Fig. 6

Nomograms for a routine and b candidate biomarker prediction models for estimation of the patient-specific risk of multiple sclerosis. Each of the nomograms can estimate a patient’s risk of multiple sclerosis (MS) diagnosis. Instructions for use (here exemplified for the “routine model”): Locate the patient’s IgG index value on the “IgG index” scale and find the corresponding point score straight above on the “Points” scale. Do this for the other axes and sum the total point score. Locate the total score on the “Total Points” axis and draw a line straight down on the “Risk for MS” axis: This is the patient’s estimated risk of MS diagnosis within the first year from onset of optic neuritis (ON). Risk of MS should be considered with an uncertainty of up to ± 10% for the “routine biomarker” model, and ± 15% for the “candidate biomarker” model according to calibration plots (Additional file 3: Figure S3). Example of patient from this study: a A patient had no OCBs, 3 leukocytes/μl, and an IgG index of 0.48, giving a score on the “routine biomarker” nomogram of 0, 1, and 9 points, respectively (10 in total). This corresponds to approximately 13% risk of developing MS. b This same patient had CSF measurements of 0.16 pg/ml IL-10, ≤ 10 pg/ml CXCL13, and 229 pg/ml NF-L. This scored 6, 0, and 2 points, respectively, and 18 in total on the “candidate biomarker” nomogram. Such a score corresponds to circa 19% risk of MS. This patient indeed had ION by the end of the follow-up period