But an analysis of the more than 95,000 studies registered since 2004 suggests that missing data is still a problem, according to Robert Califf, MD, of the Duke Translational Medicine Institute in Durham, NC, and colleagues including Deborah Zarin, MD, the database director.