This 75-minute session will present new methodological research for the modeling of sequential treatment decisions. Actual clinical practice often involves sequential treatment decisions that create modeling considerations more challenging than those typically encountered in more standard single pairwise comparison models. For example, the sequential decisions make cohort Markov modeling very difficult because they introduce problems in developing the controlling equations to take proper account of the conditionalities involved. Also, propagating uncertainty and exploring its impact in PSA cannot follow standard approaches. There are also methods issues relating to estimating parameters that are appropriate to a place in the sequence or pathway. These research presentations will be more in-depth than typical research podium sessions and will include time for moderated audience discussion.