Using DRG Analytics to Drive TCO: The Holy Grail in Healthcare Analytics

Hospitals are bleeding money.  After a period in the 1980’s where the AMA was dictating prices they could charge, Medicare regulation, Part B, and introduction of bundled payments has now severely restricted the payments coming in to cover operations.  The issue is that hospitals may only charge a certain amount for a DRG.  Known as a diagnosis related group, DRG’s are a system to classify hospital cases into one of originally 467 groups, with the last group (coded as 470 through v 24, 999 thereafter) being “Ungroupable.  It is a way to identify “products” that a hospital provides (e.g. “appendectomy”).  It was developed to convince Congress to use for reimbursement.  DRG’s have been used in the US since 1982 to determine how much Medicare pays the hospital for each “product”.  DRG’s are also standard practice for establishing reimbursements for other Medicare related reimbursements such as home healthcare.  The prior cost-based model allowed hospitals to calculate the cost of an operation, add a margin, and charge the government for reimbursement, which they did without question.

But one of the big problems hospitals are now encountering is that they are unable to bill the government for products that are purchased, but either not properly billed, or lost entirely!  There is even a metric for this used by hospitals:  “utilization”.   To illustrate the problem, consider a hospital that buys 100 stents.   They are purchased, the manufacturer delivers them and they are put into the stockroom next to the OR.  When a cardio operation occurs, the nurse pulls one out.  But maybe she pulls a stent out, doesn’t like the looks of it, and sets it aside.  Or maybe she feels bad for the patient, and doesn’t mark it down in the OR log.  Or just plain forgets to include it in the clinical record.  The operation occurs, but the patient is never charged for the stent.  Lucky patient!  Not so lucky hospital!  So maybe 80 out of 100 stents are captured in revenue, so the “utilization” is 80%.  This represents a 20% loss in working capital, but also a 20% loss in revenue attributed to cardio operations using those stents.  Without being able to track material, hospitals can’t be reimbursed for stuff they buy.  Utilization at most hospitals is around 70%, based on my discussions with experts in this area.  At some it is as low as 50%.  Utilization is a chronic problem -is it any wonder hospitals are scraping by with single digits margins at best?

The key to understanding analytics that can potentially change the hospital ecosystem forever is by understanding the total cost of any patient procedure, and charging patients/insurance/Medicare for an amount that actually allows the hospital to eke out a small margin and reinvest in the hospital.  Today, hospitals don’t know if they make money or not on any given procedure.  As data governance improves in healthcare, the material cost of a procedure becomes a known quantity – but often is the only one that is known!  And when we can understand also what was used in the procedure, we can link it to a unique Universal Procedure Name.  The UPN can be linked to a different DRG’s, which is how hospitals today are paid by Medicare.  Unfortunately, most DRG’s are not granular enough to be useful in capturing the total cost of the procecure.

Here’s how this can change.  By knowing the material cost, the physician cost associated that UPN can also be linked to a  LARGE DATA ARRAY, which includes financial, clinical, and material costs.  Material Codes are the foundation for creation of cost models that can produce analytical estimates of every component of cost for every procedure performed by the hospital.  This can create a level of visibility and insight never before available for a hospital.  Data analytics can be constructed in the cloud in real-time, allowing  CFO’s visibility into all of their revenue flows, profit margins, and cost absorption in real-time, on mobile apps.

This would allow hospitals to be able, for the first time, to answer some very basic questions.  For instance, do you want to know what procedures you are losing money on?  (Most hospitals have no idea).  Which are profitable?  Can you drill down into why they are unprofitable? (Again, most hospitals have no clue)  New forms of analytics systems are needed to allow people to answer some basic questions on their product care portfolio, which no one can do today.

This capability will grow, as healthcare needs to begin to establish a cleansed master database for their material costs.  Hospitals are working in partnership with the Supply Chain Resource Cooperative at NC State, to create new visual platforms, analytical insights, and mobile apps that will become new forms for application of these data to drive hospitals.

The future in store is exciting.  This work can create a direct pipeline between manufacturers and providers, and dis-intermediate GPO’s and software markets, which either do not work or do not provide value.  The benefit of providing a single price, into a single market that all providers, big or small, can tap into, is at the heart of how to make hospitals profitable again.  Manufacturers can get direct market intelligence into how their products are being used, and improve forecasting, new product development, introduction of new technologies, and improved logistics and production planning.  But there’s a long road ahead.