Frequency, determinants and costs of thyroid function testing in a laboratory serving a large population

To assess the rationale and frequency of thyroid function testing and to analyse factors that influence serum thyrotropin (TSH) levels.Serum TSH levels were evaluated in a hospital laboratory serving a population of 604 000 in 2018. Patients on medications or with conditions affecting thyroid function were excluded. Frequency of thyroid function testing by age and sex was assessed and the relationship between serum TSH with potential predictor variables was analysed using ordinary least square regression analysis allowing for potential non-linearity.Twenty-eight percent of the local population had their thyroid function tested at least once in 2018 with significant differences by sex (28.2% women vs 23.4% men) and by age groups, with less than 2% of <16-year-old people and more than 50% of >80-year-old people being tested. Most of the symptoms commonly attributed to thyroid dysfunction were not higher in the thyroid dysfunction groups. Serum TSH levels were higher in older people particularly after the age of 60 years, in women (by 0.1 mIU/L), during the early hours of the morning, and in winter and spring seasons. There was remarkable uniformity in the frequency of subclinical thyroid dysfunction, as well as substantial cost savings, if TSH reference intervals were recalculated across sexes, age groups, time-periods and seasons.Serum TSH is frequently tested in the population but is not a good discriminant of symptoms attributed to thyroid dysfunction. Furthermore, considering the influence of factors on TSH reference limits could significantly impact patient care and resource utilisation.