Why You Shouldn’t Waste Time Building MPI Technology In-House

Thought Leadership

You’re a healthcare technology startup building the next-generation platform in care management, revenue cycle automation, electronic health records, population health, natural language processing, etc. You’re building your platform on a modern tech stack leveraging big data, artificial intelligence, cloud services, flexible APIs, and blockchain.

As you create your magical product you realize that being able to accurately tie disparate data to a single person (patient matching) is key to delivering your ultimate vision. But instead of looking for a modern patient matching technology like you did a cloud provider, you decide to use your team – designed to build the next generation care management platform – to build a complex patient matching functionality. Why?

  1. Because patient matching is an annoying feature you need behind the scenes in your product. It isn’t the main show so it shouldn’t be that hard for your rockstar developers to crank out in a couple of hack-a-thons.
  2. Because your solution is so unique that you need to build your own patient matching engine. No commercial solution providing patient matching services for thousands of clients could ever solve your unique patient matching problem.
  3. Because commercially available patient matching technology is heavy, expensive, difficult to implement and maintain, and only marginally more accurate than something you could build with open source algorithms.

Let me address these one by one.

First, contrary to what you may believe, patient matching is an extremely difficult challenge. Verato’s sustainable growth is proof enough that solving this challenge is both very difficult and very valuable. But if you don’t believe me, The Pew Charitable Trusts just spent two years researching solutions to the patient matching challenge. Here’s their report. Spoiler alert – Referential Matching technology (what Verato does) is one of their four recommendations on how to solve it.

Second, patient matching is not a unique challenge. Ultimately, patient matching boils down to the same question, one that is ubiquitous across healthcare: “Do these two or more entities/records belong to the same person?” The only differences in solving the challenge come in the implementation (i.e. when you ask the question) and in the difficulty (i.e. how good your data quality is).

Third, you weren’t totally wrong – conventional patient matching solutions are heavy, expensive, difficult to implement and difficult to maintain. But they became that way because of how challenging patient matching is, and because they were built before today’s technologies like cloud computing, big data, and machine learning existed. So, they had to be on-prem, built on relational databases, and painstakingly configurable to get accuracy improvements worth their overall cost.

Unlike conventional patient matching solutions, the Verato Universal master patient index (MPI) is built on modern infrastructure enabling us to provide world-class patient matching services (SaaS) at a fraction of the cost of conventional solutions and with flexible APIs that make implementation simple. Our goal has been to build a patient matching utility, optimally tuned for the entire US population, with an easy ramp-up period, that healthcare tech companies can simply plug into.

Every minute your Rockstar developers spend building an MPI or tweaking a patient matching algorithm, is a minute they aren’t building the features you hired them to build – the features that differentiate you from the competition.