I served on a committee of the National Research Council that studied the challenges posed to Computer Science (and computing in general) in the area of healthcare. The report was just released and can be found here.
Needless to say, healthcare is a fascinating area, and progress will require not only collaboration among multiple disciplines (within computer science and others), but also paying attention to the workflow and constraints of the industry itself (doctors' work habits, the way insurance works, or doesn't work, hospitals as businesses, etc).
When reading the report, keep in mind that the goal of the committee was not to analyze what is wrong with the industry right now, but rather articulate the scientific challenges we should be addressing. Gio Wiederhold and Susan Davidson were the other database folks on the committee, which also had experts from other fields of computer science, bio infomratics, and from the medical estblishment.