I wrote a few months ago about how research in differential privacy seems very applicable to global public health. There is a new report from the Institute of Medicine which calls for a new approach to protecting privacy in health research, Beyond the HIPAA privacy rule. The Lancet also has an editorial about the report, which is what made me think that I should pass this reference on to theoretical privacy researchers. Lancet’s summary:
The IOM recommends that the Department of Health and Human Services (HHS) develops a completely new framework for protecting privacy. The new framework would facilitate greater use of anonymous data and include legal sanctions to prohibit unauthorised identification. If policy makers choose to keep HIPAA, then the IOM calls on HHS to make several revisions to the Rule to reduce misinterpretation and improve the usefulness of data sources for research. These revisions include the promotion of best practices for privacy protection in research and the development of guidance that clearly states that individuals can authorise use of personally identifiable data for future research.
I haven’t read the report, but I see that it mentions differential privacy once, and references a short survey by Cynthia Dwork.
Healthy Algorithms 
http://grants.nih.gov/grants/guide/rfa-files/RFA-OD-09-003.html#PartII