What you experienced is called the ceiling and floor effect. A ceiling effect occurs when you choose the highest answer possible for a question, but then you get better and need a higher number to choose from.
The same is true at the opposite end (the floor effect). The lowest number on the scale seems to describe you best at first, but later your symptoms or function have changed enough that you could pick an even lower number to describe yourself.
The problem isn't with you: it's with the tool being used to measure improvement. The survey you took wasn't able to discriminate accurately enough. There may be a better questionnaire available. Some questionnaires are meant to be general. Others are more disease-specific.
Your doctor may have what he or she is looking for, but it would be a good idea to raise this question the next time you are in the doctor's office. He or she may not be aware there is a potential problem that could be solved with a different (possibly better) survey.