OBJECTIVES: Reexamine cost-effectiveness of riluzole in the treatment of amyotrophic lateral sclerosis (ALS) in light of recent advances in disease staging and understanding of stage-specific drug effect.

METHODS: ALS was staged by the FT9 method. Stage-specific health utilities (EQ-5D, US valuation) were estimated from an institutional cohort, whereas costs and transition probabilities were informed by existing literature. Costs were disaggregated into recurring costs (RCs) and “one-off” transition/“tollgate” costs (TCs). Five- and 10-year horizons starting in stage 1 disease were examined from healthcare sector and societal perspectives using Markov models to evaluate riluzole use. Probabilistic and one-way deterministic sensitivity analyses were conducted.

RESULTS: Mean EQ-5D utilities for stages 0-4 were 0.79, 0.74, 0.63, 0.54, and 0.46, respectively. From the healthcare sector perspective at the 5-year horizon, riluzole use contributed to 0.182 QALY gained at the cost difference of $12,348 ($5,403 riluzole cost, $8,870 RC and -$1,925 TC differences), translating to an incremental cost effectiveness ratio (ICER) of $67,658/QALY gained. Variance of transition risk estimates contributed considerably to ICER uncertainty (- 30.2% to + 90.0%). ICER was sensitive to drug price and RCs, whereas higher TCs modestly reduced ICER due to delayed tollgates.

CONCLUSIONS: This study provides a framework for health economic studies of ALS treatments using FT9 staging. As mischaracterization of TCs as RCs would markedly inflate the ICER, prospective stage-specific and disaggregated cost measurement is warranted for accurate future cost-effectiveness analyses. Appropriate separation of TCs from RCs substantially mitigates the high burden of background cost of care on the ICER.