The Expanded Disability Status Scale (EDSS) is an established measure of disability in multiple sclerosis (MS). Due to similarities in the clinical presentations of MS and neuromyelitis optica spectrum disorder (NMOSD), the EDSS is also widely used to assess disability in NMOSD but has yet to be validated for this purpose.
To establish the psychometric reliability, validity, and responsiveness of the EDSS in NMOSD patients.
Analyses were conducted in a pooled population of NMOSD patients (N=178) from the phase 3 SAkuraSky (NCT02028884) and SAkuraStar (NCT02073279) studies. EDSS was assessed at regular intervals. Reliability was evaluated using standardized Cronbach’s α and test/re-test reliability. Convergent validity was assessed by comparison with the EuroQol Visual Analog Scale (EQ-VAS) and relevant outputs from the Short Form-36 (SF-36) health survey (Physical Functioning [PF], Role-functioning Physical [RP], and Physical Component Summary [PCS] domain scores). Discriminant validity was assessed against the Visual Analogue Scale for Pain (VAS-pain), and non-physical domains of the SF-36 (Vitality [VT], Mental Health [MH], Role-Emotional [RE], and Mental Component Score [MCS]). Criterion validity was assessed by comparison with the modified Rankin Scale (mRS). Responsiveness of the EDSS to changes in health status was assessed through a relative validity (RV) comparison of EDSS scores in patients who experienced an investigator-reported clinical relapse vs those without.
Cronbach’s α coefficient was >0.6, suggesting reasonable internal consistency (α=0.67). The test/retest reliability coefficient was α=0.91, with scores >0.70 representing reasonable reliability. Assessment of convergent validity revealed moderate-to-strong correlations between EDSS and other measures of physical functioning (EQ-VAS, rs –0.53; SF-36 PF, rs –0.61; SF-36 RP, rs –0.58; SF-36 PCS, rs –0.60). The EDSS showed strong discriminant validity against VAS-pain and non-physical SF-36 domains (VAS-pain, rs 0.31; SF-36 VT, rs –0.35; SF-36 MH, rs –0.27; SF-36 RE, rs –0.37; SF-36 MCS, rs –0.25). Strong criterion validity was observed in relation to the mRS (rs 0.68). The EDSS was found to be responsive to investigator-reported relapses (F-statistic=36.64, p<0.0001; RV=1.0).
The EDSS demonstrated reliability, validity, and responsiveness as a measure of disability in patients with NMOSD. Further studies to corroborate these findings are warranted.