The Costs of Interfacing

Body Politic - With Your Powers Combined...The Body Politic is a metapohor from the 17th century used to describe the various classes and cultures of a society as parts of a body. Each section had its part to play in order for the society as a unified organism to move forward. With each social class performing its function, they work together to move the nation toward the betterment of all citizens. Disparate elements combining powers to achieve a singular goal - sounds very much like the current landscape of health IT. As much as EHRs are bringing many workflows under one umbrella, there are still different programs and departments that capture different types of data that have to work together. In order for these systems and workflows to transfer information, facilities have to set up interfaces. Interfaces are what pick up or feed pieces of information between one system to the next - results of a test are entered into one computer program and then sent over into the EHR, or a scheduling program puts in a patient's appointment that will then generate a billing process in the accounting software, etc. Most facilities have interfaces between their many health IT programs since they also have a myriad of programs to serve the specific needs of that department, specialty, and/or workflow. Unfortunately each of those interfaces can come with a pretty hefty pricetag.

In a great post on, Sovita Chander writes about the reasoning for the high costs of healthcare interfaces and the unique conundrum it presents to both consumers and developers. Consumers - hospitals and other healthcare facilities -  are confused when presented with large bills for interfaces because surely these companies have had to interface before, why the staggering price for work that must have already been completed?

Developers - health IT companies providing the services - are awkwardly positioned to explain (as Chander does) why each interface is a digital snowflake, unique and unlike the others that have come before it. There are obviously shared traits between all the interfaces, and the trend is toward standardization, but it nonetheless requires real work by IT staff to make the two systems function the way consumers need them to. It's (theoretically) easy to send a PDF between two systems and just send them into a file to be read later. What's difficult is identifying the specific data elements you need from each system, in what format, and sending them back-and-forth to populate and kick off various other functions. How McKesson captures diagnoses may not use the same language/data/formatting as how Cerner kicks off orders. And if you do it once between two companies, like Epic and Meditech, it may only work a certain way for those particular products/programs/clients used.

 It takes time and hard work by development staff to work on interfaces because of those unique ways of capturing data and presenting them in each system. Those unique ways are specifically different based on many variables: the type of data being captured, the way the data will be used eventually, the version of the program capturing it, the manufacturer of the software, the environment in which it's captured (firewalls, cloud-based, server-installed, etc.), the number of systems involved, who needs access to it, and many others. Unless all of those elements are the exact same, interfaces will require time to build anew and that means money.

Make no mistake - this is an apologia for the time and costs of interfaces, not an excuse for them. These costs shouldn't be something the health IT community has to deal with and the facilities have to pay. And, hopefully, they won't be for too much longer. The key to reducing these costs is standardization. Standardization not just of the methods of sending information - HL7, CCDA, etc. - but also the type of data being captured. We've long talked about the push for standardization in structured data capture - but there's also a need in defining templates for the reports. National and International organizations need to start supporting, creating, and disseminating agreed upon templates for reporting in radiology, surgery, and many other specialties - much like CAP has done for pathology and CPAC is beginning to do with cancer surgeries. If there's a standardized set of types of data, forms, reports, and formats that will always be in the same location and can be transmitted using the same formatting language, then interfacing costs will go down as there will be less need for customization.

It's a great piece by Chander and you all should read it. These facilities all exist for one purpose only - to improve the health of people. No matter what else they specialize in or research they perform or services they offer, that is the underlying goal for every single hospital, ASC, clinic. However, within that one facility with its one objective lie a whole host of semi-autonomous sections. This body politic relies not just on each department doing its job to the best of its ability, but also to work in tandem with others in order to progress in delivering excellent care to its patients. Right now, the cost of these disparate elements working together is high - but hopefully by instituting standard messaging languages and unified templates for reporting, it won't have to be for much longer.