario.

Treatment and Follow-up

Survey findings regarding treatment decision and follow-up of pregnant women with SCH are summarized in Table 3. The preferred therapy for the management of SCH during pregnancy was LT4 (97%), using an initial fixed dose of 25 to 50 μg/d (71%) or 75 to 100 μg/d (8%). A few respondents selected a dose based on TSH level (12%) or the patient's weight (7%). The factors considered when deciding whether to start therapy in a pregnant woman with SCH are shown in Fig. 2.

Figure 2.

Factors considered by survey respondents for treatment initiation in women with SCH during pregnancy.

According to most respondents (65%), TSH levels should be rechecked every 4 to 6 weeks until midgestation; 18% would recheck TSH levels only if the TSH level was not appropriate at first check after LT4 initiation, and 2% would never reassess TSH levels during pregnancy. Most respondents followed ATA guidelines and endorsed a TSH goal for thyroid hormone therapy of <2.5 mIU/L (74%) during pregnancy or in the lower half of the trimester-specific reference range (12%).

Stopping Therapy

Respondents would stop LT4 therapy after delivery if postpartum TSH levels fell within normal limits for a nonpregnant adult (35%) of if the patient required an LT4 dose <50 μg daily (18%). Some would stop LT4 in every woman (17%), whereas others would stop it in all TPO-Ab–negative women (13%) (Table 3).


With clinical experience (in years or in numbers of women treated), respondents expressed greater self-confidence in the care of pregnant women with SCH (Fig. 3). Endocrinologists showed a higher self-confidence level than other clinicians, and internal medicine specialists had the lowest self-confidence level for the management of SCH in pregnancy.

Figure 3.

Self-confidence regarding screening, treatment and follow-up, and stopping therapy for SCH during pregnancy as reported by survey respondents.