Skip to main content
. 2021 Jul 12;8(1):e000528. doi: 10.1136/lupus-2021-000528

Figure 3.

Figure 3

Survival analysis (M32). (Panels A and B) Kaplan-Meier time-to-event curves for assignment of the M32 International Classification of Diseases, Tenth Revision (ICD-10) section over time. Curves show the per cent probability of assignment of the M32 section after T0 for the four study groups (tier-1 (multianalyte assay panel (MAP)(Tier1+)), high tier-2 (MAP(High+)), low tier-2 (MAP(Low+)), MAP negative (MAP(−)), together with anti-double-stranded DNA positive (anti-dsDNA(+)) and negative (anti-dsDNA(−)) patients throughout the study. The x-axis reports the number of days since T0. Panels A and B report the same data analysis, with panel B allowing better visualisation of the initial portion of the survival curves. (Panel C) Cox proportional hazard model comparing the MAP score (Neg.: negative; Low: low tier-2; High: high tier-2, Tier 1: tier-1 positive) versus anti-dsDNA antibodies for assignment of the M32 section. In each quadrant, the numerators represent the number of subjects that were assigned an M32 section after T0 (events, n=51 in total) while the denominators represent all subjects in that quadrant. Concordance and p value of the likelihood ratio test are also reported. For data analysis, we used the date of the visit when the M32 section was recorded in the ICD-10 list.