Table 2.

Effect of treatment delay on PROMs during the 4–10 years follow-up period: results of the multivariate mixed models regression

A: Binarised exposure (late (2–4 years) vs early (0–2 years)) DMT initiation
Outcome	Early mean score (95% CI)	Late mean score (95% CI)	Coefficient (late vs early) (95% CI)	P value
EQ-5D VAS	71.63 (67.02 to 76.56)	66.62 (60.89 to 72.35)	0.93 (0.85 to 1.01)	0.089
MSIS physical	17.78 (13.39 to 23.60)	23.29 (19.38 to 28.09)	1.31 (1.09 to 1.58)	0.004
MSIS psychological	26.41 (20.71 to 33.68)	30.11 (25.62 to 35.39)	1.14 (0.97 to 1.34)	0.102
EDSS	1.27 (1.12 to 1.42)	1.62 (1.37 to 1.86)	0.35 (0.10 to 0.59)	0.005
B: Continuous exposure (years from MS onset to first DMT)
Outcome	Reference score (95% CI)	Change in score per year of DMT delay (95% CI)	Coefficient per year of DMT delay (95% CI)	P value
EQ-5D VAS	72.55 (67.58 to 77.90)	−2.18 (−4.35 to 0.07)	0.97 (0.94 to 1.01)	0.146
MSIS physical	16.15 (12.11 to 21.55)	2.75 (1.29 to 4.20)	1.17 (1.08 to 1.26)	<0.001
MSIS psychological	25.19 (19.67 to 32.27)	2.02 (0.03 to 3.78)	1.08 (1.01 to 1.15)	0.024
EDSS	1.17 (0.99 to 1.35)	0.15 (0.05 to 0.25)	0.15 (0.05 to 0.25)	0.004

Each outcome was adjusted for disease duration; coefficients not shown.

Higher values indicate more severe disease burden in EDSS and MSIS; higher values indicate better quality of life in EQ5D-VAS2.

Coefficients are additive for EDSS (values >0 indicate higher scores).

DMT, disease-modifying therapy; EDSS, Expanded Disability Status Scale; EQ-5D, EuroQol-5 Dimensions; MS, multiple sclerosis; MSIS, Multiple Sclerosis Impact Scale; PROMs, patient-reported outcome measures; VAS, Visual Analogue Scale.