Woolie
Senior Member
Yea, this is a really big problem. The way you define the sample in the first place affects the outcome hugely.It's not helped, of course, that clinicians aren't always consistent in how they diagnose the illness. CFS clinics here keep talking mostly about 'chronic fatigue' and 'fatigue spectrum illnesses', which means they're only really diagnosing fatigue. I noticed it both on my table at NICE and at the local CFS clinic, where the specialist specifically mentioned the 'fatigue spectrum illnesses' to me. I think that many doctors in the UK are treating CFS as a wastebasket diagnosis rather than as a positive diagnosis. Not always, but enough that it's a problem.
This is a problem in any field where the illness is is defined on the basis of symptoms and the cause unknown. A few researchers have recently been trying to tease apart the various features of depression by looking at which cluster together the most - the sadness, the lack of pleasure, the negative self-evaluations, etc. Seems like a good place to start, right? We can maybe work out whether there are distinct subtypes of depression or not. But these researchers only looked at people who already had a clinical diagnosis of depression, which means they would have needed to have several of these symptoms concurrently. The people who only had one or two, well, they never got into the sample.
So this method of selection will artificially increase the chances that symptoms occur together.