The statement from spinger on this is here. I can think of some ways to fix the particular issue that caused the retractions:

• Not allow author-suggested reviewers
• More stringent process for author-suggested reviewers compared to non author-suggested reviewers
• Only allow author-suggested reviwers if they are for reviewers that are registered in some global database so that people can't create fake contact details for other people. There would need to be some method for authentication before you can register a user.

I would guess that any solution to the larger issue of scientific misconduct would need to consider Goodhart’s law and work to eliminate opportunites for people to game the system. There is a site called retractionwatch which has information on retractions that have occured.