Hi Tony,
If the nebula extends beyond the length of the slit and you want to produce a flux calibrated spectrum then yes, you need to take a separate sky background spectrum image of the sky near the nebula under as near as possible the same conditions and with the same exposure time. I would then preprocess them all with bias, darks and (halogen)flats as usual then subtract the nebula image from the sky image to produce a final image ready for extracting the spectrum profile as usual (without any background subtraction). The instrument response calculated from a nearby reference star is applied in the normal way.
Cheers
Robin