Background: Guidelines in healthcare should be evidence-based, satisfy patient needs and improve patient outcome.

Methods: We performed a cost–utility analysis in Graves' disease (GD) and estimated incremental costs after the introduction of a national guideline adding the Graves' Recurrent Events After Therapy (GREAT) score with genetic determinants (GREAT+) to predict recurrence, a thyroid nurse, preoperative calcium/vitamin D treatment and thyroid-stimulating immunoglobulins.

Findings: Antithyroid drugs (ATDs) were less costly, achieved 0.88 quality-adjusted life years (QALYs) over 8 years and dominated over radioactive iodine (RAI) treatment. The relevant incremental cost-effectiveness ratio was ATD versus thyroid surgery (Tx). Tx was more costly than ATD but was also more effective. The incremental cost-effectiveness ratio was equal to 40,488 Euro per QALY gained. In recurrent GD, the QALY weight for surgery after ATD was 0.76 compared with 0.79 when surgery was the initial treatment. If individuals requiring surgery could be identified at start of first treatment, QALYs would be higher (6.32) and the cost lower (13,945 Euro). The net cost increase after the new guideline was 17.6%, which was partially an effect from more time being spent with the thyroid nurse. If the GREAT+ score was also applied, the total increased net cost was 14.8% if 24% of the tested patients changed treatment to Tx.

Interpretation: Tx was more cost-effective than RAI when ablative treatment is advocated. Prediction score for recurrence directing patients to earlier Tx is cost-effective and enables the introduction of a specialist thyroid nurse. Health economic evaluations should accompany future guidelines.