lurkingforacure and moondaughter make a good point: it is the individual that matters. If a therapy truly works for person X, that fact is not changed if it doesn't work for person Y or, indeed, everyone else. Seeing the therapy work for other people just increases our estimation of the accuracy of our assertion that it truly works for X.

We have a partial ordering of therapies:
- works for all the people, all the time;
- works for some of the people, all the time;
- works for all the people, some of the time;
- works for some of the people, some of the time.

As if that wasn't complicated enough, throw in the distinction between a therapy that gives symptomatic improvement and one that slows disease progression. Throw in pragmatics, such as cost, genetics, PwP preparedness to take part in clinical trials, the placebo effect, the sometime non-publication of adverse results, and the accuracy and resolution of the measurement process. Mix all these up, and it's no surprise that progress is so slow.

John