Neural network and logistic regression diagnostic prediction models for giant cell arteritis: development and validation

• Published in: Clinical ophthalmology (Auckland, N.Z.) Vol. 13; pp. 421 - 430
• Main Authors: Ing, Edsel B, Miller, Neil R, Nguyen, Angeline, Su, Wanhua, Bursztyn, Lulu L C D, Poole, Meredith, Kansal, Vinay, Toren, Andrew, Albreki, Dana, Mouhanna, Jack G, Muladzanov, Alla, Bernier, Mikaël, Gans, Mark, Lee, Dongho, Wendel, Colten, Sheldon, Claire, Shields, Marc, Bellan, Lorne, Lee-Wing, Matthew, Mohadjer, Yasaman, Nijhawan, Navdeep, Tyndel, Felix, Sundaram, Arun N E, Ten Hove, Martin W, Chen, John J, Rodriguez, Amadeo R, Hu, Angela, Khalidi, Nader, Ing, Royce, Wong, Samuel W K, Torun, Nurhan
• Format: Journal Article
• Language: English
• Published: New Zealand Dove Medical Press Limited 01.01.2019; Taylor & Francis Ltd; Dove Medical Press
• Subjects: Algorithms; Analysis; Autoimmune diseases; Biopsy; C-reactive protein; Giant cell arteritis; Information management; logistic regression; Medical centers; neural network; Neural networks; Ophthalmology; Original Research; Patients; prediction models; temporal artery biopsy; Variables; Vein & artery diseases; Veins & arteries; Canada; Ontario; Montreal Quebec Canada; United States > US; Baltimore Maryland; giant cell arteritis; prediction models; rheumatology; ophthalmology; logistic regression; neural network; temporal artery biopsy
• ISSN: 1177-5467; 1177-5483; 1177-5483
• DOI: 10.2147/OPTH.S193460

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. 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. 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 ( ≤0.05). The area under the receiver operating characteristic curve/Hosmer-Lemeshow 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 ( =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. 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).