Opportunity Discovery

Opportunity Discovery allows users to quickly probe the relative performance of attributees (such as physicians or departments) or care delivery variables and to predict the possible impact of performance improvement.

Calculation

Opportunity Discovery is run within a measure. The measure is divided into cells based on the risk stratification fields and is run independently within each cell. The results are then aggregated together. It is run on the opportunity fields of the measure, of which there can be more than one. Low-volume cells are grouped into a single "Low Volume" cell, and low-volume opportunity entities are grouped into a single "Low Volume" field.

Using the example of a surgeon’s opportunity field, within each cell, a regression is run to determine the best- and median-performing surgeon. Then, predictive analytics are run for each other surgeon: If this surgeon's performance were comparable to the best performer, given his or her actual case mix, what improvements would we predict to see? The same prediction is run against the median performer for all below-median surgeons.

Overall opportunity is the sum of the opportunities within every cell. For this reason, Opportunity Discover commonly finds some opportunity for every provider, because one provider is rarely best for every cell.

Setup

In Measure Workshop, the user must identify the risk stratification fields on the measure or indicate that no such fields exist. Risk fields must be either categorical or non-string, or else they are ignored. The user must also specify the opportunity fields and the target direction. 

Opportunity Discovery is calculated automatically when a report is run. Once generated, Opportunity Discovery is available to users if they are unscoped or are leaders on the report, or if the report restriction level is permissive.

Visualization

Opportunity Discovery is presented as a bubble chart in which users can see relative performance (i.e., the coefficient of the regression) on one axis and overall opportunity on the other. The size of the bubbles are determined by the case count.

Different opportunity fields can be distinguished by the color of the bubble. The overall opportunity for each value along the risk dimensions is also shown, with a different bubble color for each risk dimension. Users can drill down into filtered view by clicking on a bubble. For example, if Primary-vs-Revision-Surgery is a risk field, users will see Primary and Revision as two bubbles on the chart, and by clicking “Primary” they can see the opportunities that exist within this type of surgery.