What I learned about CCR, Part I

This is a bit overdue for me given that I learned what I learned about CCR at the ASTM Workshop in May, 2008, conducted by David Kibbe, MD, MBA, and Steven Waldren, MD. I think what I learned is important for supporting patient-centered HIT, so I’m putting my summary here. A reminder that the beauty of Web 2.0 is that nothing is ever finished – feel free to clarify or correct anything I’ve gotten wrong.

The seminar itself was very enjoyable, nicely paced, and I learned as much from the students (including fellow blogger Vince Kuraitis and Sam Faus from Sujansky and Associates) as from the teachers – that’s always the right combination.

Why was CCR created?

CCR stands for “Continuity of Care Record. (link to official site here) ” I think the “why?” of anything is the most important – knowing where things come from tells us where they’re going…

In 2003, there was a desire to figure out a way to take the “green form,” which health care providers used to transmit a patient’s medical record summary to the nursing home, digital.

(In the clinical arena, the transfer to a nursing home is one of the most delicate times for a patient and their family. Inaccurate or missing information can have an enormous impact when the receiving institution is not staffed to reconcile and adjust therapy the same way a hospital is.)

In their discussions, David told us, they began asking about making this XML-based standard “a record for the patient” that could be used for all kinds of care transitions, not just to the nursing home. And so, CCR was born.

(Comment from me – (a) yet another innovation from those in our profession who care for our geriatric peers (b) CCR is fairly new on the scene, 2003 is recent in Informatics history)

CCR was vetted through ASTM International, which from my understanding, is atypical for this organization, which now has a very small focus on Health Informatics. There are other standards bodies, most notably HL7, that have traditionally working in health information.

What exactly is CCR?

Note: The paragraph below is superceded by the comment attached to this post by David Kibbe, MD, which more accurately explains the content and licensing of CCR. Thanks, David!

CCR is two basic things – a “Schema Definition” that resides an a fairly easy to read spreadsheet. And an implementation guide, which provides the “how to” of every element. Both are licensed under the Apache GNU License. This is not 100% the same as open source – but I think it does mean anyone writing software can use this standard to move data around.

CCR uses eXtensible Markup Language, or XML. XML is very standard in modern Web applications. For example, the piece of software that I am writing this blog post on is going to send my words via XML to my blog, and my blog doesn’t really care what software I used to send it my words from, as long as it uses the right XML code to contain it in. This is important in a standard – it’s connected in a logical way to standards already used by other industries to move data (getting us away from the “health care is different” model).

Is CCR supposed to help with portability and interoperability?

As David explained to us. These are two different things. Portability means you can move the information to another place, via paper, fax, or digitally. Interoperability is the ability of “XHR” software (“X” can mean “P” or “E” for EHR or PHR – get it?) to communicate and exchange data so it can be used for its intended purpose, for example to trend with data like it, or to be used to support population health.

CCR is supposed to help with both, using an approach that is not based on a paper chart – it is concerned with vendor-neutral, human and machine-readable structured data.

Why isn’t CCR based on the concept of a medical chart?

I have worked in large systems that computerize medical charts using state of the art software. This part took me some time to understand. A medical chart is divided into various sections, organized a certain way (depending the doctor/hospital where it’s used unfortunately), and serves various clinical and legal functions. It’s a lot of things to a lot of people.

CCR on the other hand, is more like, “What do I need to know about you (and what do you, the patient need to know about you) for you and I to take care of you right now.” It’s a summary of the most important things, but not necessarily everything. At the same time, it can have multiple observations, like a list of past blood pressures for example. It can be a summary of just one outpatient visit, or of 10 years in a healthcare system. It’s flexible that way. We learned that MinuteClinic, Inc. uses it in the former way, to “summarize” one visit to a practitioner and transmit it as needed by the patient.

Part II coming soon: More of What I Learned about CCR


Ted: Thanks for the post about the CCR standard. A couple of minor corrections. In you section "What exactly is CCR?" I would say it this way:

The CCR standard is two basic things – it is a personal health data set, that is, content about a person such as diagnoses and medications, which resides in a fairly easy to read spreadsheet. An implementation guide, or IG, includes the spreadsheet and explains the "how to" of every element, most of which are optional and may or may not be included in a CCR file produced for an individual. The CCR standard is also an XML "schema," which is programming code that describes precisely how each content element in the CCR should be arranged in XML, eXtensible Markup Language. The schema is a bit of XML technology that connects the CCR standard in a logical way to standards already used by other industries to move data (getting us away from the "health care is different" model). It also makes the content machine-readable.

The CCR standard is licensed by ASTM International, a non-profit. ASTM asks that companies who wish to use the CCR in their products or services pay a one time fee of $100 for the standard IG and schema, but after that they are free to use the CCR standard without any royalty or additional fees. Some additional components related to the CCR standard, such as the XSLT that transforms the .xml file to .html , making the content readable in a web browser, are open source licensed under the Apache GNU License.

I hope these minor changes make sense! The blog is very accurate otherwise.

Kind sir, a question has arisen on the Google Health discussion group http://is.gd/wbeG about Google's implementation of Functional Status in the CCR. They only support the child elements for pregnancy and breastfeeding, and I want to know whether there ARE other elements. I've been totally unable to locate the schema anywhere – can you help?

Hi Dave,

David Kibbe, MD, responded to your comment to me in e-mail, and I am reposting it here for your review. As always thanks to both David and Steven Waldren for being such great and responsive stewards of the CCR standard.

The full schema for the CCR standard is available from ASTM International, the Standard Development Organization under whose auspices the CCR standard has been developed and funded.

If you go to http://www.ccrstandard.com you'll see in the left hand menu an entry "Get the Standard from ASTM" Click on that and you'll be taken to the ASTM website.

The full standard costs $100, but there are no fees for using the CCR standard after accessing the components of the standard. Another way to access the standard is to join ASTM E31, which is $75 per year. The advantage here is that you get to participate in the further development and maintenance of the standard, and vote when the next version comes to ballot.

Google Health supports only 6 data objects, or sections, of the CCR standard, and Functional Status is not one of them.

Hope this is helpful, and hope you're doing well.

Regards, DCK

Yes, I quickly found the ASTM page. Here's a problem, IMO: how are patients supposed to participate in discussing good use of the standard, if it costs $100 to get a copy?

I'm open to discussion but methinks the new world of participatory medicine requires open access to the standards that are being discussed … especially since there's plenty of reason to think that implementation decisions could use some scrutiny.

In this case, the patients contributing to Google's discussion can't even tell whether Google has implemented the full Functional Status element or even a part.

To whom would we speak to request open access for this schema?


For what it's worth, I recently needed to do a little bit of research on the Google Health API, which is based on CCR, and found that they did a nice job of documenting what they're using, here:


You might start there to get a feel for how it's used,


Yes, Doctor. :–) Been there. I know what they're implementing: "Google used the <Function> element to indicate whether the user has specified she is pregnant, or breastfeeding" but as I just said, that gives me no info on what available Function values they didn't implement.

Honestly, you think I'd be whining about insufficient info when I didn't look first? :–)

Your fan,


Sorry Dave, just wanted to make sure the information is in the thread. So noted that you did look ! Ted

So, now I'm really wondering: @CarlosRizo and others at Health 2.0 Meets Ix were talking about various useful applications for CCR; I'd really like to assess them, as I'm sure many other e-patients would.

Whom do we ping or nag about freeing this up? I understand that ASTM needs to cover its costs but the "T" is for technical, indicating that these are usually industrial docs, not something where the public might have a vital stake.

Ideas, anyone? (I'll tweet this.)


I've pasted the e-mail exchange that followed this conversation in a post of its own, here. Feel free to add comments/keep the conversation going there. It's a great one,


Ted Eytan, MD