What is it about?

Background: Giant cell arteritis (GCA) is an autoimmune inflammation of the blood vessels in the elderly. GCA can be a difficult diagnosis especially when it does not present with typical symptoms, signs and abnormal bloodwork. Temporal artery biopsy is the conventional standard confirmatory test for giant cell arteritis (GCA), but is an invasive procedure. Ultrasound and MRI are other methods to investigate GCA. What we did: Mathematical prediction "formulas" for GCA were made using ten variables: age, gender, new onset headache, jaw muscle discomfort with chewing, temporal artery tenderness or pulselessness, vision loss, double vision, erythrocyte sedimentation rate, C-reactive protein and platelet levels. Using the data from 1,201 patients in whom the result of the biopsy was known, we developed a calculator that can estimate the risk of GCA prior to temporal artery biopsy, ultrasound or MRI. The link to the calculator is https://goo.gl/THCnuU

Featured Image

Why is it important?

Humans have difficulty assigning appropriate weights ("significance") to multiple patient variables. Although no mathematical model is infallible, the models can objectively determine the probability score for GCA with more accuracy than most clinicians. In addition to the pretest calculator there are posttest probability tables for ultrasound, MRI and a negative temporal artery biopsy.


The mathematical models (logistic regression and neural networks) help clinicians and patients decide the likelihood of giant cell arteritis (GCA) but decisions must be individualized for each case. In this dataset of 1,201 patients, 99% of patients with GCA had a probability threshold of at least 7%. However, patients are unique. Also, since the symptoms and signs of giant cell arteritis can overlap with many other diseases the calculator cannot be completely accurate. Notwithstanding, the mathematical models are objective and accurate "card counters" when you are forced to gamble in the "casino of giant cell arteritis. This pretest calculator can be combined with the results of ultrasound, MRI or negative temporal artery biopsy to determine the post test probability of GCA. It is important to use the calculator in conjunction with your doctor's clinical opinion.

Dr Edsel B Ing
University of Toronto

Read the Original

This page is a summary of: Neural network and logistic regression diagnostic prediction models for giant cell arteritis: development and validation, Clinical Ophthalmology, February 2019, Dove Medical Press,
DOI: 10.2147/opth.s193460.
You can read the full text:

Open access logo



The following have contributed to this page