The Kelley School of Business at Indiana University hosted a Healthcare Conference on “Patient-Centric Healthcare Management in the Age of Analytics” last Friday and Saturday. I was able to attend on Friday and enjoyed the conference. Special thanks go out to Kurt Bretthauer for organizing the conference and special issue of POM attached to the conference. There were over 70 talks, with 2-3 parallel tracks allowing for 15 minutes/talk.

A couple interesting talks to me:
A group from the University of Cincinnati and Cincinnati Children’s Hospital (Elham Torabi, Craig Froehle, and Christopher Miller) are looking at the triage of patients classified to be “level 3” on the ESI triage scale. The simplest cases of those patients could probably be seen in the Fast Track, alleviating strain on the emergency department during busy times. This dovetails nicely with my work on triaging patients calling EMS for an ambulance transport.

A group from the College of William and Mary and the Bon Secours Health System (Jim Bradley, Chalit Fernando, and Rajiv Kohli) used survey data to identify patient perceptions of hospitals. The data are strongly correlated, making typical regression models difficult, so the group uses PCA in their exploratory analysis to reduce the dimensionality and correlations in the data. The principle components are then incorporated into a stepwise regression to determine the appropriate model. While I wasn’t particularly interested in their problem, I thought it was cool that PCA and stepwise regression were being used together. I haven’t seen either of them used in an operations management paper lately.