A study published in the BMJ [see here] found that 21% of papers published in 2008 in 6 major medical journals had guest or ghost authors. This is a slight decrease since a similar study in 1996 found 29% but still a cause for concern. Guest authorship appears to be a particular problem in research articles.
COPE Chair, Liz Wager, has used COPE cases to show the problems editors sometimes face when they try to work with institutions on cases of suspected misconduct. The report has been published online in the BMJ this week. It is available to BMJ subscribers here and on the COPE website here.