As we head into Fall, HL7 International has now published (or reaffirmed) 39 new specification versions 2023 year-to-date. If you're interested in getting notified as soon as new specifications are published, be sure to follow the #Announcements > New Publication thread on Zulip.

We celebrate these specifications because they represent the ongoing work of reaching consensus about how a community of practice wants to address a particular interoperability problem.

Now more than ever we're seeing people around the world using FHIR to create interoperability solutions. And the specifications underlying those solutions are increasing built using our auto-build pipeline.

The auto-build pipeline runs around-the-clock, creating continuous integration builds of specifications under development by HL7 International, HL7 Affiliates (15 or so of them), and many others. There are now more than 390 implementation guides connected to this pipeline!

Behind the scenes, the FHIR pipeline and infrastructure development team have made a continual set of performance and feature improvements to the IG Publisher, FHIR Validator, and validator.fhir.org.

You can track the updates made in each release here. A few of key highlights include:

• Significant progress towards enabling a federated approach to the terminology services needed by the publishing ecosystem (description of the requirements are here | test cases are documented here)
• Significant progress in enhancing our Language Framework towards enabling multi-lingual IGs
• Many many updates, enhancements, and bug fixes to the Validator and Renderer components
• Performance improvements to publishing process
• General updates for security enhancements from our vulnerability check process (OWASP, trivy, CodeQL)

Looking ahead, we already begun some exciting work on visually highlighting changes between releases, adding version-switching banners to publications like THO and US Core, better handling for retired specifications, and more.

The process of seeking, receiving, and incorporating community (and expert, test, and real-world experience based) feedback is the circulatory system of our collective progress in standards development. It takes sustained, ongoing commitment from each of you in the HL7 community to keep this forward momentum. 

Here's what 2023 has looked like in terms of the number of resolved issues each day across all of our specifications.

Since 2021, we've had 668 people submit comments on HL7 specifications. How wonderful. I'm particularly pleased to share that for 2023, of the 306 issue submitters so far, about a 30% of them were new.

Since launching the HL7 3 Yr Plan, we have continued a series of conversations about how we're doing as a community and what we could be doing better. In particular, we've had robust discussion on how we might better optimize for contributor happiness (and reduce burden).To better understand where we stand and help plot the course towards what we want to become, we conducted a survey of Work Group Co-chairs and Management Group members to better understand their perspectives.

A detailed writeup of the key findings from that survey and some of the next steps we're taking based on the feedback are available in this post: Findings from the 2023 Co-chair Perspective Survey. The post summarizes our discussion points and thinking about next steps in the areas of process, infrastructure, and people.

We walked through these key findings with the community on the summer TSC Update Webinar. If you missed that session, you can watch the recording on HL7's YouTube channel.

We have begun working towards implementing the ideas gleaned from this survey, some of which I'm happy to share updates on within this post.

We approach our work in standards development with a learning mindset - always looking to learn how better to create, curate, and deliver data standards that enable interoperability. This mindset certainly applies to how we think about our technical support infrastructure. This Fall we're excited to see the fruits of several key infrastructure investments come online. Here are a few of them.

As a large community of innovators, we want foster a spirit of idea sharing that helps us work together more productively and efficiently. With support from our Cooperative Agreement with ONC (#90AX0035/01-00) we have collated, synthesized, and are now in the process of disseminating several collections of best practices, including:

• Using Terminology in HL7 Standards. A new collection of frequently asked questions (with answers!) has been curated and incorporated into the Guidance for FHIR IG Creation publication. A Confluence page of supplemental guidance focused on HL7 International specifications has also been developed, with answers to questions like "should my custom code system live in my IG, in THO, or somewhere else?"
• Conducting eVotes using a Template. Electronic votes (eVotes) can be a great way to get decisions on routine business, enabling your real-time committee meetings to focus on the issues that most benefit from live discussion. We've now written up detailed instructions on how you can use a Confluence template for conducting eVotes. There's also a short video on the eVote process on HL7's YouTube channel.
• (Coming Soon) Template Jira Dashboards for Work Groups, Accelerators, Project Teams, and Individuals. Jira's issue tracking has become a centerpiece of HL7's standards development workflow. With so much happening across the community, it can be challenging to keep tabs on all of it. We've been collecting ideas from folks who have created nice dashboards in Jira to help organize and visualize their work. We're building template dashboards tailored for 4 different points of view, and once we have them in a refined state, will look for ways to automate their creation:
  • A Work Group managing all of their projects,
  • An Accelerator managing all of their use cases / work streams and the specifications
  • A Project Team managing the development of a particular specification (from inception → balloting → publication)
  • An Individual Contributor managing the work they are responsible for
• (Coming Soon) Continuous Reference Implementation Development. Having reference implementation software for each HL7 specification is a long-term vision of ours. Many specifications today have them, and we want to encourage more. This guide will describe best practices for continuous reference implementation (RI) development that aligns with the FHIR implementation guide (IG) development lifecycle. 

In October 2023, we'll be launching single sign-on (SSO) capabilities for Confluence, Jira, and HL7.org. You'll now be able to use a single (integrated) set of credentials across these platforms - simplifying and securing your access. We're building this capability with HL7's new association management system (Fonteva, which is based on Salesforce). 

Stay tuned for instructions and details on how this will work.

Looking forward, we have a vision of rolling this capability out to additional applications we use across HL7.

HL7's Unified Terminology Governance (UTG) process is how we maintain, extend, and enhance the terminologies that HL7 develops in support of specifications across V2, V3, CDA, and FHIR. The output of this governance process is the publication of the HL7 Terminology (THO), which contains the terminology artifacts that support other HL7 specifications.

When a member of the HL7 community sees a need to update or enhance the HL7 Terminology, they submit a UTG Change Proposal that starts the process in motion. The UTG process is supported by a robust technical infrastructure that we are continuously improving. This cycle, we went live with a major set of enhancements that have been in the works for quite some time, including:

• Moving from BitBucket as the version-controlled repository to GitLab. Sounds mundane, but with that came browser-based built-in editing capabilities that eliminated the need for change requesters to install special tooling or download large files.
• We've also integrated automated builds from the IG Publisher, which validates the proposed content without the need for a local build.
• The process now generates links to the "change objects" (i.e. stuff in the proposal that represents changes to what was there) based on differentials from the commits.
• We've improved the validation checks, such as ensuring committed changes are present prior to proposal submission
• There's also now "just-in-time" dynamic documentation links that provide help text appropriate for the context state of the proposal.

A major thanks to the UTG development team for making this happen. We'll all benefit from these enhancements!

We have a vision for an integrated end-to-end workflow from idea to standards development to publication. We're tackling this big vision in iterative cycles of improvement, under the banner of what we call the Specification Lifecycle Project. 

In this cycle, we focused on improving the Jira-based workflows for the early project phases: Project Proposals and Project Scope Statements (often referred to PSSs). The main goals were to give Work Groups better tools for the review workflow, improve communication across the community about which projects might be of interest to them, and to simplify creation of Project Scope Statements. To do that, we:

• Enabled identification of Co-Sponsors, Interested Parties, and Work Groups who may be interest in a particular Project Proposal.
• Enabled a tagging workflow for Work Groups to identify and track Project Proposals they have reviewed
• Simplified creation of a Project Scope Statement by migrating relevant information from the progenitor Project Proposal and automatically establish a navigateable link between them

You can read more about the updates in this cycle in this post.

Just a few days ago we released a re-worked edition of the Ballot Desktop. From the outside, you won't see much difference because the updates, while significant, were more internal.

First, we fully decoupled logins to this application from the old (GoMembers) association management system to the new one (Fonteva). Second, we excised the dependency on a very arcane web application framework called Fusebox. These two updates put the Ballot Desktop on a more secure and reliable footing while we focus on our other modernization efforts.

HL7 International is keen to have a vibrant Affiliate network around the world. One way we can help is by scaling our infrastructure to support common Affiliate processes.

This Fall, we're excited to be working with HL7 Australia and HL7 New Zealand to extend the HL7 International Jira balloting capabilities to support balloting of their standards. It's been our ambition to support Affiliate balloting since launching with our Atlassian tool suite, and now with our new Association Management System and Single Sign-On coming online, we can enable this more effectively.

Our plan is to establish the technical and process capabilities with these first two Affiliates for their balloting activities in 2024, and then open up the invitation to other Affiliates who may wish to take advantage of this service.

More good things still coming in 2023...

There's a lot packed in this update, but the year's not quite through. Stay tuned for more to come on additional best practices, infrastructure improvements, and process refinements.

Approach

We conducted an online survey from April 15 to May 19 2023. Participants were recruited by the co-chair mailing list and in person at the May 2023 WGM+ meeting. Either anonymous or identified responses were allowed, though we encouraged respondents to identify themselves. Two lucky (randomly selected) respondents received a $50 credit for HL7 swag as a token of our appreciation for their feedback.

Here is what the survey looked like:

Summary Results

The table below summarizes the main respondent characteristics. Overall our response rate was 19% out of the 124 co-chairs and 12 management group members who aren't co-chairs, all of which were technically eligible to respond. We received responses from (at least) 3 first time co-chairs.

Table 1. Respondent characteristics

Based on human review by the CSDO and TSC leadership, assisted in summarization by ChatGPT-4, here are the main themes in response to each survey question.

Challenges with fulfilling duties

• Understanding process steps, especially when it changes

• Uneven influence and consistency of governance among WGs, management groups, and TSC

• Tracking and tracing project progress

• Time commitment, unequal “sharing the load”

• Complexity of the ballot process

If you could snap your fingers to change 1 thing…

• Improved automation and usability of tooling (esp Jira and ConcCall scheduler)

• Have available human “concierge” support for process and technical questions

• Have a broader pool of engaged participants to share the work

• Financial support for attending WGM

What would you eliminate, automate, or delegate

• Automated end-to-end project management (and reporting)

• Delegation of admin tasks (minutes, scheduling, etc)

• Continue simplifying the standards development process

• Unified WG management console (unified dashboard)

What tools would help?

• Automated end-to-end project management (and reporting)

• Unified WG management console (unified dashboard)

• Better calendar and call scheduling

• Support staff

• Collaboration and learning resources (on-demand FAQ, mini tutorials, quick start guides for common tasks)

What is the most personally rewarding aspect of serving?

• Engagement in a true community

• Working with passionate and knowledgeable colleagues

• Making a difference in health around the world

• Personal and professional growth

Closing comments

• Appreciation for the HL7 staff and the HL7 community

• Concern for the financial commitment it takes to be a Co-chair

• Concern for our gender dynamics and respectful communication

• HL7 is complicated; it can be difficult to find what you need when you need it

• Concern for focus on FHIR (to the exclusion of large V2/CDA implementation base)

Discussion and next steps

We presented the findings of this survey at the joint session of the Technical Steering Committee (TSC) and the HL7 Board of Directors during the Board strategy session this summer. Coming out of that discussion, we identified some key themes and specific next steps (in the form of Board motions). Here I've organized these themes into 3 categories for focus: process, infrastructure, and people.

Process

The survey findings underscored that we must keep a continuous improvement mindset about our processes. To resist overgrowth and unnecessary bureaucracy, gatekeepers, or busywork, we want to identify low hanging fruit for aspects to eliminate, automate, or allocate. Ideas for these top-priority tune-ups are always welcome.

Survey respondents also challenged us to scale capacity and availability for key staff and our contributor leaders. When our work and community of contributors grow together, we’re scalable.

The survey responses also prompt us to think about whether we're appropriately matching the rigor of the process with the readiness of the project. What might we be suffocating by applying too heavy a process? Within the TSC we are beginning to explore potential new designations for certain kinds of standards to help better match the process and the project.

Lastly, the survey findings challenge us to enhance our communication and education to all participants. There were several suggestions to try to deliver those resources "just in time" or at the right place where you need it. Our documentation is pretty vast, so just pointing people to this "Encyclopedia of Confluence" site is not going to be very efficient. We may also experiment with the concept of office hours for key staff or seasoned, efficient leaders who are willing to share what has been working well for them.

Infrastructure

The survey findings re-enforced the importance of pushing forward on our vision of having a smooth, efficient, end-to-end specification lifecycle tooling platform.

We also should explore the concept of a unified dashboard for co-chairs/workgroups. We can start be sharing best practices for the component queries and visualizations that our "positive deviants" are already using to help manage their workflows.

People

The survey findings also illustrated that for many people, the most rewarding aspect of serving in a co-chair or management group role is the community and the professional growth that comes from interacting with brilliant colleagues from around the world. Combined with the personal meaning derived from pursing work that is transforming health and health care, these are powerful motivations for why people choose to contribute their time and talent to HL7.

Survey respondents also highlighted the financial commitment required to serve as co-chair, which suggests we should reassess that impact. We didn't develop a specific action plan for this, but did launch some background financial analysis on different approaches.

Last, a few responses noted that we are not always living up to our ideal of fostering an open and welcoming environment. We must do better.

Like the Apache Foundation's maxim "community over code", the essence of HL7 is not the specifications we produce, but rather having a diverse, respectful, well orchestrated community. Our success in addressing healthcare interoperability challenges depends on it.

To that end, HL7 is committed to ensuring an environment of respectful behavior and communication for all. We recognize that to continuously improve, we'll need to be proactive in emphasizing and educating about our community expectations, including how to report unacceptable behavior or misconduct if you witness it in any of our community gathering places (whether in person or only). In addition to the board motion noted below, we are also reviewing how best to incorporate Human Resources expertise into our handling of reports.

We hope that the processes, governance, education, and any corrective actions necessary will help us become the kind of community we aspire to be.

Thank you for your contributions to HL7's collective work.

As always, we welcome your participation and feedback in how we keep refining our community habits, processes, tools, and roles to provide self-sustaining (not burn-out inducing) effectiveness and growth.