Big Data, Analytics, And Hospital Readmission Rates

The US government will start tracking hospital readmission rates. Why? Because we spend some $15B each year treating returning patients. Many of these would not need to return if they followed instructions — which involve meds, follow up out patient visits, diet, and you get the picture. To be fair, it's sometimes not the patient's fault. They often do not get a proper discharge summary and in some cases they are just not together enough to comply. They may lack transportation, communication skills, or the ability to follow instructions. Doesn't it make sense to figure out those at-risk patients and do something a little extra? It does. No question. And translates to real money and better care, and this is where big data comes in — and it's nice to see some real use cases that do not involve monitoring our behavior to sell something. Turns out — no surprise here — the structured EMR patient record, if one exists, is full of holes and gaps — including missing treatments from other providers, billing history, or indicators of personal behavior — that may provide a clue to readmission potential. The larger picture of information —mostly unstructured —can now be accessed and analyzed, and high-risk patients can have mini workflows or case management apps to be sure they are following instructions. IBM is doing some great work in this area with the analytics engine Watson and partners such as Seton. Take a few minutes to read this article.