The translation of clinical studies into the actual clinic is the holy grail of biomedical research, but it's a lot harder than it sounds. On the one hand, says Forbes' David Shaywitz, to really study and understand something, you need to take it apart and focus on the most basic aspects. On the other hand, putting everything back together to the point where it can help patients is difficult. "While tenure committees and the Nobel Prize selection committee are typically enamored of captivating scientific experiments, the rest of us (who, incidentally, are paying for much of this fascinating, federally-supported science) care most about whether these insights will ultimately translate into something that will improve people's lives," Shaywitz says. It's become increasingly clear that it can be challenging to extrapolate from model organisms to humans, and equally hard to generalize from controlled clinical studies to the actual clinic. In addition, there are many biological and environmental factors that can vary patients' responses to certain drugs, even when they have the same disease. "As if all this wasn't enough, the many factors associated with variability interact in a non-linear fashion, which essentially means that even if you correct one problem, the overall variability improves distressingly little," Shaywitz says. So what's the answer? "My view: if you really want to maximize the chance you'll be helping patients, you need to focus not on developing individual health products (whether drugs or technology) but rather on developing health solutions, integrated approaches: potentially combining a drug, companion diagnostic, monitoring device, and patient support platform likely including a social network component," Shaywitz adds. That's certainly an ambitious approach, but given the challenges of translation, it might be necessary to get ambitious.