All you need to do is upsample the 16/44.1 source to 48khz and up the bit depth to 24-bit. You should be able to so this in Soundforge or Wavelab. Just make sure when you put the two pieces together that you have a matching point so there is no artifact in the transition.
I'd recommend applying a crossfade to smooth the transition from one source to the next. And you probably don't even need to convert the bitdepth. Just <1> SRC the patch source from 44.1kHz to 48kHz, <2> add the files to your multitrack editor, <3> align them, <4> apply crossfades, and <5> when generating the mixdown file, simply specify the proper output bit-depth (24) and sample rate (48k). The multitrack editor should simply pad the 16-bit source with zeroes for the least significant 8 bits.
The final file would like like this (since I can't resist ASCII):
24A = the first 24-bit file
PS = the patch source
24B = the second 24-bit file
Green = the audible portions of each file
Maroon = muted portions of the patch source
[----------24A----------]
[---------------------------PS---------------------------]
[----------24B----------]