Advocacy by nonprofit scientific institutions needs to be evidence-based: a case study

Scientific institutions, including universities and research centers, occasionally engage in advocacy to gain financial support. However, this can be problematic if they selectively present scientific evidence. We describe a case involving a French semi-public foundation dedicated to clinical research on four adult psychiatric disorders: autism without intellectual disability, treatment-resistant depression, schizophrenia, and bipolar disorder. The foundation has claimed that an initial assessment at any of its Expert Centers led to a 50 % reduction in hospitalization days. We analyzed communication directed at the public in the French press, advocacy efforts towards members of the French Parliament (MFPs), evidence supporting this claim within those activities, and MFPs' initiatives that addressed the foundation's request. However, the reduction in hospitalization originated from a single study of bipolar disorder patients, which lacked a control group and had other methodological flaws. No scientific publications supported similar claims for the other three disorders. On May 2, 2024, 70 members of the French Parliament introduced a bill aimed at integrating these Expert Centers into the healthcare system. Justifications for the bill explicitly cited the 50 % reduction in hospitalization for all four conditions. This case highlights the need for policy makers and journalists to verify the robustness of scientific claims before these become policy. It also emphasizes the responsibility of scientists and journal editors to recognize and mitigate spin in research studies and opinion articles and to develop tools that help evaluate advocacy and lobbying claims in scientific contexts.