abstract
- Background: Thyroid-stimulating hormone (TSH) is commonly ordered to investigate delirium. Previous studies have recommended against the routine testing of thyroid function in in-patients due to low clinical yield and excess use of resources. This has not been investigated in in-patients admitted for delirium specifically. We aimed to investigate the clinical yield of thyroid function testing in in-patients admitted for delirium. Methods: This was a retrospective study of 84 patients admitted to hospital with delirium identified as one of their in-hospital issues. Both patients with and without TSH tested in-hospital were included. The primary outcome was a composite of the proportion of patients with TSH tested who had further thyroid-related testing or medications ordered during their hospital stay. Secondary outcomes included number and types of thyroid-related symptoms documented during admission. Multivariate regression was performed to determine factors associated with ordering TSH and with the primary outcome. Results: TSH was ordered in 65 (77.4%) patients and the result was within reference range in 53 (81.5%) cases. Thyroid disease was identified as the etiology of delirium in zero cases. Only 3 (4.6%) patients had additional thyroid-related testing or medications ordered. Multivariate regression demonstrated a significant association between known hypothyroidism (P=0.001) and pre-existing prescription for levothyroxine (P=0.028) with the primary outcome. The average number of thyroid-related symptoms reported per patient was 1.095. Thirty-one (36.9%) patients had zero symptoms of thyroid disease documented. There was no significant difference in number of symptoms reported between patients who had and did not have TSH testing during admission (P=0.325). Conclusions: Thyroid function testing does not often result in a change in clinical management in the investigation of delirium in in-patients. Routine TSH testing in delirium without thyroid-related symptoms or when another cause of delirium is apparent may use resources excessively with low clinical yield.