Testing Bundled Care Analytics to Improve Care and Lower Costs

Colorado APCD website screenshot
June 2013

Contact: Center for Improving Value in Health Care
Email:  ColoradoAPCD@civhc.org

Background: HCI3 has developed an analytics system that is a multi-tiered model that begins with individual patients and the episodes or events they trigger at the lowest level, to an association logic that ultimately rolls them up to global populations, permitting both specific individual drill-downs and global trend estimates. At all these various levels (5 total), it is possible to distinguish clinically indicated care and its costs from the costs of care due to “defects” in the provision of care. And since reducing care defects is an important federal payment policy focus in addition to having been shown to reduce costs of care while improving quality, we believe the fruits of this research could greatly benefit all Coloradans.

Goals: Test the applicability of certain costs of care metrics such as total severity-adjusted patient costs, episode-level severity-adjusted costs, quantity and concentration of super-users, and a basket of utilization metrics, and; test the applicability of certain quality metrics, in particular the NQF-endorsed rates of potentially avoidable complications for certain chronic conditions and acute medical events.

Data Analyzed: All claims data (inpatient, outpatient, professional, ancillary) from 2009-2012.

Results: HCI3 was able to demonstrate the kind of wide variation in price, frequency and mix per episode of care that they have observed in other parts of the country and show how actionable the analytics can be for payment reform projects.

Related Links