That would not surprise me in the least. People underestimate how difficult it is to analyze results when said results depend on subjective measures -- in this case, a doctor's determination of how much improvement there is (or isn't).