Determinants for Thyroid Hormone Replacement Therapy in Subclinical Hypothyroidism: A Multicenter Electronic Health Records-Based Study.

Approximately one in five patients diagnosed with subclinical hypothyroidism received thyroid hormone replacement therapy.

• 796 total patients with SCH were included across four academic centers in the United States and Mexico
• 165 of 796 patients (20.7%) were treated with thyroid hormone replacement therapy
• The study period spanned January 1, 2016, to December 31, 2018
• 65.2% of the overall cohort were women

Female sex was independently associated with higher odds of receiving thyroid hormone replacement therapy for subclinical hypothyroidism.

• In multivariable logistic regression, female sex had an odds ratio of 1.71 (CI 1.13–2.59), p = 0.01
• 72.7% of the treated group were women compared with 63.2% of the untreated group (p = 0.03)
• The association persisted after adjusting for other covariates in the multivariable model

Higher index TSH level was independently associated with higher odds of receiving thyroid hormone replacement therapy.

• In multivariable logistic regression, index TSH level had an odds ratio of 1.97 (CI 1.56–2.49), p < 0.001 for every standard deviation (2.75 mIU/L) change
• This was the strongest independent predictor of treatment in the multivariable model
• TSH level was evaluated as a continuous variable with effect estimates reported per SD change

Treated patients with subclinical hypothyroidism were significantly younger than untreated patients.

• Mean age of treated group was 51.0 years (SD 18.3) vs. 55.3 years (SD 18.2) in the untreated group
• p = 0.008 for the age difference between groups

The majority of patients did not have confirmatory thyroid function tests before the treatment decision was made.

• Only 46.7% of patients in the treated group had confirmatory TFTs before the decision to start thyroid hormone replacement therapy
• 65.6% of patients in the untreated group had confirmatory TFTs before the decision not to treat was made
• This indicates that for both treated and untreated patients, decisions were frequently based on a single set of abnormal TFTs

Positive thyroid autoimmunity testing was more frequent in the treated group, but overall thyroid autoimmunity assessment was underused.

• Positive thyroid autoimmunity was present in 48.2% of the treated group vs. 20.3% of the untreated group (p < 0.001)
• There was no statistically significant difference in the frequency of thyroid autoimmunity evaluation between the two groups
• The authors concluded that thyroid autoimmunity assessment was underused in the overall SCH population