Thyroid hormone testing is widely used in primary care; however, many asymptomatic patients are tested, leading to the overtreatment of subclinical disease. Hypothyroidism occurs in 4.6% of the US population (0.3% of cases are clinical, and 4.3% are subclinical).^1–3 Biochemical hyperthyroidism occurs in 0.5% to 1.3% of adults in the United States.^1–3 Levothyroxine is the third most prescribed medication in the United States, with 6.0% of the population taking it.⁴ One study found that in 1 year, 18% to 25% of patients had thyroid function testing.⁵ There is a disconnect between prevalence of the disease and testing and treatment.

Unnecessary levothyroxine treatment of subclinical disease may lead to increased fractures and cardiac arrhythmias.⁶ Discordant test results between thyroid-stimulating hormone (TSH; thyrotropin) and thyroxine (T₄) or triiodothyronine (T₃) levels can lead to unnecessary follow-up, patient anxiety, and increased health care costs without clinical benefit. The US Preventive Services Task Force states that there is insufficient evidence for screening asymptomatic, nonpregnant adults for thyroid disease.⁷ Given the risks of unnecessary treatment and overtreatment, it is critical to use a patient-centered approach. When should thyroid function tests be performed, and when should they be avoided?

For patients with symptoms of thyroid disease (eg, fatigue, depression, difficulty concentrating, memory impairment, weight gain or loss, constipation, dry skin, voice changes, cold or heat intolerance, weakness, palpitations), and when there is clinical suspicion for thyroid disease, testing for TSH levels is preferred initially.^1,2 If TSH levels are elevated, free T₄ levels should be obtained.¹ If TSH levels are below normal limits (overly suppressed), free T₄ and total T₃ levels should be obtained for diagnosis of hyperthyroidism.²

Universal screening for thyroid disease during pregnancy is not recommended because treating subclinical hypothyroidism does not improve pregnancy outcomes. Screening select patients is recommended due to the effects on maternal outcomes and fetal development. Untreated thyroid disease can result in preeclampsia, preterm delivery, low birth weight, abnormal neuropsychologic development in the infant, and fetal demise.⁸ Measuring TSH levels is also recommended as the first-line test for pregnant patients with a personal or family history of thyroid disorders, type 1 diabetes, or a clinical concern for thyroid disease. In pregnant patients with hypothyroidism, TSH levels should be maintained between the trimester-specific lower limit and 2.5 mIU/L, and TSH measurement should be obtained every 4 to 6 weeks to ensure appropriate levels of levothyroxine.⁸

Other causes of abnormal TSH levels, such as medication use, sick euthyroid syndrome, and pituitary or hypothalamic disease, should be considered before the diagnosis of primary thyroid disease.⁹ In a pooled analysis of 2,335 older adults in Europe, 60.8% of those with subclinical hypothyroidism experienced spontaneous normalization of thyroid function within 1 year.¹⁰ Additionally, biotin supplements should be held for 2 days before measuring TSH levels because biotin can suppress TSH, leading to false-positive findings for hyperthyroidism.²

When treating primary hypothyroidism, levothyroxine dosing should be adjusted based on normalizing TSH and not free T₄ levels. Treatment of subclinical hypothyroidism is recommended if TSH is greater than 10 mIU/L or thyroid peroxidase antibody is elevated.¹ A previous American Family Physician article discusses dosing recommendations for hypothyroidism.¹

Subclinical hyperthyroidism is defined by suppressed TSH and normal free T₄ levels. TSH should be retested in 3 to 6 months to confirm subclinical hyperthyroidism before starting treatment. If a patient is at high risk of complications from untreated subclinical hyperthyroidism, TSH and free T₄ should be retested within 2 to 6 weeks.⁹ Because hyperthyroidism is associated with an increased risk of cardiovascular disease, atrial fibrillation, heart failure, and decreased bone mineral density, the American Thyroid Association recommends treating subclinical hyperthyroidism if TSH levels are less than 0.1 mIU/L in two groups: adults 65 years and older and adults younger than 65 years who have heart disease, osteoporosis, or symptomatic hyperthyroidism.^2,9

Clinicians should follow a guideline-based approach for the evaluation of thyroid disease. Routinely performing full thyroid panels should be avoided. Testing for TSH levels should be performed initially, with reflex testing when indicated. Screening should be avoided in asymptomatic patients. When treating primary thyroid disease, TSH levels should be used to guide dosing. By reassessing how we approach thyroid function testing, we can reduce overtreatment, minimize unnecessary health care costs, and improve patient outcomes.