Cost-Effectiveness Analysis Of Repeat FNA For Thyroid Biopsies Read As Atypia Of Undetermined Significance

Abstract

Introduction: The 2007 National Cancer Institute (NCI) Thyroid Fine Needle Aspiration (FNA) State of the Science Conference proposed a 6 tiered classification scheme for thyroid FNA. The indeterminant category was partially replaced by a new category “Atypia of Undetermined Significance” (AUS). The malignancy rate for AUS FNAs should be between 5 and 10%. In the absence of risk factors or radiographic features of thyroid cancer, repeat FNA in 3-6 months is recommended. We found that from 2000-2009, 28% of thyroid nodule FNAs at our institution initially read as AUS had no repeat FNA and no surgical resection, 43% went directly to surgery, and 29% had repeat FNA. 53% of nodules initially categorized as AUS were benign on repeat FNA. 34% remained AUS on repeat FNA, and the malignancy rate in this group was 21%. We hypothesized that repeat FNA would be more cost-effective than diagnostic lobectomy after an initial AUS FNA result, and compliance with repeat FNA recommendations would be a cost-saving measure and improve cancer detection rates. Methods: Cost-effectiveness analysis was performed comparing diagnostic lobectomy with repeat FNA for lesions initially classified as AUS. A Markov model was developed based on a reference case scenario of a 40 year old with a solitary nonspecific nodule read as AUS by FNA. A third-party payer perspective was used. Treatment outcomes and probabilities for each strategy were identified based on literature review. Costs were estimated in 2009 US dollars using Medicare charge and reimbursement data and the US Nationwide Inpatient Sample. It was assumed that total thyroidectomy or completion thyroidectomy would be performed for thyroid carcinoma. The cost of lifelong thyroid hormone was included. Outcomes were weighted using established quality of life utility factors in order to yield quality-adjusted life years (QALYs) as a measure of effectiveness. All future costs and QALYs were subjected to a 3% annual discount rate. Sensitivity analysis was used to examine the uncertainty of probability, cost, and utility estimates in the model. Results: The diagnostic lobectomy strategy cost $6,140 and produced 23.971 QALYs. Repeat FNA for AUS cost $1,225 and produced 24.042 QALYs, making it the dominant strategy by demonstrating cost savings of $4,915 as well as a gain of 0.071 QALYs compared to the diagnostic lobectomy strategy. Repeat FNA remained the dominant strategy until the cost of FNA rose to $5,270 and continued to be cost-effective until the cost of FNA exceeded $8,000. Dominance of the repeat FNA strategy was not sensitive to the cost of surgery or changes in the rate of surgical complications on one-way sensitivity analysis. Conclusion: This study confirms that the recommendations from the NCI conference regarding AUS are cost-effective. Improving compliance with these guidelines should lead to lower overall costs, higher quality of life, fewer unnecessary thyroid resections, and fewer undiagnosed thyroid malignancies.