I have posted a number of previous notes about analytics (see, for example: Lab Analytics Emerges as Hot Area for Software Development; Relevance of Lifestyle Analytics for Healthcare Organizations; Much of the Future for Pathology and Lab Medicine Rests with Analytics; Successfully Screening for Lung Cancer Based on Predictive Analytics). I believe that the future of clinical pathology, which is to say the mission of hospital-based clinical labs, will be significantly intertwined with analytics. I had begun to categorize the various types of analytics in previous posts. However, Gartner has come up with a very good schema for understanding medical analytics, which they refer to as the analytics continuum (see: Attending to value and sophistication degree, Analytics is organized in 3 levels) and I show their analytics diagram below:
The Gartner term of descriptive analytics in the diagram above corresponds to what I have referred to in past notes as organizational analytics (see: Much of the Future for Pathology and Lab Medicine Rests with Analytics). By this I mean the collection of data concerning the daily operations of the labs with, for example, the goal of improving such processes as lab staffing or deciding which esoteric tests to send out to reference labs. The Gartner use of the term of descriptive analytics is broader than mine. Labs interested in taking advantage of this type of data can purchase a LIS module such as Sunquest Analytics or a system from a specialized analytics and business intelligence vendor such as Visiun. However, we are now also on the verge of deploying a wide range of diagnostic analytics in pathology, which is to say moving up the slope of the Gartner diagram above.
For various reasons, some of which have been described in previous notes here, much of the hospital market, particularly large health systems, now tends to favor enterprise-wide EHR solutions like that provided by Cerner and Epic. This means that they will often install an EHR and a lab module from the same vendor. As a result of this, such lab modules then perform the standard LIS functions such as order-entry, result reporting, data storage, and billing. What "work" then remains for a classic LIS? Is it to slowly evolve as a type of middleware? My opinion in response to this question is that the classic LIS will function, in part, as a specialized lab system to support diagnostic and predictive analytics. To use Gartner's vocabulary, it will provide diagnostic insight and foresight in the delivery of healthcare. How this will work is speculative at this time but I will try to delve into these ideas in upcoming notes.
Comments