Without a doubt, making LOINC freely available under an open license is a key factor in its adoption around the world. Although we often explain the basics of LOINC’s license, I haven’t written a lot about our approach. Recently, I received a thoughtful set of questions about the rationale behind LOINC’s licensing model that provide a nice framework for exploring why the license is crafted the way it is.

LOINC is free to use

LOINC is a freely available international standard, and can be used for any commercial or non-commercial purpose, subject to the terms of use. Has Regenstrief always used the same approach to licensing LOINC, and why did you choose this approach?

In general, Regenstrief’s terms-of-use for LOINC have had the same core principles since its first release back in 1995. Certainly, the license has gotten longer and more explicit about certain things, but the main premises remain:

1. LOINC is free for use in both commercial and non-commercial applications,
2. LOINC is protected against change by end-users or creation of another vocabulary so that it can fulfill it’s goal of being an international standard, and
3. When LOINC is used in databases, software applications, etc. users need to acknowledge the license and copyright so that end users of that product know that LOINC is protected and licensed in a specific way.

LOINC’s development and distribution model was patterned after the Internet standards. The Internet has been one of the world’s most successful approaches to standardization. The IETF provides its standards at no cost, and does not attempt everything at once (one small step at a time). It used “small” amounts of federal funding to underwrite some of the work, but volunteers do a significant portion of development. It requires that application software be provided with the standards, and requires an implementation of the standard in software. This approach has led to the development of more than 2000 interlocking standards, No other standards effort has had anywhere near as large an impact on the information-processing world. None has moved faster.

The LOINC effort imitates the Internet approach. The LOINC standard has tackled one important part of the health information exchange standardization problem – universal codes for clinical observations. We have used extensive volunteer efforts to fuel the development (and also help ensure that we focus on useful, relevant domains). We provide the database of codes and software tools (like RELMA) to facilitate the use of the standard at no cost and distribute them over the Internet.

From the beginning, Regenstrief wanted LOINC to be used as widely as possible, because standards demonstrate the network effect: they become more valuable as more people use them. If you are the only person in the world with a telephone, it isn’t that helpful. But, when every person on the planet has one, a phone becomes a much more valuable tool. When health data around the world are coded in a consistent way, the possibilities for improving health through health information technology are endless. Distributing LOINC at no cost was a key part of keeping the barriers to adoption as low a possible.

Yet, for a standard to be a standard, it has to be the same for everyone. If users could change the content or “fork” it into their own creation, it wouldn’t be universal. The license prohibits altering the LOINC content, though users can augment it with other information, and prohibits using LOINC content to create another code system (vocabulary standard). Both of these situations would defeat the purpose of having a standard in the first place.

To help downstream users recognize that LOINC is protected and licensed in this way, users need to include a brief acknowledgement statement when incorporating LOINC into their products. Such a notice is not needed when exchanging instances of health data like a lab test result between organizations. That would seem excessive.

Did you consider excluding commercial use in order to generate income from licensing?

Licensing LOINC at a cost was certainly considered in the initial deliberations, but ultimately felt to be at odds with the mission of widest possible adoption. Commercial use can have different meanings in different contexts. In order for LOINC to be successful, Regenstrief wanted to encourage use within commercial laboratory information systems, electronic medical records, medical device interfaces, and other software. Patterning LOINC after the Internet standards, Regenstrief thought it would be wonderful if a commercial ecosystem of products, services, and implementations blossomed around LOINC.

What impact do you think the relatively unrestricted terms of use have had on the uptake and use of LOINC?

It would be hard to quantify, but I believe strongly that the open approach to licensing has accelerated the adoption of LOINC and broadened its use. Removing a key upfront cost gets people thinking and moving towards implementing faster. The “open” approach is certainly part of our ethos and culture, so in some ways it is hard to imagine LOINC being developed any other way.

Of course, licensing is only one piece of why someone would adopt LOINC. Being free helps, but LOINC must also be a high quality product, have tools/solutions to help implement it, and an open/ongoing content development approach.

LOINC is open

On your web site, it mentions that LOINC has been developed as an open standard that is distributed freely. What does this mean exactly? Is it the process of development of the standard that is open?

When LOINC is called open, it means several things. First, of course, is that LOINC is distributed worldwide at no cost. It is openly available without fees.

In addition, LOINC is developed in an open way. Anyone can request a new code or suggest edits to any term. The vast majority of the content included in LOINC has been requested by a community member. We consider this to be an important part of our development process. It helps LOINC stay nimble, keeping up with the latest in diagnostic testing advances, and also pragmatic – we only add codes that users have a legitimate need for. In this regard, we welcome requests from all comers: individual users, large laboratories, national health systems, EHR vendors, in vitro diagnostic manufacturers, and everyone in between.

Further, the LOINC Committee is the advisory group that helps set naming conventions and policies. Members of the Committee are volunteers, and all of the meetings are open to the public. We welcome participation from brand new LOINCers, experienced LOINCers, and everyone in between.

Did you consider making LOINC available under the terms of an open licence such as Creative Commons or other?

Well, Creative Commons wasn’t around in 1995 when LOINC began, but their model has informed our evolution of the LOINC license over time. Creative Commons provides a great set of template licenses for documents and media, and there are other variants of open source software licenses that we’ve reviewed. The Creative Commons licenses work wonderfully for documents, and we tend to license our educational content for LOINC under one of the Creative Commons licenses (typically attribution-noncommercial).

There are a few key differences between our approach with LOINC and the Creative Commons licenses that prevented us from using one off the shelf. For example, Regenstrief obviously needed a license that provided indemnity for medico-legal issues. Regenstrief also wanted to to allow users to make some kinds of derivative works such as augmenting the LOINC file with additional data or deleting the content they didn’t need, but did not want to allow other kinds of derivative works that would defeat the purpose of having a standard. Regenstrief was informed by the copy-left / share-alike idea in its approach language translations. We wanted to ensure that any language translations of LOINC could be made available to the global community. Similarly, Regenstrief believed that LOINC users had an obligation to contribute back requests for new codes if they saw gaps in the content, but did not want a share-alike clause for the whole database; a copy-left approach would limit LOINC’s uptake in systems sold by commercial vendors.

Are there any plans to make LOINC more open in the future?

As Regenstrief continues to add more content and resources to the LOINC distribution, we are committed to keeping our open approach. We are actively working on several fronts to make LOINC content more openly available, including as an API through the FHIR terminology services standard. Regenstrief also keeps exploring ways to make LOINC development more open.

LOINC is an internationally adopted standard

LOINC is used in 174 countries, which is very impressive. How many languages has LOINC been translated into?

Yes, it’s quite remarkable. LOINC is used in nearly every corner or the planet. At present, LOINC content has been translated into 18 variants of 12 languages. In addition, the online search browser at search.loinc.org has been translated into 13 linguistic variants. You can find out more about the language translations and which content is available in these translations at loinc.org/international.

Are there any formal arrangements with countries eg national licences to use LOINC? Have any countries formally adopted and mandated the use of LOINC within the healthcare systems?

Because LOINC’s licensing model is so open, there is no need to have a specific country-level level license. About 30 countries have officially adopted LOINC as a national-level standard, requiring its use in certain contexts or adopting it in national level programs and services. (Side note: because of LOINC’s open licensing model, we often don’t hear about these adoptions until after the fact). I list the countries where we know about national adoption in my LOINC Overview and Introduction presentation. For example, earlier this year both France and Italy passed legislation requiring the use of LOINC codes in laboratory data exchange.

How is LOINC development funded?

Since LOINC is made available at no cost, how is LOINC development funded?

The Regenstrief Institute is a non-profit research organization that develops LOINC with the support of external funding. Over the 20+ years that LOINC has been around, many sources have contributed to its development. LOINC’s funding portfolio over the years is diverse, and includes the U.S. National Institutes of Health, U.S. federal agencies, private foundations, industry, the LOINC community itself (through initiatives like the Premium Membership Program), and both the Regenstrief Institute and the Regenstrief Foundation. Today, there are about 8 active external funding sources, with the U.S. National Library of Medicine accounting for about 2/3 of LOINC funding.

Who are your major stakeholders and do any of your key stakeholders (e.g. governments) contribute financially?

We view LOINC as a public good, and consider all of our users as stakeholders. LOINC is used in settings of all kinds, from clinical institutions to health systems, IT vendors, research projects, government agencies, and international e-Health projects spanning multiple countries. As described above, the U.S. government has been a key financial contributor to LOINC development, through the National Institutes of Health and other agencies like the Centers for Disease Control and Prevention and Centers for Medicaid and Medicare Services. Other national programs have contributed in smaller ways through the LOINC Premium Membership Program.

Do any of your stakeholders contribute in-kind, e.g. to the development and updating process?

We rely extensively on our user community to fuel LOINC’s continued expansion and enrichment of content through new content submissions. In addition, the LOINC Committee members serve in this role as volunteers.

LOINC fits well with other standards

How do you see the future for standards such as LOINC, bearing in mind the increasing requirements for interoperability within health information systems? Do you think that the core standards developers need to cooperate more in the area of interoperability (eg mapping standards) and distribution?

Although it sounds attractive, it is clear that no single vocabulary standard can cover all of healthcare’s terminology needs. From the beginning, Regenstrief approached the problem one step at a time, focusing LOINC on a very specific need (standard codes for tests, measurements, and observations), but recognizing that full interoperability across all of healthcare would require other standards. To that end, Regenstrief is committed to working together with other standards development organizations to help health data move seamlessly when and where it’s needed. Regenstrief has ongoing collaborations and working partnerships with the IHTSDO (developers of SNOMED CT) and the RSNA (developers of RadLex) to align vocabulary content, and collaborations with other standards developers like HL7, IEEE, etc to ensure that LOINC’s vocabulary content fits well with health data messaging and exchange standards. Regenstrief has been a longtime member of the Health Standards Collaborative, an executive forum and process for senior leadership of the U.S. healthcare standards development community. I recently became chair-elect of this group and look forward to continuing to find ways for standards development organizations to work more effectively together.

In general, mapping between standards can be a bit of a trap; they can perpetuate divergent ways of doing the same thing and generate ongoing work to keep the multiple approaches in sync. Which, of course, is the very problem that standards are trying to solve. This is not to say they don’t provide value, but in my view should be thought of as an interim strategy. A much harder, but more worthwhile and efficient path is that of unification. Or, at least, agreeing on a coordinated division of labor. This is the general approach Regenstrief has taken with its collaborations.

End users would like the most straightforward path as possible. I believe that finding ways to bundle all of the standards needed to solve a particular interoperability problem would go a long ways, and there is a lot of work that SDOs could do to make this more seamless for users.