@valeri I think the sample size is too small to subdivide it further. The plan they mention on the website was to have two trials, one in the UK for tinnitus, one in the US for hearing loss. Evidently they couldn't do both trials at this time, so this hearing loss requirement allows them (hopefully) to go to their investors saying, "Our drug helps tinnitus sufferers and those with moderate hearing loss."
One of the things that has to be accounted for in a clinical trial is people dropping out for various reasons. If you start with a sample size of 50 people, and 15 drop out, you've only got 35 cases you can point at and say, "The drug works." And what if it doesn't work in all cases, because as you pointed out, tinnitus is a complex condition? Say it works in 25 cases. Is a major pharmaceutical company going to invest millions in a drug that helped 25 people? The best you could hope for in that scenario is that they'd fund another study.
With a sample size of 152 persons, even if half drop out and a third of those remaining don't get clinical benefit, you're still better off than the three way split you propose.
I think one of the points of contention here is that there doesn't seem to be a definite connection between tinnitus and hearing loss. But that objection misses the point. AUT-00063 was designed to help people with tinnitus and people with hearing loss from the same source. The tinnitus is coming from hypoactive neurons and the hearing loss is coming from hypoactive neurons. That's the theory and what the drug is meant to combat. So the best way to test that is to find people with tinnitus and hearing loss. If it helps those people you've proved all three points of your theory:
1) Tinnitus is generated when certain neurons in the auditory brain become hypoactive
2) When these neurons become hypoactive, hearing loss is often a symptom
3) AUT-00063 improves the functioning of these neurons, lessening or eliminating tinnitus and improving hearing comprehension
That last point contains a clarification worth making. AUT-00063 doesn't regrow cochlear hair cells, so there is a ceiling to what it can do for hearing improvement. However, part of Autifony's theory is that a good portion of hearing loss is actually a failure in signal processing, which is a function the drug can improve.
Understand that I'm not really arguing in favor of Autifony's decision so much as trying to understand it. It'd be much easier to find people to report back on their experiences if hearing loss were not a criterion. But I expect a good part of the reasoning came down to economic reality. There may well be concern at some levels that the market share for AUT-00063 may not meet expectations. This could be based on the belief that since many people "just live with" their tinnitus, they are not guaranteed to seek out the cure even when it is widely available. None of us think that. But the decision-makers at GSK who do not having personal experience of the condition may do.