comparing multiple (12) Likkert-Items - Kruskal-Wallis-Test (?)

Dear all,

I already found useful information to calculate my tests however minor questions are left. I have a data-set of around 200 participants in 4 groups, unequally distributed. For every participant I have 12 Likert-Items (called A_219 to A_230)
To compare these Items with each other I decided to use the Kruskal-Wallis-Test, wich brings me to tests as e.g. A_219 with A_220 to all. This is comparing one Likert-Item with the other.

While comparing the answers of one group with the other I am going to first compare them all within the Kruskal Wallis-Test and then use rank.sum to compare two groups.

Is there any idea on calculate the overall differences between all twelve items A_219 - A_230 and then in the second step calculate rank.sum for two items? Or do I mix sth. up in here and this may simply not be possible?

Thanks for your assistance in advance!!!