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

Materials and Methods

Survey Design

Two authors (S.M. and F.J.K.T.) prepared an initial draft of the questionnaire according to the study objectives and previously issued surveys in the field.[18–20] The survey included demographic data (specialty, geographical location, years of clinical practice, community type) and multiple-choice questions based on two clinical scenarios describing variations in TSH levels, thyroid autoimmunity status, and thyroid physical examination results to widely assess the diagnostic evaluation, decision on initiation of therapy, and follow-up in pregnant women with SCH. The survey used in this project is publicly shared in an online repository.[23] The main topics covered by the survey were screening, TSH diagnostic cutoff, use of TPO to guide therapy, types of therapy, and follow-up. Because we intended to assess current clinical practices, an initial screening question was added to exclude clinicians who do not care for pregnant women with SCH. Most questions required a single best response to be selected from multiple choices and were constructed to omit phrasing that could lead respondents to the "right" answer. Some questions allowed multiple items to be simultaneously selected. We limited questions to achieve a survey response time of less than 15 minutes. Subsequent survey drafts were distributed among the coauthors, and after an iterative process of feedback and discussion, a final version was prepared. There was an additional review process by the Endocrine Society Clinical Affairs Committee, which provided feedback and ensured survey relevance to its members. The study was considered exempt by the institutional review board of the University of Arkansas for Medical Sciences.

Survey Distribution and Data Collection

An anonymous online survey was sent to 5914 US medical doctors who are members of the Endocrine Society between 5 September and 16 November 2018. They received an e-mail invitation to participate from society administrators, which described the survey and contained an electronic link to the survey website without offering incentives to participate. Three reminders were sent after the first e-mail, each 2 to 3 weeks apart. Survey responses were anonymously collected and stored electronically by an online survey service (Google Forms, Mountain View, CA), and data were password protected. Repeated submissions from the same IP address were automatically blocked by the survey service. Only members of the Endocrine Society were surveyed because according to previous survey-based studies in thyroidology that included members of the ATA and the American Association of Clinical Endocrinologists, the majority of survey respondents came from the Endocrine Society. In addition, there was substantial overlap between the respondents' memberships, and a small percentage did not have Endocrine Society membership.[24–26] We also attempted to collaborate with the ACOG regarding distribution of the same survey to its members; however, we were unsuccessful.

Statistical Analysis

Summary statistics are presented as frequencies (percentages) for categorical variables and as means and SD or median and interquartile range (IQR) for continuous variables according to the normality of the variables. The response rate was estimated for each question. Statistical analyses explored the relationships between respondents' demographics and adherence to ATA guidelines or self-confidence level in the management of SCH in pregnant women. Differences in categorical variables were analyzed with the χ 2 or Fisher's exact test, and differences in continuous variables with the independent t test or Mann-Whitney test as appropriate. Simple linear regression was used to analyze correlations between guideline adherence and the guideline-reported strength of the recommendation or quality of evidence. ANOVA was used to assess differences between demographic characteristics of the respondents and self-confidence or adherence to ATA guidelines for the management of SCH during pregnancy. To assess the possible drivers of adherence to ATA guideline recommendations, a multivariate analysis adjusted for geographic location, specialty, years in clinical practice, number of pregnant women with SCH treated over the past 6 months, previous reading of ATA guidelines, and type of clinical practice was performed. All analyses were two-tailed, with α set at 0.05, and were conducted using IBM SPSS Statistics version 25.