The datasheet seems clear enough to me. You read the calibration constants from the chip
using the I2C interface when your process commences, and then you use the
complicated formula to calculate the actual pressure from the measured pressure and
temperature on each calculation cycle.
It looks very similar to the process for the other 3 pressure chips in common use.