Dealing with increasingly smaller patient groups

Teaching ourselves how to assess personalized health technologies

Duration: October 2012 - October 2013.

For personalized or stratified medicine, various societal partners must assess health technologies for very small groups or even individuals. Methodology and statistical approaches for personalized medicine are evolving, but this is not a smooth process. This may be due to a knowledge gap and/or to conflicting values when trying to quantify medical uncertainty at the individual- and at the group-level. It will take many years, and the involvement of many stakeholders, to align scientific evidence for personalized medicine in Europe and to adapt our regulatory systems.  

The project aimed to add knowledge about how personalized medicine is being implemented, by identifying current bottlenecks and describing solutions proposed by stakeholders. A second aim was to stimulate information exchange and discussion among stakeholders who are concerned with health technology assessment in the context of personalized medicine, i.e. medicine for small groups of patients, such as the pharmaceutical industry, drug regulators, the health insurance board, physicians, statisticians, patient representatives and journalists. 

The researcher collaborated with academic researchers, clinicians, pharmaceutical industries and patient organizations representing nine institutions in five countries. Interviews with 10 Dutch stakeholders showed that insufficient evidence, regulations in the Netherlands, containment of high costs and reimbursement of health technologies were barriers to the implementation of personalized medicine. Diverse solutions were mentioned; notably an overarching solution of early collaboration between stakeholders. An invitational symposium provided educational lectures on the suitability of Bayesian statistics for small groups, reimbursement policy in relation to personalized medicine, and regulatory issues in personalized medicine. Focus groups discussed solutions for implementing personalized medicine and 13 recommendations were distilled.

The project shows that, while stakeholders are able to suggest solutions to existing bottlenecks, conflicting values prevent their immediate incorporation into policy. Salient examples deserve further dissemination (education, filling knowledge gaps) and discussion (identifying conflicts as a prerequisite to new consensus). Such activities can provide input for European Commission’s ongoing efforts to foster innovative research and policies aimed at getting the best out of personalized medicine.