Early-stage cost-utility analysis of novel diagnostic tests for giant cell arteritis: a modelling study in UK...

To identify the key characteristics required for hypothetical diagnostic tests to be cost-effective for diagnosing giant cell arteritis (GCA).

Combined decision tree and Markov cohort state-transition models were used to evaluate the cost-utility of new diagnostic tests compared with the standard pathways of biopsy and clinical judgement, with and without ultrasound. Input parameters were derived from secondary data and expert opinions. The analysis adopted a lifetime horizon and the UK National Health Service (NHS) perspective, using a willingness-to-pay threshold of £20 000 per quality-adjusted life year (QALY). Bivariate deterministic sensitivity analyses identified the maximum test price at varying diagnostic performance levels, and probabilistic sensitivity analyses over 5000 simulations provided 95% CIs.

UK.

Patients with symptoms suggestive of GCA.

Percentage of GCA-related and glucocorticoid-related complications avoided, maximum test price and incremental QALYs at each sensitivity and specificity combination.

A biomarker test incorporated into a hypothetical diagnostic pathway with perfect accuracy (100% sensitivity and specificity) can be priced up to £7245 (95% CI £5763 to £8727) and remain cost-effective compared with a standard pathway of temporal artery biopsy and clinical judgement. Against a standard pathway including ultrasound, the biomarker test can be priced up to £8606 (£6741 to £10 471). The test's value was more strongly influenced by improvements in specificity than in sensitivity. The maximum prices decreased with earlier starting age, lower clinician adherence, shorter time horizons and shorter durations of glucocorticoid-related effects.

The study highlights the potential for hypothetical tests to improve GCA diagnosis and reduce glucocorticoid toxicity, while demonstrating their market viability for use within the NHS. It also illustrates how early-stage economic models can provide valuable insights into potential cost-effectiveness to inform the test development process.