To determine the diagnostic accuracy of the 'halo sign' on temporal artery ultrasound in people with suspected GCA, using the ACR 1990 Classification as a reference standard. We will consider two tests:

1. triage test before temporal artery biopsy, if temporal artery biopsy is considered: a sensitivity of at least 95% with a minimum specificity of 50% will be required to rule out GCA and avoid biopsy in patients who test positive; a specificity of 95% and a sensitivity of 50% will be required to rule in GCA and refer for biopsy patients who test negative.
2. replacement of temporal artery biopsy as one of the ACR 1990 criteria: this will be achieved if ultrasound shows sensitivity and specificity similar to biopsy.

We will also conduct an assessment of sources of heterogeneity at the study level, considering the following sources:

1. unilateral versus bilateral halo sign;
2. concurrent use versus no use of steroids;
3. different thresholds for the halo sign: 0.3 mm (primary cut-off), 1.5 mm or 6 mm; however we anticipate that, either studies will not have reported the threshold (present versus absent) or that they will have adopted 0.3 mm cut-off;
4. average age of participants;
5. training of ultrasonographers;
6. additional use of axillary artery ultrasound in conjunction with temporal artery evaluation.