I can't get over how happily they spin their results: http://content.karger.com/ProdukteD...howAbstract&ArtikelNr=313691&ProduktNr=223864 Background: Patients with medically unexplained or functional somatic symptoms (FSS) are prevalent in primary care. In this pragmatic cluster-randomised controlled trial we aimed to test the effect of a training programme (The Extended Reattribution and Management model) for general practitioners (GPs) in the treatment of FSS. Methods: 38 participating GPs were randomised to the control group or the training group. The GPs included consecutive 18- to 65-year-old patients presenting during a 3-week period for new health complaints. We assessed a stratified subsample with the psychiatric interview Schedules of Clinical Assessment in Neuropsychiatry. Of 701 patients interviewed, 350 fulfilled the diagnostic criteria for any ICD-10 somatoform disorder (SD) and 111 presented FSS without fulfilling these criteria (sub-threshold SD). Patients completed questionnaires at baseline and after 3, 12 and 24 months. The questionnaires included assessment of health status (36-item Medical Outcomes Study Short Form; SF-36), health anxiety (Whiteley-7) and physical symptoms (Symptom Check List-90, somatization subscale). Results: Patients with SD consulting trained GPs improved more on our primary outcome of physical functioning than patients consulting control GPs at the 3-month follow-up (p = 0.004), but the improvement was not statistically significant at later follow-up. We found no significant differences in improvement between patients with SD and those with sub-threshold SD. Results for other SF-36 subscales, physical symptoms and health anxiety only showed statistically significant differences between the intervention and control groups for patients with SD; patients consulting trained GPs had less improvement in vitality, health anxiety and physical symptoms at 24 months compared with the control group. Conclusions: GP training may accelerate improvement in physical functioning for patients with SD. However, the effect is small and may not be clinically significant.