Blockchain, Pragmatic Interoperability, and the Workflow-ization of Health IT

What three health IT trends are top-of-mind for me right now?

  • The “workflow-ization” of health IT
  • Pragmatic interoperability in healthcare
  • Blockchain

These health IT trends are all top-of-mind for me because they are coming together and interacting. Individually, they are notable. Together, they may be transformational, to use a tired and over-used word. But in this case it is completely true and appropriate.

By “workflow-ization” of heath IT I refer to the diffusion of workflow technology into healthcare. 15 percent of HIMSS16 conference exhibitors refer to “workflow engine” or “business process management” someplace on their websites. Five years ago virtually zero exhibitors used these workflow industry terms of art. Health information systems are increasingly proactive, transparent, flexible, and improvable, when it comes to workflow, and therefore when it comes to data too, since workflow drives the creation, transformation, and use of data.

“Pragmatic Interoperability” is a phrase I introduced to health IT in 2014, though it existed outside health IT before then. Pragmatic interoperability is the third leg of the healthcare interoperability stool. Syntactic and semantic interoperability are the other two legs. All of these words, syntax, semantics, and pragmatics come from linguistics. Syntax is about the structure of health data. Semantics is about health data’s meaning. Pragmatics is about health data’s use to achieve goals, and to assign, monitor, and accomplish healthcare tasks. When healthcare workflows cross organizational boundaries, this is pragmatic interoperability. When a message is sent from one healthcare entity to another, does the actual effect of the message match the intended effect of the message? If so, pragmatic interoperability is achieved.

Blockchain addresses one of the most important aspects of pragmatic interoperability: trust. Healthcare needs more than just trusted data; it needs trusted workflows. Back in 2015, in a five-part series titled Task and Workflow Interoperability in Healthcare, I argued that workflow interoperability requires workflow transparency between collaborating healthcare organizations. Also see my 10,000 word, five-part series on Pragmatic Interoperability, the linguistic theory behind Task Workflow Interoperability. By combining blockchain and business process management (BPM) technologies, healthcare can achieve exactly this.

To understand how blockchain and BPM can come together to achieve pragmatic interoperability, you have to understand trust. Trust is a hypothesis about future behavior used to guide practical conduct. I trust you, if I believe you will, though action or inaction, contribute to my well-being and refrain from inflicting damage on me. My hypothesis is supported by rationality (it is in your best interest to not harm me), routine (you’ve always delivered in the past), and reflexivity (I trust you because you trust me). Further more, if your goals, resources, intentions, plans, workflows, and activities are transparent to me, I am more likely to trust you. Pragmatic interoperability can be achieved through workflow transparency.

Untrustworthiness does not require nefarious intent. Simple ineptitude can make you, or your organization, an untrustworthy partner. Does that sound harsh? Please read Atul Gawande’s For the First Time in Human History, Ineptitude is a Bigger Problem than Ignorance.

Is workflow transparency possible? Yes. To support my claim, I draw your attention to an important paper, presented at BPM 2016, in Rio de Janeiro, Brazil, Untrusted Business Process Monitoring and Execution Using Blockchain. The use case is supply chain workflows among five organizations and individuals. As I read it, I imagined a similar paper, with a healthcare focus, titled Untrusted Clinical Workflow Monitoring and Execution Using Blockchain.

If I can see your workflows, I am more likely to trust you. In the paper I just referenced, workflow models, and their execution status, are shared across multiple interacting suppliers and consumers in a distributed manner. Instead of a single central meditator directing workflows among subordinate partners (orchestration), blockchain shares workflows as smart contracts among co-equal partners (choreography). Blockchain keeps everyone apprised as to which steps in which workflows achieve what status. There’s even a cool YouTube video demonstrating, step-by-step, workflow execution and changing workflow state.

(The following paragraph is for programmers! Feel free to skip, or not!)

The aforementioned paper is a good introduction to not only blockchain, but also a number of important BPM concepts, such as orchestration (workflow requiring a central conductor) vs. choreography (true peer-to-peer workflow). The researchers translate BPMN (Business Process Model and Notation) models of workflow into a programming language executed on blockchain nodes (Ethereum Solidity). GoLang and Node.js are also involved, so geek out! The research software is a one-of, but future similar platforms will be wrapped in APIs (Application Programming Interfaces) and access workflow, task, and patient data in other health IT systems through FHIR (Fast Healthcare Interoperability Resources) and non-FHIR APIs (see Blockchain as a Platform) Note: trusted choreography among healthcare organizations, to create virtual healthcare enterprises, is especially relevant to workflows between healthcare competitors (trust, but verify!). Finally, you don’t get something (automated trust/trustless consensus) for nothing. A blockchain implementation of BPM-driven workflow across organizations is slower and more expensive than a similar setup without blockchain. However, I believe both can be managed and made small enough to be tolerable.

I’ve been in medical informatics and health IT for over three decades. As an industrial engineer who went to medical school, I’ve long been frustrated by what I regard as insufficient emphasis on not just healthcare workflow, but workflow technology, in health IT. But I’ve never been as excited about the possibilities for creating trustworthy process-aware cross-organizational health information systems. These systems will greatly surpass current EHR and health IT systems in terms of clinical outcomes, efficient use of resources, and patient and user satisfaction.

Viva la workflow-ization of trustworthy healthcare interoperability!

@wareFLO On Periscope!


Leave a Reply

Your email address will not be published. Required fields are marked *