In this introductory blog post we discuss the many terms for the convergence of digital data and healthcare that we’re interested in and also raise questions about where, or through which media, we might locate debates or shifts in this field.
One of the most challenging aspects of doing social science research is answering the question: what are we really studying? Not only are the most interesting aspects of social life sometimes elusive, underground, rapidly changing or difficult to characterize but the people we are studying inevitably have their own perfectly reasonable definitions of what’s going on. In addition, interesting objects of study don’t stay put these days: they are rarely confined to particular industries or institutions or even media and circulate rapidly in an increasingly interconnected world.
In starting this seed project, we chose to think through the question ‘what is digital health?’ first using a selection of freely available digital tools. The use of tools like these are not unproblematic in the social sciences but they offer interesting possibilities for exploring a topic in the early stages before the interviews, participant observation and document analysis are completed. It also seemed appropriate to use ‘big data’ style tools because much of what we have been studying concerns making sense of mountains of data. In this post we will first use Google’s N-Grams viewer to introduce the topic and then turn to hashtag analysis on Twitter.
At the start of the project we quickly adopted the term ‘digital health’ as signifying the sorts of technologies and innovations we were interested in: the rise of digital data, wearables and genomics. But we did have questions about whether or not it was the right term? Using Google’s N-Gram explorer we plugged in a number of terms which have been associated with the area of health and innovation in the past: m-health, e-health, telehealth, digital health, connected health etc. Google’s N-Grams uses the dataset of all google books, digital copies of (most) books in print or in libraries around the world, and graphs the occurrence of particular search terms in all printed material for that particular year.
The graph above shows the relative prominence of these terms in relation to each other. We can see that in the 90s telehealth, that is the development of technologies to deliver healthcare to remote locations, exploded as a term. This of course coincides with the rise of mobile phone technology, hype around globalization and concerns about the ‘digital divide’ – that the spread of digital, networked technology was uneven and people were being left behind. E-health, the purple line, coincides with the rapid rise of the web – many businesses and initiatives in the late 90s and early 2000s branded themselves with the prefix e (for electronic). Other prefixes like cyber- virtual- e- and i- have all had their moment in the last fifteen years or so.
Around this time the term ‘personalized medicine’, which describes the catering of medical services to increasingly individual needs, began to emerge as well. This term seems helpful for understanding our topic: how both public health providers and private insurance companies are moving from treating populations as massive aggregate groups to fine tuning both provision and pricing to the individual. However, we are mainly interested in understanding the role of digital technologies in this shift.
What is particularly interesting for our purposes is what begins to happen in the mid 2000s as several other terms start to emerge and jostle for position. By eliminating telehealth and e-health, which are more established, we can see more variations on the networked theme – ‘mobile health’ and ‘m-health’ and ‘connected health’. Notice the spelling difference between ‘personalized’ medicine and ‘personalised’ medicine: the former is a term which seems to have much more currency in America, though recently, President Obama rebranded personalized medicine under the banner of Precision Medicine. The British spelling seems to drop off around 2008, however, the National Information Board employ the the title ‘Personalised Health and Care 2020’ for their 2014 report, a phrasing we might also want to keep tabs on.
So it is important to remember that these different terms have country specific geographies which are important to take into account. Digital health is one of the small players at this point but is certainly on the rise even in the late 00s before the focus of the current study.
The key thing to remember is that terms and neologisms are not just neutral descriptions of the world– various players have strong interests in promoting certain terms over others. If we look at the Wikipedia page for digital health and click on the history tab, we can see that it was mostly written by the user Paul Sonnier who also hosts a popular LinkedIn group and a blog on the topic. So Sonnier has aligned himself to this particular packaging of healthcare innovation. Some companies and institutions may also feel some allegiance to certain terms: a recent Royal Society of Medicine conference we attended was entitled ‘Recent Developments in Digital Health’ but was organized by the Telemedicine and e-health working group. So terms also overlap and co-exist.
However it is also important to remember that different terms may be wedded to particular types of media as well. Google N-Grams only goes up to 2008 and it might be pointed out that some of the most rapid developments may happen off the pages of printed books, which inevitably lag behind the vanguard of technology debates. For this reason we turned to Twitter to get a sense of how the intersection of digital technology and health is talked about at the present moment.
Last year we decided to take a week long sample of activity on Twitter around the following terms (the ones emerging in the above graph and a few more which are specific to wearable devices like fitbit: “connected health”, “digital health”, “precision medicine”, “quantified self”, “tele-health”, “wearable tech”). In addition to reading and analyzing these tweets we can also visualize them in various ways. The following graph is called a co-hashtag network, based on a method called co-word. Simply put it shows which hashtags appear in tweets together – if they appear together in a tweet, they are connected by a line or ‘edge’ and the line becomes thicker the more times they appear. The graph has then been arranged so that more associated hashtags are drawn together for ease of reading. The graph has also been reduced so that only the most connected hashtags appear – again for ease of reading.
Image: Co-hashtag network based on co-word method
One of the most interesting developments here is that insurance appears even without it being queried – that is people are spending a significant amount of time talking about digital health and wearables in relation to insurance without us looking for that. It should be said that this map is very US focused, judging by the co-occurrence of #health and #insurance. Despite this being a dataset in English, the NHS and the UK barely register. Again, the point is that terms are not universally relevant, they have a geography.
The network is dominated, however, by #iot or ‘internet of things’ a term which describes new physical devices equipped with RFID and Bluetooth and are connected to the internet. #bigdata and #wearables also feature prominently. On closer inspection it appears that part of the reason for this is that in the particular week chosen there was an announcement about a smart watch which created lots of chatter in relation to wearables (but not so much in relation to healthcare). Sometimes then we need to adjust our queries when unexpected events skew things.
We added to the previous query the term ‘health’ – so that we would only see text related to iot or wearables in relation to shifts in healthcare.
Image: Co-hashtag network based on co-word method
Keeping in mind, one of the queries we used was ‘digital health’ it is interesting to see that the hashtag #digitalhealth emerges as one of the most connected (that is most used along with other hashtags) along with #mhealth as opposed to #digital and #health separately. The fact that digitalhealth and mhealth are connected also suggests some sort of campaign is going on. Looking at the Tweets we can see that the two terms are often used together by a handful of accounts including (@Paul_Sonnier, @DrAmishShah, @Alex__Butler, @HealthTap which are frequently retweeted:
RT @Alex__Butler: Survey: Few Providers Discuss #Wearables #mhealth Apps With Patients http://t.co/GqaWkT0vIt #DigitalHealth
The goal of using several hashtags in this way is to broaden the possible audience of the tweet – to reach users interested in the intersection of health and technology regardless of what term they use.
So perhaps it is less important to decide on the correct term for our area of study and more important to consider how terms are used by various people, strategically to promote certain understandings of the issue. But as we already discussed, these terms are not value free: buzz words, including the biggest buzz word of recent years, ‘big data’, have the power to reshape entire industries regardless of whether or not the term is understood by all concern. So we need to get past the hype but also understand how hype becomes a part of the story.
What term do you use to describe the intersection of digital data and healthcare? Tell us here.