Anyway, even easier since the OP seems to be dealing a lot with the time between samples, would be to just make sure that at least 2.5 ms have passed before reading the sensor again. Right?
[1] Not sure if older blog posts detail the hardware, I do remember reading about this product on here before.