There’s another way to look at being an analytics-focused organization — one that’s unrelated to this discussion but just as important. Yes, it’s important to discuss decision automation vs. recommendations in the EMR. But the first step is looking at the data for your hospital or enterprise to discover areas of unwarranted variation in care and cost, which reveal opportunities for improvement. And that requires a data warehouse that combines medical records, billing, and other systems in a smart way. Outcomes improvement doesn’t usually rely solely on treatment recommendations in the EMR, but a broad range of workflow processes and culture at the hospital. It’s hard work digging in and understanding a problem— asking “why” enough times to discover the root cause — but this activity is vital if you want to understand what to recommend in the EMR and what other changes need to happen to solve that problem. And this analysis depends on data in an enterprise data warehouse.