By the fall of 2009 medical societies and companies had begun releasing new data sets with sometimes dire, but more often bold proclamations of ‘a new era of how physicians are communicating’, and more specifically, how physicians are using new, social technologies platforms. It seems like a race had erupted to survey different groups of physicians and then release white papers and blog posts about this new data. By the fall of 2010 I had collected 6 or 8 of these data sets and I began to get an eerie suspicion that the groups behind the data were making some flawed assumptions and that the people interpreting this data were drawing the wrong conclusions…often with some potentially hazardous outcomes

Each of the datasets that was released in 2008, 2009, and 2010 had fallen in the same fundamental trap - repeating the same series of mistakes in their design - and, as a result, after nearly 24 months and countless hours and resources wasted the only thing that had been gained was some very broad misconceptions about these new communication strategies within medicine.

Here are some example questions from these early data sets:

1. Do you use facebook?
2. Do you spend time on social media websites?
3. Do you use ‘youtube’?
4. Have you ever used a social networking site?
5. Which of the following (applications) do you use for professional purposes?

Each question may seem valid on its surface but each raises more questions than it answers, and if we were going to do anything with this data than we needed to deconstruct the validity of these surveys and what the datasets actually mean to medicine…but nobody was applying this level of critical review.

Each of the examples above asks a variant of the question, do you/have you used social media. None of the questions above provides us the slightest indication of what the planners mean by the term ‘use’ – and this is a huge problem. The datasets were being shared widely, even highlighted by the American Medical Association on their website: “Nearly all U.S. doctors are now on social media” yet the data collection methods and the surveys themselves were not being discussed. (http://bit.ly/qGb3BS)

One dataset in 2010 suggested that, 2% of physicians use twitter professionally and 4% of physicians use facebook professionally. Another dataset suggested that 45% of physicians use social media websites ‘for professional purposes’. And a third suggested that 84% of physicians use social networks. From these first surveys we were told that 2% or 45% or 84% of physicians were using social media - this is not the type of precision in data that offers great assurances that we are getting closer to an answer.

In the months that followed data from the American College of Surgeons suggested that 64% of surgeons have a facebook account, 20% use twitter, 35% have participated in blogs, and 82% have used YouTube. In one of the first peer-reviewed publications we learned that 42% of doctors had ever used a social networking site, of these 97% used facebook and 14 % used twitter, but the vast majority of use, more than 89%, was for personal purposes (http://bit.ly/mUWrYK).

The largest survey to date was conducted by a technology company called QuantiaMD (http://bit.ly/nTIneD). In a survey of more than 4000 doctors, of which 79% were already users of their education technology platform, we were told that 87% of respondent use social media for personal purposes and 67% of respondents use these technologies for professional purposes. This is the data that the AMA highlighted and which begat the most buzz about the ‘dawn of the new era’ in physician communication.

But the fundamental question remained unanswered – what did these survey authors mean by the term ‘use’, and maybe more importantly, what did the respondents to these surveys think the term ‘use’ meant when they were answering the question?

Let’s look at this another way. If your boss walked up to your desk in the middle of the day and handed you a note that said, “Do you use social media? Please circle yes or know” How would you answer the question? Would you circle yes because you spend time on Facebook every week, or b/c you once set up a LinkedIn account (ironically enough, because you were looking for another job)? Or, would you think that since your boss is asking then maybe she wants to know about how you use social media in relation to your job. And if she means do you use it for work, does she mean to engage with colleagues, or customers? And beyond this, what could the word engage mean – does it mean you post content, post comments, or just lurk and learn; once a day, once a week, once a year. There are dozens of ways that you could interpret that simple yes or no question – simply put the question was way too vague.

But she wants an answer, she has a $25 Starbucks gift card with your name on it waiting for your answer and so you think you about all the different times you use social media and you circle yes. As you hand the note back to her she looks at it, nods her head appreciatively, places it in an envelope, and says, “Great, keep up the good work…”

But you have no idea why she said that because when you circled yes to the question ‘Do you use social media?” you were thinking about the fact that you are planning your 25^th high school reunion – or about the account you started to connect to family and to monitor your kids, but in your mind you would never ‘use’ social media for work.

When your boss gets back to her desk she has collected 100’s of responses to the survey and the great majority of your colleagues also chose yes – maybe as many as 2/3^rd of your organization – so she rings the head of IT and she advocates for a large investment to leverage this wide spread social media use to spearhead broad company processes. The investment is made, the platform is introduced, everyone in the organization is trained…and 3 months later no one uses the technology.

This scenario may seem like a stretch, or a worst case example, but it happens in business and it has undoubtedly already happened in medicine. It turns out that how you define the word ‘use’ makes all the difference in the way the question is interpreted and what the data means and, if you do not have absolute certainty that the meaning of the word is crystal clear, then there is a great chance that the data you are gathering will lead to faulty conclusions. And we cannot afford to take this risk in medicine.

In medicine there are some great case studies for how these new technologies are supporting the practice of medicine, or perhaps it is better said, there are great cases studies for how these new technologies are being ‘used’ by the medical community. But in order to understand the impact they may eventually have on healthcare quality we needed a very different way of exploring this question of use. As a friend from Mayo Clinic said to me back in 2010, ‘it is a rare opportunity to be able to begin an entire new area of science…to get in on the ground floor, but this also means that there is a huge responsibility to get things started the right way – one false interpretation and anything that follows may end up barking up the wrong tree.”

To understand the role of social technologies in medicine we had to start on the ground floor, we had to establish some very simple definitions, and we had to ask some very specific questions. It was for this reason that I proposed, for the sake of conducting some meaningful research, that clinicians use social media in 3 meaningful ways.

1. To treat – using social technologies as a means of providing direct patient care.
2. To teach - using social technologies as a means of providing a credible opinion and review of breaking medical news and reports for the public.
3. To learn - using social technologies as a means of supporting their own life-long learning – providing a learning and decision-making resource based on the collective knowledge of their own ‘network’.

Each of the three definitions of use has very different impact on the medical profession and each elicits very different reactions.

Doctors using social media to treat patients is considered to be a very high risk endeavor – there are clear issues pertaining to privacy and liability when a physician uses open social technology channels to communicate directly to a patient in reference to their health. And to be sure, every time I’ve ever introduced the topic of social media and medicine the first reaction has been to raise the risks associated with direct patient care. This is understandable, the relationship and the interaction between a physician and patient are given very unique protections within our society – the act of patient care is the hallmark of medicine – so many see the barriers to this use to be exceedingly high. As do I. But a few innovative physicians have been quite vocal about the possibilities of ‘bucket one use’. Jennifer Shine Dyer a self-described, ‘stylish pediatric endocrinologist and tech entrepreneur‘ from Columbus, Ohio has spoken about the online public relationships she has with her younger patients. Through facebook she is able to keep up to speed with what is going on in their lives and remind her patients, most of whom have Type 1, or early-onset, diabetes about the choices they need to make each day to remain healthy. She can remind them about testing their blood glucose and she can answer questions they may have about nutrition.

Doctors using social media to educate patients – as a modern public education channel – is far-and-away the most prominent use today. In many ways this is one of the most logical and natural extensions to traditional practice. From standard websites, to blogs, to content curating services like Tumblr, Delicious, and Scoop.it physicians are beginning to establish a credible online footprint allowing them to ensure that the patients have access to regionally specific or disease specific information. There are still real risks in ‘bucket two use’ but these are risks with which many more physicians are seemingly comfortable. Herein there is a large pool of great case studies to learn from: Wendy Sue Swanson, a pediatrician and blogger from Seattle, Washington; Bryan Vartabedian, and gastroenterologist from The Woodlands, Texas; Jennifer Gunter, an Ob/gyn from the bay area each bring a different voice and cover different topics in health and medicine. Each uses blogs and twitter and other channels to address new, sometime controversial, healthcare issues, from nutrition, to vaccination, to sexual abuse. As Dr Gunter explains her desire to provide public education through the web and social media:

When my children were born extremely prematurely I was deposited squarely on the other side of the stethoscope. As I began to research my children’s medical conditions I was appalled at a lot of the content I read on-line. Information twisted and distorted by bias (the reporter’s bias, the blogger’s bias, or even the investigator’s bias). The fact that a lot was just out of date. And that’s just for starters…(http://bit.ly/vnGeFZ)

But most integral to our topic herein is that beyond some personal experiences we know very little about how physicians are using new and emerging social technologies to support their own life-long learning. So that is the question that I set out to answer in the fall of 2010. I had been bouncing around this idea for nearly a year when I had a conversation with a friend and peer, Maziar Abdolrasulnia. Mazi is a researcher who had previously looked at how to measure physician adoption of electronic health records. As we got to talking, we believed that we could not only measure the adoption of social technologies as a channel for physician life-long learning, but we could also answer some questions about why a physician would or would not adopt these technologies – he suggested we could predict what factors increased or decreased the likelihood of adoption.

Having secured funding from my then employer Pfizer, I used my social network to find some experts to support this new research program and there was great interest in doing this right, rigorously, scientifically. Bryan Vartabedian was the first to join the team. Robert Miller, a breast oncologists from Johns Hopkins was next on board. And, Mazi recruited Molly Wasko. Molly was an academic and department Chair, and an Associate Professor at University of Alabama Birmingham. The five collaborators began to develop a research project to answer the very specific questions: Do physicians use social media to support their life-long learning (bucket three use) and if so, why?

The 1^st part of the study was fairly straight forward; by crafting a series of simple and straightforward questions we would be able to measure the current rate of adoption of new information communication technologies (ICTs) including social media. The 2^nd part of the study used a well described model of adoption to explore why a physician would adopt these technologies. The first part of the study would provide data describing the current trends. The second part of the study would provide data that helped us predict future trends.

Without getting too technical, the adoption and use of a new technology is thought to be driven by a unique combination of attitudes and beliefs about the technology (http://bit.ly/vUnX9r). Do you think it is easy to use? Do you think it is useful? Do you think it helps you perform your job more effectively? Do other people you respect use the technology? Do you see yourself as an innovator, and therefore you strive to be a pioneer with new technologies?

To be sure we were headed down the right path, we refined the definition of use over and over again before settling on the following: use = to exchange information, advice, ideas, reports, and scientific discoveries with other physicians in the medical community. We validated the research instrument – a sophisticated survey – and we collected responses from nearly 500 physicians. In the months that followed the team spent hours analyzing the data; what does it mean, where could we have improved the instrument – we spent a lot of time trying to understand what conclusions we could reasonably draw…remember that this is the first real scientific exploration of this question and we had no intention of making the same mistakes of the numerous data sets that had been loosely collected and unapologetically heralded in the preceding months.

By June of 2011 we had a good sense of what the data meant and we presented the first half of the data – the descriptive data set – at a conference on the campus of Stanford University in September 2011. (http://bit.ly/p2zsJy)

It appears that, depending on the specific technology or application, up to 52% of physicians had adopted these new technologies to support their life-long learning; but there is a sizable separate group of physicians that claim that they will never use these technologies to support their life-long learning; we found only a very small group of physicians that are still unaware of these applications.

From the predictive data we confirmed our belief that the adoption of these new technologies is dependent on the existing beliefs that they are easy to use, provide added benefits in terms of efficiency and effectiveness of medical practice, and have barriers to use which are easy to overcome. Physicians are more likely to use these new technologies in support of their life-long learning if they believe other physicians they respect are using them and if they believe that the technologies are a sign of their personal innovativeness...and we learned much more:

• 33% of physicians believe that these technologies are an essential use of learning time.
• 32% of physicians believe that these technologies are beneficial for learning.
• 36% of physicians believe that these technologies are an engaging means of learning.
• 37% of physicians believe that these technologies are a good way to get information.
• 31% of physicians believe that these technologies are a source of high quality information.
• 58% of physicians believe that these technologies help them care for patients more effectively
• 51% of physicians believe that these technologies improves their professional performance
• 61% of physicians believe that these technologies improve the quality of their patient care

So what does this all mean?  

I see this as a cup that is very much half full, but the picture is not entirely rosy - anywhere from 44-51% of physicians are still on the fence with these new technologies, and 1 in 5 physicians believe that when it comes to learning these new technologies are a ‘waste of time’ and a ‘source of low quality information.’ This research gives us a roadmap to 1) create the value proposition for new technologies as a means of life-long learning, and 2) drive adoption my leveraging this roadmap. Although we are in the earliest days of the adoption many new platforms have sprung up to support physician learning, some with very interesting social layers. These platforms allow us to pilot test our roadmap and move from studies of prediction to studies of causation – and once we begin collecting this data then the proverbial cat will be out of the bag. But importantly, these data anchor the discussion and should help us move from the days of vague and ambiguous surveys to an era of real science, real data, and real lessons on learning.