Arijit Sengupta presents a summary of his work as the CEO of BeyondCore in the presentation “Advanced Analytics for All: Enabling business users to act on length of stay patterns at a leading hospital system.” This presentation was part of the Strata Rx Online Conference: Personalized Medicine, a preview of O’Reilly’s conference Strata Rx, highlighting the use of data in medical research and delivery.
Sengupta’s vision is to bring analytics to people throughout an organization who can use them in their work. He hopes to bring analytics that have traditionally been available only to those at the top of a large organization down to those making everyday decisions. Users of analytics should not need to know statistics or computer science. In this presentation, he shows how hospital employees can correlate the length of a hospital stay with other variables.
Key points Sengupta’s session include:
- Analytics benefit from social interaction: many colleagues can look at the same graphs with the same settings. [Segment begins at the 9:06 mark.]
- To protect patients, make sure you cannot isolate a trait to a small group. In this health application, BeyondCore set 35 people as the cut-off point, but users can change this. The trade-off is between the risk of patients being re-identified and the risk of missing important correlations by making samples too big. [12:40]
- Examples from more than 9 million calculations on 247,000 patient outcomes [14:33]:
- Patients with altered mental states were staying an average of seven days in the surgery ward longer after surgery. The solution was move them to a psychiatric facility and free up the surgery ward.
- Patients with more generous insurance were also staying 10 days longer.
- The tools can show the exact variables used for a calculation and the formulas used to produce a graph, so that someone who knows some statistics can check the accuracy of the graph. [21:43]
Sengupta’s full presentation is embedded below: