Share this:

Jul 16, 2013 · 5 Replies

Health Tag Ontology Project: Adding Meaning to Social Media?

By Matthew Katz, MD @mattkatz

Patricia Anderson (@pfanderson) and I started discussing an idea of organizing hashtags for cancer care about two weeks ago.  I wrote up an initial post for the American Society of Clinical Oncology, and the positive response has been remarkable. But we wanted to share with the Mayo community as well.

The idea to organize is based around the success of the #bcsm [breast cancer social media] community on Twitter, which now has developed into a real resource for support and advocacy.  In addition, it has attracted physicians to participate.  Wouldn't it be great if similar resources were available for other cancer patients?

To Patricia's and my surprise, since that first post there has been support from all types of stakeholders: patients, clinicians, patient organizations, professional societies, and more.   Patricia has a great explanation beyond my initial idea. And we're not close to having "the answer".  But maybe the time is right for hashtag evolution to be embedded with meaning in a different way than before.

We're calling it the health tag ontology project. A tag ontology is a structured system, more 'top-down' than 'bottom-up'.  While the conventional wisdom would suggest the latter works better in social media, for certain areas where there is a hunger for reliable content, support, and resources a top-down approach that's responsive and collaborative might work.

In my opinion, this kind of system might work best if it is patient-centered, disease-specific and open source for flexible development.  If there's a way we can use hashtags to attract clinicians to get online and interact with each other and with patients, it's worth a try.  But we want to listen first before proposing anything definitive.

What do you think? We'd love to hear your ideas here.




Tags: #hlthtop, Cancer, Hashtag, Tag Ontology


Posted by @skram, Jul 17, 2013

This sounds like a great idea but I am trying to understand how it will be different and/or complement (or supplement) @Symblur's work on at

As the developer behind (which won an HHS app challenge called "Now Trending 2012") and also the idea guy and implementor of other health information social media work (like

Looking forward to hearing more about this and helping out in any way I can!


Posted by @mattkatz, Jul 17, 2013

Thanks Mark. Symplur's healthcare hashtags are great for tracking and analytics. However, it is retrospective, and only tracks created and submitted tags.

My idea is instead to develop an intellectual/structural system that's based on patient-used tags like #bcsm but then expanded so that people with different conditions will be able to form communities more easily. So it's more proactive at predicting the need for such tags.

Some would argue it's too 'top-down'. Maybe so, but so far I've had overwhelming positive feedback from patients, doctors and others. I'll take a look at Mappy Health, haven't heard of it before.

If you would like to add more and see how this develops, Patricia Anderson has set up a Google+ community: We're calling it HLTHTOP for HeaLTH Tag Ontology Project. All are welcome!

Mark Silverberg likes this

Posted by @skram, Jul 17, 2013

Great - I'll join that community and keep an eye out for ways I can help. This sounds familiar to something I also want to do in terms of detecting what people are using and not necessarily just what people have added to a database (a la Symblur).

In order to "intelligently" know what people are using, however, one needs access to the Twitter firehose (or like service) in order to ingest all tweets, look at the hashtags, determine if its health related, and then log. Is that something you all are looking to do with this or will it be manually curated for now? Is this an official MayoClinic initiative and therefore funded/supported (in terms of tech and finances) by Mayo or more grassroots/community led?

Very interested so very curious 🙂




Posted by @pfanderson, Jul 17, 2013

Mark, there are a LOT of ways in which this could potentially have impact.

1. As Matt said, if there becomes a somewhat standard strategy for creating new condition hashtags, or if they already exist even if not yet heavily used, it could possible help to diminish that awkward start-up period in which many micro-communities are trying to do the same thing, and don't know about each other. There are advantages to that, and this doesn't preclude the creation of microcommunities, but it does make it easier to immediately engage with others.

2. If the tags have a standard format and are already listed in tools like the Symplur database, then they are more easily discoverable.

3. If the tags actually fall into the kind of hierarchical structure Matt has designed, that makes many other things possible. Can we archive and retrieve not only the individual hashtags, but groups of related hashtags? Can the streams be scraped for research purposes, to discover common lifestyle concerns across a variety of related conditions? [On the other hand, do we WANT streams scraped for research?]

Or to build community across related conditions! I've wanted to build community across #facialdifference communities for several years now, but everything we've tried has been unsuccessful in bridging the conversation gaps between groups such as cleft and TMJD and oral/ocular cancers, despite the lifestyle issues faced by each group having tremendous overlap.

4. If these hashtag ontologies are endorsed and supported by professional organizations such as ASCO and Mayo, that both raises the likelihood of easy discovery and also may serve to lower the barrier to entry for clinicians and healthcare professionals. That might even make social media more attractive to these professional communities, at least for lurking, and hopefully for engagement.

5. If the tag ontology terms are endorsed and adopted by patient communities as well as professional communities, this may serve to help open connections and communication between the various aspects of healthcare, supporting the epatient movement, supporting improved patient-clinician communication in other venues, and generally improving the quality of health information in patient communities.

Just brainstorming about all this.

Matthew Katz, MD likes this

Posted by @mattkatz, Jul 17, 2013

Patricia is much more eloquent than I. What she said 🙂

The crux to me about success will not be designing it but how it's used. Point #3 is key. If we are going to make something scalable and public for patients to use, we need to educate them about the risks regarding privacy and data scraping. I wrote on this for Mayo back in August 2011 but not sure this hyperlink works now:

In making it, in my opinion it's essential we work closely with patients, advocates and organizations that can help ensure that there is appropriate learning about the risks and benefits that go with such a system.

Please login or become a member to post a comment.

© Mayo Clinic Social Media Network. All Rights Reserved.