Nobody has said that patient selection for scientific studies does not matter. Please reread the prior posts. Patient selection is important for studies into ME. So is the selection of controls, and decisions made about standards of testing. Selection bias is not so important in an OMI style database. Its the studies that determine criteria, create matched controls etc. Its the sad reality that doctors being unable to handle uncertainty in medicine and diagnosis that makes them want certainty. The only "certain" diagnosis for ME in many doctors eyes is a psychosomatic one. This is dangerous. Medicine needs to embrace uncertainty, or these mistakes will be driven by their need for certainty. which makes them vulnerable to latching on to half-baked theories such as that deconditioning, depression or the wrong ideas about their illness can cause ME, or actually be ME. Once they have the wrong ideas firmly entrenched, then work committments, confirmation bias and cognitive dissonance make looking at contrary data difficult. It gets dismissed. Why should they look at data about ME? They already know its psychosomatic, or not a real illness or whatever their view is. What needs to happen is that false hypotheses, false claims about ME, are disproved. Then we have to find a way to break through the political and economic structures that make current BPS models so appealing. Meanwhile studies like the phase 3 clinical trials into Rituximab may yet present us with a reliable cure for many of us, and a treatment for most. At that point we will need to push for treatment, and lobby for government support in every country and acceptance by the insurance industry. It was patient demand that got antibiotics accepted for use in peptic ulcers, the medical profession resisted it. Other studies include possible biomarkers, novel pathogens etc. Lots of interesting research may come out in the next few years.