ABSTRACT
Thyroid tests are done in acutely ill patients who often have confusing transient thyroid abnormalities, despite a lack of clarity about intervention and cost benefit. A retrospective analysis of patients admitted to the Medical Assessment Unit (MAU) was undertaken in 2004 to assess the frequency and utility, pattern of abnormalities and cost of thyroid testing. Guidelines were issued and the audit was repeated in 2008. 53.8% of 1593 subjects were offered thyroid tests in 2004 with a significant reduction to 21.7% of 1176 in 2008 (p<0.001). Free T4 or TSH was outside the reference range in 11.2% (2004) and 7.5% (2008) (p=0.10) and low TSH (52.7% in 2004 and 64.3% in 2008) was commonly combined with normal free T4. Appropriate indications for testing were documented in 43.9 vs 73.7% of patients with abnormal thyroid results (p=0.004) and in 14.3 vs 16% (2004 vs 2008) of a random sample of subjects with normal thyroid results, respectively (p=0.77). Documentation of intervention (25.5. vs 92.9%; p=0.001) and follow-up (45.5 vs 85.7%; p=0.001) had also improved significantly in 2008. We have demonstrated a significant reduction in thyroid testing in acutely ill patients after audit and the issue of guidelines. We currently recommend thyroid tests only in those with previous thyroid disease, the presence of clinical features and risk factors for thyroid disease, the use of relevant drugs, and unexplained tachydysrhythmias. The difficulties in interpreting results, the lack of clarity about intervention and follow up and possible cost savings would argue against an unrestricted policy.
