Severity of multiple sclerosis (MS) is typically assessed using the Expanded Disability Status Scale. This score can take values from 0 (no deficits) to 10 (= death due to MS) and a disease progression is defined as an increase in this score by at least one point. The primary endpoint in a study on progressive MS was time to disease progression, defined as the time between randomization and the first initial progression assessment (IPD) that was confirmed by the next scheduled assessment at least 12 weeks later. In this disease, confirmation of an IPD is key, in order to separate progressions from relapses. The latter are clinically quite different, tend to be of short duration, and typically allow patients to recover. In this talk, we will discuss a proposal for the construction of an estimand in this setup. A key difference of this example to other situations is that intercurrent events cannot only happen between between randomization and IPD, but also between IPD and the subsequent confirmation. This requires careful assessment of all potential event courses a patient can experience in order to define an unequivocal estimand.