Insightbanner
Event Summary

HALICON 2015

HALICON 2015

Minneapolis, MN

The 2015 HALICON (Healthcare Analytics Lean In Conference) took place October 23, 2015. HALICON is Minnesota’s premier event for Data Science in Life Science and Healthcare, run by MinneAnalytics, a nonprofit organization dedicated to serving Minnesota’s data science and analytics community.

Trexin was a Silver sponsor of the event, which took place at Optum’s Eden Prairie facilities and attracted over 1000 attendees from the Twin Cities Health Care and Analytics community. Over 55 speakers delivered presentations on a wide variety of topics related to Analytics and Big Data in Health Care. The one day conference was highlighted by featured speakers including John Santelli, CIO of Optum, and Cris Ross CIO of Mayo Clinic, speaking about the progress of Analytics and Big Data at their respective organizations.

Trexin’s Chief Data Scientist in the field of Health Economics, Walter Linde-Zwirble, spoke on the topic of “The Use of Systematic Review in Healthcare Data Science”, as per the following synopsis:

One of the greatest challenges in modernizing healthcare delivery is that there is a bifurcation between published knowledge in the medical literature and the body of data on all healthcare encounters in large computerized databases. Most of the established risk scores are derived from clinical data, around small populations, translated from clinical terminology to coded data. While these scores have been of utility, they always produce a disclaimer of ‘we don’t know the effects of what wasn’t included.’

Using data science, Trexin has employed claims data from very large populations to derive richer results. We began with the data from the care and outcomes of large population to systematically examine all possible codes to identify factors associated with the outcome of interest.

halicon2015

Walter described how the systematic review of all diagnosis codes created a population health score for Medicare beneficiaries by using an iterative process to compare each code to the context created with iterative modeling. This process can be generalized to a variety of outcomes and situations, using the scale of modern computing to allow for the efficient examination of all possible scenarios.

Nine Trexin experts attended either the event or the VIP reception the previous evening. The consensus was that conference was striking in its caliber and size, feeling more like a national conference rather than a local one.

Help Desk

Let us help you "Get to Done"

We’ve been in your shoes. Let us help you find a better way.