Annual Disability-State Transition Probabilities Estimated from a Large Sample (-6000) of Australians with Multiple Sclerosis from MS-Base

OBJECTIVES: Multiple sclerosis (MS) is a chronic autoimmune/neurodegenerative disease associated with progressing disability mostly diagnosed between 20-40 years of age in females (approximately 75%). We aimed to estimate transition probabilities from no disability through to severe disability and death. Transition probabilities are a vital input for health economics models that are used to inform healthcare resourcing decisions including for MS-specific disease-modifying therapies (DMT).

METHODS: Data were obtained from Australian participants of MSBase, the largest international registry that collects observational data as part of routine clinical care. We used a five-state continuous-time Markov model to describe how people with MS transition between disability milestones defined by the Expanded Disability Status Scale (scale 0-10); no disability (EDSS=0), mild (EDSS=1-3.5), moderate (EDSS=4-6), severe (EDSS=6.5-9.5), and dead (EDSS=10), all recorded by the treating neurologist. Model covariates included sex, DMT usage, MS-phenotype (relapsing-remitting (RR)MS or secondary-progressive (SP)MS), and disease duration. Subgroup analyses were performed.

RESULTS: N=6,369 participants (mean age 42.5 years, 75.0% female) entered the study, yielding 38,994 person-years of observation (mean (standard deviation) 6.1 (4.5) years) and 54,699 clinical reviews (mean 8.5 reviews). Holding all covariates at their mean values, one year transition probabilities included: remaining disability-free (54.0% [95%CI: 43.6%, 56.2%]); remaining in mild state (82.0% [60.4%, 82.6%]); from the no to mild state (42.1% [40.1%, 43.9%]); mild to moderate (11.4% [10.9%, 11.9%]; moderate to severe (9.5% [8.8%, 10.3%]); severe to death (0.5% [0.3%, 0.7%]). Hazard ratios showed that SPMS was associated with a 148.2% (95%CI:120.2%, 180.0%) increase and a 66.8% (71.3%, 61.5%) decrease versus RRMS in the hazard of disability progression and recovery, respectively.

CONCLUSIONS: MS progresses quickly from the no disability state, particularly with the SPMS phenotype. Estimated transition probabilities will be applied in a health economics simulation model for Australia, intended to support reimbursement of interventions including medications to reduce progression of MS in Australia