Background/Purpose: Presenting signs/symptoms of GCA can be nonspecific
and patients can initially present to a spectrum of clinical specialties.  While temporal artery biopsy (TAB) is
recommended for confirming a GCA diagnosis, many patients without GCA are
referred for TAB and started on high-dose steroids while awaiting the
procedure, which can result in toxicity. 
Several studies have evaluated clinical predictors of positive TAB among
patients with suspected GCA but none have evaluated if the specialty of the
provider referring for TAB affects the likelihood of a patient ultimately being
diagnosed with GCA.

Methods: All TAB performed at a tertiary care center over 5
years were identified.  Medical
records were reviewed; those without adequate clinical information were excluded.
Whether a patient was referred for TAB by a rheumatologist, internist,
ophthalmologist, or neurologist was noted.  Subjects were categorized according to 1990 ACR criteria and
TAB status as having GCA or no GCA: those with GCA and negative TAB met ³3 ACR
criteria and were treated as GCA for at least 6mths based on reviewing chart
following TAB.

 Results: Of 188
subjects analyzed, 52 (27.7%) ultimately were recognized as having GCA.  23 (12.2%) had positive TAB and
clinical GCA, and 29 (15.4%) had negative TAB but were clinically diagnosed as
GCA.   There were a similar
number of referrals from each of the 4 specialties. Rheumatologists were
responsible for referring 28/52 (53.8%) of cases ultimately diagnosed as GCA
for TAB.  Compared to all other specialties,
rheumatologists had the highest odds of referring a case ultimately deemed to
have GCA for TAB. Patients referred by
rheumatologists had 5.43 (95% CI: 2.70, 10.99) times the odds of having GCA
compared to patients referred by internists, ophthalmologists, or neurologists
(p < 0.001).  Rheumatologists
were also more likely to have referred a patient with a positive TAB than other
specialists OR 3.41 (95% CI 1.40-8.33) p=0.007.  Using rheumatologists as the reference group,
patients referred by internists, neurologists or ophthalmologists all had a
lower odds of being diagnosed with GCA 
(Table 1)   Adjusting
for presenting symptoms (PMR, vision loss, headache) that may cause referral
bias, patients referred for TAB by rheumatologists remained more likely than
those referred by other specialists to have GCA, OR 5.49(95% CI 2.60-11.63),
p<0.001.     

Conclusion: In this cohort, patients referred for TAB by
rheumatologists had a significantly higher likelihood of having GCA or positive
TAB compared to those referred by all other specialties.   These data may support more
wide-spread implementation of fast-track pathways, where feasible, which would permit a rheumatologic evaluation for all patients
with suspected GCA prior to TAB.

Table 1: Odds of GCA Diagnosis
By Referring Specialty

*Odds of having GCA compared to
patients referred by rheumatologists