How mature can population-health analytics be if we can’t get our polling data right for a national election? That’s the important question raised by some pundits for healthcare organizations pursuing population health: If Big Data analytics wasn’t sharp enough to predict Trump winning the presidency, how can we expect it to drive population health under value-based care?  “We’re working on it,” is the answer in both cases. The key is the quality of the data, which requires data models that are time-consuming to design and build. In that vein, our Inside Edge report on population health begins at the beginning: defining populations. In a later IE we’ll focus on how to design care models based on those definitions, which stratify populations according to risk.