Validating a diagnostic GCA ultrasonography service against temporal artery biopsy and long-term clinical outcomes

Currently, there is no mechanism for service validation of diagnostic ultrasonography (US) for giant cell arteritis (GCA). Temporal artery biopsy (TAB) and classification criteria are poor benchmarks. We validated our service against physician-verified diagnosis at 100 weeks (100wD). Twenty-five patients underwent US within 7 days, and TAB within 28 days, of commencing prednisolone. US, TAB and baseline diagnosis (bD) were all compared with 100wD using Cohen's kappa. Fourteen US and 8 TABs were positive. Twenty at baseline and 14 at 100 weeks had diagnosis of GCA. The kappa (95% CI) were 0.4 (0.1, 0.7) for US vs. TAB; 0.5 (0.2, 0.8) for US vs. bD and 0.2 (0.0, 0.4) for TAB vs. bD. Versus 100wD, the kappa (95% CI) were 0.8 (0.6, 1.0) for US; 0.4 (0.1, 0.7) for TAB and 0.6 (0.3, 0.9) for bD. Seven cases were US+/TAB-. Four had alternate confirmation: 18FDG-PET (n = 1), CT Aorta (n = 1) and US at relapse (n = 2). At 100 weeks, 4 cases (all US-/TAB-) with bD of GCA had alternative diagnoses including cancer (n = 2). This is the first study validating US service provision for GCA. Twenty-five US with a robust kappa on comparison with long-term diagnosis validates our service. A diagnosis of GCA should be made with extreme caution for US-/TAB- cases. Key Points • This is the first study offering a way to validate a new diagnostic US service by validation against TAB and long-term physician-verified diagnosis. • US has substantial to near-perfect agreement with long-term physician-verified diagnosis and is more reliable than TAB in our hands. • Alternative diagnoses should be sought in patients with dual negativity for US and TAB.