When you measure anything, there's an error. I've read BG monitors may have errors up to 15%.
If you make multiple measures, take the average, or median.
Or do only one measure and assume the real value is somewhere close to the reading (up to 15% up, or down).