A Predictive Model for Temporal Artery Biopsy in the Setting of Suspected Giant Cell Arteritis: A Validation Study.

Abstract

A previously published predictive model based on threshold parameters for erythrocyte sedimentation rate, c-reactive protein, and platelet count demonstrated that 40% of patients who underwent biopsy may not have required it. The current study was performed to evaluate the model's performance on an independent data set. This is a retrospective consecutive series of patients undergoing temporal artery biopsy (TAB) in a single health region in Canada. The model was applied to a multicenter cohort of patients undergoing TAB by a variety of surgical services. A centralized pathological database serving multiple institutions and surgical services was used to identify patients undergoing TAB. Over a 7-year period, patients undergoing TAB were identified via a central pathological database. Those who had concurrent illnesses which would likely affect erythrocyte sedimentation rate, c-reactive protein, and platelet count, patients on steroids for >2 weeks by the time of biopsy, and those with missing serum markers were excluded. The previously developed model was applied to the 222 patients enrolled. The model correctly identified 29% of patients with a pretest probability of 0% for a positive biopsy and 9% with a pretest probability of 100%, suggesting that in total, 38% of patients could have avoided TAB. The results of this independent data set support the previously published predictive formula. Utilizing a simple, clinically applicable predictive model of the pretest probabilities, approximately 38% of TAB currently being performed may be avoided. The results suggest that evaluation with a prospective multicentre study would be appropriate.