Here’s a great (not to mention right) place to start.

We hear a lot of discussions around leveraging big data, cognitive intelligence, machine learning (ML), deep learning, unsupervised learning, and the like to transform healthcare supply chains. There is a mad rush to sign up with the latest and sexiest technologies (the logic goes, more the buzz words, higher are the capabilities of the company!). A few of the large integrated delivery networks and health systems have gone ahead and announced grand partnerships with consulting companies, analytics providers, and software companies. It’s a virtual magic wand, or so they believe, that will usher in a new era of innovation and automatically elevate the supply chain to unheard of levels.

If other industries have successfully leveraged these technologies to transform themselves, they ask, then why can’t healthcare supply chain do the same? Finance, retail, logistics, transportation, manufacturing, bio technology, cancer research, and more have made significant progress in leveraging machine learning, cognitive intelligence, and predictive  intelligence to drive innovation. But if you closely examine each of these successful industries, you will see a common denominator and that’s significant, deliberate and systematic investments in data foundation.

And therein lies the difference for the healthcare supply chain. Take a hard look at some of its unique challenges and enabling factors:

  • Quality of the product data- Our experience in the last 3 years is that even at highly sophisticated institutions the cleanliness of the data is at best a 5/10. This is fundamental to cross-referencing, evaluating functionally equivalent products, comparing procedure costs of material and standardization.
  • Pricing discrepancies- Because of the complex nature of health care supply chain there are multiple players (GPO’s, manufacturers, dealers, distributors) and multiple avenues for buying products and services leading to a cacophony of data.
  • Variance in procedure nomenclature- Every hospital, every nurse, and every provider has their own way of describing and documenting a procedure and associated costs for a case. This lack of standardization leads to inefficiencies, re-work and sub-optimal results.
  • Absence of normalization, integration of item masters, and data- Whether this is a result of M&A activity or legal ERP systems, continuing to run multiple systems results in duplication and loss of scale and consolidation.
  • Indirect spend/Purchased services- Most companies would not even know where to start because neither do they have the tools to analyze the data nor the category expertise to implement the recommendations from the analysis.
  • A leadership team in dire need of a totally new perspective: This conversation actually happened:

John Doe (CEO): We want to build the largest and strongest data analytics competency in our company, like a veritable Empire State Building of data capabilities. Can you guys do it?

SupplyCopia: Yes Sir, we are excited to build this Empire State Building for you but this is going to take a considerable foundation cost.

John Doe (CEO)- Why the foundation cost is so high?

SupplyCopia: Sir, the Empire State Building is 103 stories tall and based on its design a solid foundation is required to support it.

John Doe (CEO)- (shocked) I was told that you have smart scientists in machine learning. Why can’t they start from the top? Why do you need a foundation?

SupplyCopia: We’ll get back to you.

David Kenny, General Manager of IBM Watson, said it very simply and beautifully in his interview with Fortune Magazine:  “Watson is only as smart as the humans training it. You literally have to tell Watson, ‘Yes, that meant this. Yes, those go together.’ “

In conclusion, we suggest that healthcare systems that have a transformation mind set start with two fundamental things:

1. Lay a great data foundation .

2. Have first class subject matter experts connect the data and teach the machines that can deliver the transformation. Once this framework is created and tested then it’s simply a matter of replicating the success across your organization.