According to this, Dragon Naturally Speaking can transcribe recorded speech from a sound file (this feature may only be available in the “pro” version - check your user manual).
Assuming that is correct, then you just need to export your recording from Audacity in a file format that Dragon Naturally Speaking supports. The export default format in Audacity is 16-bit WAV and I would suggest trying that.