Context: Thyroid function testing is commonly based on measurement of TSH first, then reflex testing for free T4 if TSH is outside reference range limits. The utility of different TSH cut-offs for reflex testing is unknown.
Objective: To examine different TSH cut-offs for reflex testing of free T4.
Design, setting and patients: We analyzed concurrent TSH and free T4 results from 120 403 individuals from a single laboratory in Western Australia (clinical cohort) and 4568 Busselton Health Study participants (community cohort).
Results: In the clinical cohort, restricting measurement of free T4 to individuals with TSH below 0.3 or above 5.0 mU/L resulted in a 22% reduction in free T4 testing compared with TSH reference range limits of 0.4 and 4.0 mU/L; using TSH cut-offs of 0.2 and 6.0 mU/L resulted in a 34% reduction in free T4 testing. In the community cohort, the impact was less: 3.3% and 4.8% reductions in free T4 testing respectively. Using TSH cut-offs of 0.2 and 6.0 mU/L, elevated free T4 would go undetected in 4.2% of individuals with TSH 0.2-0.4 mU/L; in most, free T4 was marginally elevated, and unlikely to indicate overt hyperthyroidism. Low free T4 would go undetected in 2.5% of individuals with TSH 4-6 mU/L; in 94%, free T4 was marginally reduced and unlikely to indicate overt hypothyroidism.
Conclusion: Setting TSH cut-offs 0.1-0.2 mU/L below and 1-2 mU/L above reference range limits for reflex testing of free T4 reduces the need for free T4 testing, with minimal impact on case-finding.