Saturday, July 19, 2014

High Tech Crystal Balls – Predictive Analytics

Data! Big Data! Bigger Data!

“Big data” and data analytics are all the rage. Moneyball was a successful book and movie telling a story about the use of data and predictive analytics to improve a baseball team.

Providers create vast oceans of data, much of which escapes analysis in the crush of daily operations and the horrors of the revenue cycle. In the past the system could be navigated with reasonable effort and physician compensation was generous enough, so analytics was not necessary to success.

We are in a new era, where bundling and reference pricing and ACOs will change the nature of health care operations.

Some good news – most of your work does not require massive data sets or highly sophisticated computer models – most of your work requires reasonable amounts of data and a well designed spreadsheet model.

Predictive analytics is a data rich version of a classic question - “what if.”

Sensitivity analysis is analytics focused on the change of a single variable. Scenario analysis is a change involving multiple variables. Assuming the group has a sophisticated budgeting process built on a spreadsheet platform, those same spreadsheet model can be altered for focused predictive predictive work.

Don't have a sophisticated budgeting process? Time is a wasting.

One field of accounting knowledge is critical, the ability to separate variable, fixed and mixed costs. This is critical to building proper incremental assumptions in your models. A 20% change increase in procedures does not create a 20% increase in all costs, and knowing what does change is critical.

Executives and administrators have plenty of work to do, the work load is going to get heavier, and more sophisticated analysis and planning will be required. Adding predictive capabilities provides important tools to improve feedback and to improve decision making.

No comments: