At the HL7 Genomics meeting in Washington, D.C., last month, Intermountain Healthcare Chief Medical Informatics Officer Stan Huff, M.D., made a compelling argument for the nonprofit Health Services Platform Consortium as an approach that would allow for the sharing of knowledge, not just data, which could become increasingly important in the age of genomic data.
Huff started out by noting that Utah-based Intermountain provides advanced clinical decision support covering 150 items, ranging from ventilator weaning to sepsis and pneumonia, all of them very valuable. “What we realize, though, is that while we have 150 of those, the opportunity is for 5,000,” he said. “We have focused on ones involving high-volume activities or where the patient is at great risk. But we are not doing anything for hyperthyroidism or other clinical syndromes. But with the budget we have and other constraints, we’ll never get from 150 to 5,000. We might get to 300. We realized we needed to change the paradigm,” he said.
“Our goal is not just to share data, but to share knowledge in executable form,” he said. “We want to share things we have created and get things from others. We can create 150, and somebody else can do 100. The goal is plug and play so we are sharing more than just data.”
As he has mentioned in other talks, Huff stressed that meaningful use has done more harm than good at Intermountain, in his opinion. “Policy has gotten in the way,” he said. “Meaningful use, for all the good intent, in our system was disruptive. It had people focusing on teaching to the test and not being innovative or asking important questions.” He said is worried that Stage 3 is going to do the same thing.
Huff said that in order to share knowledge, health systems need standards for data and data representation, and standard ways to get data out of the EHR, and like HL7 and others, he likes the idea of using FHIR-based services and APIs.
So Huff and others have come together to create the Healthcare Services Platform Consortium with the mission of enabling the acceleration of application development through an open, standards-based, services oriented architecture platform and business framework that supports a new marketplace for interoperable healthcare applications.
The group envisions developers creating applications that don’t need to know about the physical structures of databases in EHRs. So the same application in the cloud using the SMART approach developed at Boston Children’s Hospital integrates into the EHR, and the exact same app can run in Epic, Cerner, Allscripts or Greenway as long as those systems support standards-based services. “Think of the implications,” Huff said. “Today, any useful program has to be recreated in every EHR. As a society, we pay the cost of every one of those pieces of software in vertical architectures.” The services model offers the potential for tremendous advantage and decrease in cost, he stressed.
Currently there are 10 or11 applications running on Cerner, Epic, athenahealth, and others, he said. He envisioned a program for genetic test ordering or an interpretation program that could be accessed in a standardized way. Or developers could make an app that integrates knowledge and data to manage chronic disease, he said.
Ultimately, this could lead to the sharing of genetic knowledge as executable applications, a powerful idea, Huff said. In this approach, you wouldn’t have to move the genomic data; instead you create standard APIs and access it where it resides.
Huff said that if providers make their case strong enough, EHR vendors would follow. “The hardest challenge is changing the environment to support a new way of doing things with standards-based services vs. a proprietary stack. We want to set up an app store,” he said. Right now app stores are provider-specific. “We want to create one that is provider-neutral and vendor-neutral. He thinks the consortium has a good chance of success, but even if it is not, he believes the idea is so persuasive that it will succeed at some point in the future