The impact of glucocorticoid initiation on the diagnostic accuracy of ultrasound
Published Date: 01st June 2020
Publication Authors: Sze JT, Dawson J
Background
Temporal and axillary artery ultrasound (US) has been recommended by EULAR as the first-line investigation in patients with suspected giant cell arteritis (GCA).1 US is reported as having a 77% sensitivity and 96% specificity for GCA.1 However, these figures have largely been derived from studies carried out in specialist centres where US was performed rapidly following the onset of symptoms. When performed by highly experienced sonographers, halos can still be detected on US weeks after initiation of glucocorticoid treatment.2 Little is known about the relationship between the dose and duration of glucocorticoid and diagnostic accuracy of US in real-world experience.
Objectives
We evaluated the impact of glucocorticoid initiation on the diagnostic accuracy of US in patients with suspected GCA in routine clinical practice in a district general hospital.
Methods
This is a single-centre retrospective study of all temporal and axillary artery US performed since its inception in November 2015 until October 2019. Patients who were aged ≥50 years and assessed by a rheumatologist were included in the study. US was performed by either a musculoskeletal consultant radiologist, musculoskeletal sonographer or vascular sonographer. US was considered positive for GCA when a halo, occlusion or stenosis was seen. Patients’ medical records and investigation results were reviewed in a systematic manner. The reference standard for GCA was the final clinical diagnosis after a minimum of 3-month rheumatology follow up.
Results
311 US performed on 305 patients were included. 62% of the scans were requested by rheumatologists, the rest by ophthalmologists and general physicians. 57 of these episodes had a final clinical diagnosis of GCA. US had an overall sensitivity of 39% and specificity of 100% for GCA. Overall positive and negative predictive values were 100% and 88%, respectively. Sensitivity was 31% for US done in the first 2 years (n=160) which was lower than sensitivity of 45% in the latter 2 years (n=151) (p=0.2663). Specificity remained the same in the two periods. When performed on patients who were not on any glucocorticoids, US had a sensitivity of 89% which was significantly higher than sensitivity of 29% in those who had been treated with any dose or duration of steroids (p=0.0007).
Sze, JT; Dawson, J. (2020). THU0542 The impact of glucocorticoid initiation on the diagnostic accuracy of ultrasound in giant cell arteritis experiences from a district general hospital in the UK. Annals of the Rheumatic Diseases. 79 (Supp 1), 511.