The steep rise of advanced technology in healthcare and the required spe- cialisation of staff causes healthcare systems to have become increasingly complex. This increased complexity poses great challenges on the risk management of systems. Many different methods have been developed the past decennia to identify potential risks in these system to enhance safety. A commonly used prospective risk analysis method is HFMEA, which is based on the linear view on system safety. Another relatively new method is FRAM which meets the dynamic systemic view on safety and focusses more on potential risks in a system due to "everyday performance". An example of a complex healthcare system is proton therapy; a novel type of radiotherapy to treat tumours in the proximity of the central nervous system. The first operational clinic in the Netherlands providing this therapy is HollandPTC. During the therapy the tumours are irradiated with a high precision in multiple sessions. When anatomical variation is observed between these sessions, the treatment plan of a patient has to be adjusted. This critical process is called plan-adaptation and has to be both time-efficient and safe. To ensure the safety of the plan-adaptation process at HollandPTC, currently controls are designed based on potential risks identified with HFMEA. The aim of this thesis is to identify an effective strategy to identify which controls are able to monitor and mitigate risks associated with the process of plan adaptation at HollandPTC. Independently of the HFMEA, a FRAM was conducted on the plan-adaptation process at HollandPTC. The resulting set of potential risks were compared with the po- tential risks identified with HFMEA. Furthermore, the effectiveness of the controls proposed by HFMEA were assessed on the set of potential risks identified with FRAM. Based on these results a strategy is proposed to monitor and mitigate the potential risks. The analysis of the FRAM models revealed among others potential risks related to: informal communication lines between caregivers, discrepancies between caregivers ideas about task division and identified multiple causes for time delays. These risks were not identified with HFMEA. The controls proposed by HFMEA do not mitigate the potential risks identified with FRAM. By combining the strengths of both HFMEA and FRAM, an effective strategy is proposed to monitor risk and to identify effective controls. This strategy can be used as a prospective risk-analysis method and on ongoing processes.