(ii) Non-state led services: the state provides financial support to non-governmental rehabilitation services; or (iii) A hybrid model: integration of non-governmental specialized rehabilitation services into the national health infrastructure. The aim of the discussion on rehabilitation services and funding models was to share good practices in the field of national rehabilitation services and funding models, as recommended in the outcome report of the OSCE workshop in October 2016. The discussion was framed around the following key question: based on existing systems and structures (including the legal framework) and the in-need population, what model (state-led, non-state or hybrid) would work best to ensure rehabilitation is available, accessible and appropriate? Could the existing public-health system and budget be used as a starting point for establishing specialized rehabilitation services? At the beginning of the working group discussion, experts explained two key principles that should apply to rehabilitation services regardless of the funding model in place. First, civil society participation must include the active participation of NGOs and associations providing services to survivors on the local, regional and national level. GOs must be free to assist survivors regardless of the will of the latter to co-operate or not with the competent authorities. Their access to funding must not be dependent on the collaboration between the assistance and the law enforcement agencies. Second, a multidisciplinary and cross-sectoral approach to rehabilitation services means that a wide range of actors with distinct knowledge and expertise should be systematically involved in the design and implementation of any offer provided to torture survivors. Clear responsibilities and mandates must be defined to ensure smooth cooperation according to agreed procedures and quality standards of assistance. Participants shared their national experiences with the aim of having an open and constructive discussion addressing the challenges of the different approaches. During the discussion, it became clear that participants were very interested in enhancing their understanding of the different rehabilitation models and how they function in different contexts. The main conclusion of the discussion was that the best model for a country very much depends on the national context, including the structure of the public health care system, financing systems and tradition for cooperation with NGOs. During the discussion, it was underlined that none of the models were considered to be the golden standard as they all entail advantages and challenges. Participants agreed that State-led services had the advantage that they usually have more stable and long-term funding. However, some victims of torture might feel reluctant to make use of the public health care system because of a lack of trust in state authorities. Another challenge could be related to limited access due to complicated application procedures. Participants pointed out that funded but non-State-led services had the advantage that sustainable funding is provided by the State, while the rehabilitation services are provided by specialised NGOs who can respond to the specific needs on the ground. Some participants pointed out that NGOs are often more sensitive to the challenges that victims of torture may be facing, while a challenge with the model is to fully keep track of what is being implemented on the ground. 15