Working on databases we should always be aware of the possibility of a re-identification attack on the data we release in our publications. Such possibilities are in fact very rare because we aggregate all our data in the articles we submit to publishers. Nevertheless the odds that our data could be re-identified are to be known. Mathematicians have brilliantly addressed the issue in the article referenced below:
Title: A Game Theoretic Framework for Analyzing Re-Identification Risk
Authors: Zhiyu Wan, Yevgeniy Vorobeychik, Weiyi Xia, Ellen Wright Clayton, Murat Kantarcioglu, Ranjit Ganta, Raymond Heatherly, Bradley A. Malin
Clearly the balance has to be found between on one hand the fines paid for the consequences of privacy violation and on the other hand the gain for authors and publishers in terms of respectively grants,notoriety,transparency, and audience for releasing accurate data and practicing data sharing which is what research is heading toward.