Of course a between-groups comparison isn't actually testing this, but a small effect probably wouldn't yield highly sensitive/specific clinical discrimination from that measure alone. The question is whether your aims/inferences are actually attempting to address this. Probably not.
