The promise of personalized medicine - pharmacogenomics to predictive testing - is that when you tailor medicine to the individual, you have a much better chance of effective treatment. But there's an assumption built into that model: The "medicine" you deliver has to be legitimately effective. That's the lesson of this British study that looked into tailored treatments. The problem: these were herbal medicines, not pharmaceuticals. The catch is that the research was basically a meta-analysis, based on previous clinical trials of herbal medicines. There were only three such trials available, however, so the conclusion that trials of these particular medicines turned up no real positive effects isn't quite the same as saying that all herbal medicines are bunk.
The premise of traditional Chinese medicine is that each treatment is tailored to the individual, a combination of various herbs by a practicioner. It sounds like a far cry from the sort of personalized treatments that genomics might offer. But it's also a warning that simply saying things are "personalized" or "tailored to the individual" doesn't mean that they'll in practice be any more effective.