| Summary: | Edsel B Ing,1 Neil R Miller,2 Angeline Nguyen,2 Wanhua Su,3 Lulu LCD Bursztyn,4 Meredith Poole,5 Vinay Kansal,6 Andrew Toren,7 Dana Albreki,8 Jack G Mouhanna,9 Alla Muladzanov,10 Mikaël Bernier,11 Mark Gans,10 Dongho Lee,12 Colten Wendel,13 Claire Sheldon,13 Marc Shields,14 Lorne Bellan,15 Matthew Lee-Wing,15 Yasaman Mohadjer,16 Navdeep Nijhawan,1 Felix Tyndel,17 Arun NE Sundaram,17 Martin W ten Hove,18 John J Chen,19 Amadeo R Rodriguez,20 Angela Hu,21 Nader Khalidi,21 Royce Ing,22 Samuel WK Wong,23 Nurhan Torun24 1Ophthalmology, University of Toronto, Toronto, ON, Canada; 2Ophthalmology, Johns Hopkins University, Baltimore, MD, USA; 3Statistics, MacEwan University, Edmonton, AB, Canada; 4Ophthalmology, Western University, London, ON, Canada; 5Queens University, Kingston, ON, Canada; 6Ophthalmology, University of Saskatchewan, Saskatoon, SK, Canada; 7Laval University, Quebec, QC, Canada; 8Ophthalmology, University of Ottawa, Ottawa, ON, Canada; 9University of Ottawa, Ottawa, ON, Canada; 10Ophthalmology, McGill University, Montreal, QC, Canada; 11University of Sherbrooke, QC, Canada; 12University of British Columbia, Vancouver, BC, Canada; 13Ophthalmology, University of British Columbia, Vancouver, BC, Canada; 14Ophthalmology, University of Virginia, Fisherville, VA, USA; 15Ophthalmology, University of Manitoba, Winnipeg, MB, Canada; 16Ophthalmology, Eye Institute of West Florida, Tampa, FL, USA; 17Neurology, University of Toronto, Toronto, ON, Canada; 18Ophthalmology, Queens University, Toronto, ON, Canada; 19Ophthalmology & Neurology, Mayo Clinic, Rochester, MN, USA; 20Ophthalmology, McMaster University, Hamilton, ON, Canada; 21Rheumatology, McMaster University, Hamilton, ON, Canada; 22Undergraduate Science, Ryerson University, Toronto, ON, Canada; 23Statistics, University of Waterloo, Waterloo, ON, Canada; 24Ophthalmology, Harvard University, Boston, MA, USA Purpose: To develop and validate neural network (NN) vs logistic regression (LR) diagnostic prediction models in patients with suspected giant cell arteritis (GCA). Design: Multicenter retrospective chart review.Methods: An audit of consecutive patients undergoing temporal artery biopsy (TABx) for suspected GCA was conducted at 14 international medical centers. The outcome variable was biopsy-proven GCA. The predictor variables were age, gender, headache, clinical temporal artery abnormality, jaw claudication, vision loss, diplopia, erythrocyte sedimentation rate, C-reactive protein, and platelet level. The data were divided into three groups to train, validate, and test the models. The NN model with the lowest false-negative rate was chosen. Internal and external validations were performed.Results: Of 1,833 patients who underwent TABx, there was complete information on 1,201 patients, 300 (25%) of whom had a positive TABx. On multivariable LR age, platelets, jaw claudication, vision loss, log C-reactive protein, log erythrocyte sedimentation rate, headache, and clinical temporal artery abnormality were statistically significant predictors of a positive TABx (P≤0.05). The area under the receiver operating characteristic curve/Hosmer–Lemeshow P for LR was 0.867 (95% CI, 0.794, 0.917)/0.119 vs NN 0.860 (95% CI, 0.786, 0.911)/0.805, with no statistically significant difference of the area under the curves (P=0.316). The misclassification rate/false-negative rate of LR was 20.6%/47.5% vs 18.1%/30.5% for NN. Missing data analysis did not change the results.Conclusion: Statistical models can aid in the triage of patients with suspected GCA. Misclassification remains a concern, but cutoff values for 95% and 99% sensitivities are provided (https://goo.gl/THCnuU). Keywords: giant cell arteritis, temporal artery biopsy, neural network, logistic regression, prediction models, ophthalmology, rheumatology
|
|---|
