Safety and efficacy data in clinical trials are mostly analyzed separately. However, especially the treatment of life-threatening disease such as cancer requires a good understanding of benefit and associated risks to make an informed therapy decision for an individual patient. Recently approved immunotherapeutic drugs in oncology are associated with potential side effects such as immune-related hypothyroidism, rash and colitis. There is some biological reasoning that the occurrence of immune-related adverse events and corresponding management may compromise the drug response. On the other hand, it has been observed that patients responding to treatment might face a higher likelihood of adverse drug reactions. A multi-state model is able to explore these hypotheses and offers the opportunity of insights into potential associations while addressing some of the methodological challenges. For example, the necessity of a time-dependent approach to accommodate the fact that safety and efficacy events can occur throughout the treatment. Moreover, longer treatment duration can impact simultaneously the likelihood of efficacy as well as safety events, i.e., introducing immortal time bias. The multistate model is able to unfold this spurious correlation. We present an approach for analysis and exemplify the methodology with simulated data.