Background Internationally, stakeholder involvement is recognised as adding value throughout the Health Technology Assessment (HTA) process. Stakeholders in seven countries took part in a European project, “INTEGRATE-HTA” which developed methods for more integrated, patient-centred approaches to HTA for complex interventions. The methods were applied in a case study evaluating models for palliative care service delivery, informed by patients, carers, practitioners and commissioners of health care.

Methods The initial plan was to set up combined lay and professional panels and use consistent evidence-based approaches to lay and professional involvement across all countries and stages of the project. At an early stage this proved not to be feasible as philosophies and views of the best methods for stakeholder involvement differ across Europe. Therefore, local co-ordinators adapted approaches to take account of the differences in context, usual practice, and ethical requirements between countries. We also adapted methods used for different stages of the project, which ranged from scoping, identifying and assessing evidence, to decision-making and discerning appropriate key messages for different stakeholder groups.

Results One hundred and thirty-two stakeholders (82 professionals and 50 lay people) in seven European countries (England, Germany, Italy, Lithuania, the Netherlands, Norway and Poland) were consulted at the outset of the project. Each country co-ordinator devised and implemented consultation methods with stakeholders as deemed appropriate locally, either using a qualitative research approach or consulting stakeholders as research advisors. In some countries, individual, face-face interviews were conducted with patients and analysed thematically. In others, participatory methods used group meetings. Subsequent input from additional stakeholders was elicited using a wide range of methods, including face-to-face, telephone and Skype interactions, either using a formal topic guide or less structured discussions. Some methods involved both lay and professional stakeholders discussing issues collectively. Other consultations were conducted with separate groups where it was felt combining groups would inhibit free expression of views. Adopting a wide range of methods ensured we could still get comparable and relevant input from across different countries.

Conclusion Using consistent methods of stakeholder involvement across different countries and settings, at different stages in a research project proved not to be feasible, acceptable, appropriate or efficient. All methods of stakeholder involvement have advantages and limitations which must be considered in light of local philosophies and views about such activities. Whilst stakeholder involvement in research may require the application of underpinning core standards and common principles, careful tailoring of approaches and flexibility is also required.