Is it ok to use z-scores means for inter-sample comparison?

I have scRNA-seq from 2 patient. I want to show particular genes’ enrichment in both patients and compare them. So, first I integrated the datasets to be able to compare them. Then I want to compute the mean z-scores of gene A in the cells of patient 1 and the mean z-scores of gene A in the cells of patient 2. Cause in this case z-scores means for each patient won’t be 0, right? Will that make sense from the statistical point of view?