When the distribution is normal, the z-score is a measure of the the number of standard deviations a value is from the mean.
Given a measured value, one has to standardise it by converting it into a z-score. This allows for comparison of different results from different studies.