Supernova Award Category

The Problem

One of the largest costs for a healthcare provider is running a perioperative unit, and Mercy has over 30 of them. A large portion of this expense is related to surgical supplies, and those can vary dramatically between different types of procedures (which you’d expect) and between individual physicians as well (which you might not expect). With medical reimbursements being driven down and supply costs going up, Mercy had to find opportunities to better manage these costs.

The Solution

Once Mercy realized the variability it had in its cost/case metric, it looked for a solution that could deal with not only a vast amount of data (over 10 million individual surgical supplies used each year) but that could also quickly adapt to an analyst’s changing questions. Mercy looked for tools that could support multiple levels of analysis (from operational executives to data scientists) that could support rapid information delivery and more important agile delivery of new datasets. These datasets were made available appropriately as both pre-defined guided analysis applications as well as self-service enabled information spaces.

The results

Prior to the delivery of these analytic applications, Mercy was unable to zero in on the greatest areas of opportunity in their perioperative supply costs – it was known that certain locations/physicians/procedures spent more on supplies than others, but not if that was justified, and if not, why. With our new solutions, analysts can now tell which procedures or supply categories have the most variability, then dig in and see which surgeons have the most opportunity and why. For instance, it was found that certain providers were using more bone cement than others, and leaders and physicians were able to discuss why and make appropriate changes. Beyond supplies, Mercy also used this solution to provide insight into other cost-drivers, such as block utilization.

Metrics

Mercy was able to quickly achieve results from this solution on a number of fronts. From a pure cost perspective, analyzing the supplies used on just total knee replacements saved over $1.2 million dollars in the last fiscal year. Operationally, Mercy was able to improve operating room block utilization (the case minutes as a proportion of total blocked minutes assigned to a surgeon) by 12%. Finally, we were able to eliminate an auto-transfusion device from a number of cases, saving unnecessary process and cost without negatively impacting the procedure’s outcome.

The Technology

Mercy used its existing analytics toolset (SAP BusinessObjects) and added SAP HANA as a sidecar database for performance. SAP HANA not only sped up the dashboards and datasets, but was also able to deliver the solutions much faster because that performance enabled very rapid development. We were also able to utilize the SAP HANA Predictive Analysis library.

Disruptive Factor

The game changing step in this project was taking an area with known opportunity and breaking it down to an actionable opportunity. Previous methods to gather this data took huge amounts of time to compile, and weeks to slice and dice, which prevented any follow up questions from reasonably being answered. Today, operational users can spot an anomaly in the data presented and quickly drill in to see what location or supply category or vendor or provider stood out, and with just a few clicks after that they can see alternative options that are working in other circumstances.

Shining Moment

Beyond traditional cost analysis, Mercy is now able to apply statistical models to identify which projects might bear the most fruit. An application was built which allows an analyst to dynamically group procedures and calculates the nth percentile cost for that procedure and determines the total opportunity for Mercy if all physicians were at or below that line. Now with just a few clicks, an analyst can first determine where they should spend their time, then dig deep on that area.