MARINA

No, we do not follow such a policy (it may be unique to your facility!).
To address this issue operationally, I use the AntigenPlus program. http://www.antigenplus.com/
This program standardizes the rule-out process (excluding antigens to which the patient may have antibody). This is only the first of the many steps required for antibody identification.
Everyone in my staff of generalists is required to use the program.

We use the 2 and 2 rule.
 
The way I understood the math is that 3 and 3 rule = 0.05 p value if you're running only 6 cells total. We use a three cell screen and an 11 cell first-line panel. You should be able to get an acceptable p value based on those numbers of cells tested. On that note, there's a lot of discussion about the relevance of p value in the first place.
 
Kanter, M. H., Poole, G. D. and Garratty, G. (1997), Misinterpretation and misapplication of p values in antibody identification: the lack of value of a p value. Transfusion, 37: 816–822.
 
We use one homozygous cell for rule outs, with some exceptions.