From “Pull” to “Push”: A Transformation in Medicine

Weitzman et al just published a very interesting article in JAMA Internal Medicine called “Participatory Surveillance of Hypoglycemia and Harms in an Online Social Network.”  They looked at using the online social network, TuDiabetes, as a method for surveillance for hypoglycemia.  You can read their very novel article here.

I was flattered to be asked to write a commentary on this article.  Below are the first few paragraphs of my commentary and a link to the full text PDF for download here: JAMA Internal Medicine, Feb 2013, Aaron Neinstein, From “Pull” to “Push”: A Transformation in Medicine: Comment on “Participatory Surveillance of Hypoglycemia and Harms in an Online Social Network.


Consider the words we use to describe what a physician does when she or he sits across from a patient to perform a history. Take. Obtain. Elicit.

These words all conjure images of physicians extracting information from patients. We pull information not just from our patients but also from our information sys- tems, calling up vital signs and laboratory results when we want them, on our time and our terms. However, this is rapidly changing, as information will be coming to us from the patients themselves to create “push” medicine. Are we ready? Not yet, but with some pivoting and some preparation, we can be.


Several synergistic technological and cultural trends are leading us toward “push” medicine. Increasingly ubiquitous technologies such as broadband Internet, smartphones, and cloud computing have created fertile ground. There is increased focus on patient-centered decision making. Patients are increasingly well-informed; nearly 60% of adults have looked online for information about health topics.1

Data are coming from many new sources. Mobile applications enable patients to actively create data, such as by answering symptom questionnaires, or allow wireless sensing devices to semipassively generate data like heart rate or physical activity. Other mobile applications use your calendar, text messages, and e-mails to passively generate meaningful health information, such as mood or quality of life.2 The realm of data collected in the home is expanding beyond blood pressure and glucose log books to tracking daily pain and functioning scores for rheumatoid arthritis. Patients are also contributing data through social networks and personal health records and by direct entry into the electronic health record. Patients are increasingly requesting their personal genomes—and to do so they need only curiosity, an Internet connection, and a credit card. These activities are increasingly common, and 27% of Internet users, or 20% of all adults, have tracked their health online.1


The Tools That I Use For Productivity, Creativity, and Research

As an early-adopter and a bit of a productivity nut (I am a big believer in David Allen’s Getting Things Done), I use a lot of different tools in my day-to-day life.  Friends, family, and colleagues know this, and often ask me what I use and what I like.  So, here it is, my list of the digital productivity tools and medical tools that I use and like (I put them into categories but there is some overlap):

** Please note that I have no financial stake or conflict of interest related to any of these companies… I’m just passionate about useful tools and these are the ones I happen to use.  To cement that point, I am purposefully not providing links to these sites so that you can be sure that I am not getting any kickbacks or free space from clickthroughs!

Tools for Personal Organization

Evernote, multi-platform: A Swiss-army knife of a productivity tool.  This is my peripheral brain, and I use it for just about everything.  I use it for taking notes (or photographs of hand-written notes), clips from web pages and blogs, saving important emails and receipts, travel documents, favorite recipes and bottles of wine, magazine articles, and a whole lot more.  Everything is taggable and searchable (including characters recognized within photographs you import) so I can always find it later.

Omnifocus, Mac and iOS: What I use for task management and to-do lists.  There are a lot of other good ones out there like Asana, Nirvana, and Things, and they all have their strengths and weaknesses.

Workflowy, web and iOS: The newest addition to my suite of tools.  Useful for outlining, as in preparing a lecture or article.  Useful for shared task lists and brainstorming.  It is fast, lean, and an incredibly flexible tool.

Tools for Content Creation

Google Drive, multi-platform: Collaborative document creation and editing.  If team members are going to be editing a document simultaneously, this is the tool to use, where Dropbox fails in this regard.

Dropbox, multi-platform: File sharing of all types among a group.  Collaborative document editing that allows you to use a more formal and powerful tool like Powerpoint or Word than does Google Drive.  Cannot simultaneously edit documents, however.

WordPress: For writing this blog!

Tools for Content Consumption and Discovery

Pocket, multi-platform: I have a friend who, before leaving home for the airport, always opens 25 web browser windows with different articles to read on the plane.  Pocket solved this problem (sadly, not for him, because he doesn’t use it).  Any time you find a blog, article, video, etc on the web that you want to read later (the company used to be called ReadItLater), you just send it to Pocket and it is there for you to read at your leisure.  A lot of people also like Instapaper, which does essentially the same thing.

Reeder, multi-platform: A nice cross-platform tool for browsing and reading items in your Google Reader RSS feed account.  If that sentence was in another language for you, go to Google Reader and sign up and then come back.

Docwise, iOS: Despite being an early adopter, I had not owned an iPad until last month.  The feature I really wanted an iPad for was to be able to collate articles from all the different medical journals I like to read into one place.  I heard about Docwise, and then I bought an iPad.  Docwise allows you to mark any number of journals as your favorites, to browse articles from those journals and add them to your reading list, and then to come back when you want and read the items in your reading list.  Note that you still need your institutional library EZProxy or VPN to get access to full-text articles, though you can do this inside Docwise if you have such access.

Instacast, iOS: A very simple and straightforward iOS app for downloading and organizing podcasts.

Much of my web content discovery is done through Twitter and an increasing amount through LinkedIn.

Research Tools

Papers, multi-platform: A few years ago while in the midst of a research project, I purchased and tried using EndNote to manage my references.  It was a pretty miserable experience.  Luckily, a friend suggested Papers, and I have not looked back.  The easiest explanation is that it is like iTunes for your academic PDFs (without so many headaches as iTunes).  You can tag and store your journal articles here, and add annotations like notes and highlights.  This is really great on the iPad.  Their newer version really blew everything else out of the water because it allows you to insert citations and references from your Papers library into any document and it will create footnotes and a reference list for you.

Google Scholar, web: A great tool for starting a literature search.  I always use this in conjunction with Pubmed (below).  I find this to be more “sensitive” than Pubmed, ie the search results might be less relevant, but you get more of them.  An especially nice feature is the “cited by” feature, which shows you the list of articles that have cited the article you are looking at.

Pubmedweb: A must-use for any literature search, I use this in conjunction with Google Scholar (above).  I find this to be the more “specific” of the two tools, ie if you don’t use the proper search term you might find nothing, but if you have better search skills, you will be much more precise in what you find.


I hope some of you find this helpful, and if there are any tools out there that you use that I did not list, I would love to hear about them!


Dreamforce 2012

It was a true honor and a pleasure to be invited to be part of the UCSF Unusual Thinkers panel at Dreamforce 2012 yesterday in San Francisco.  Jenise Wong and I joined the panel to talk about our plans to help revolutionize the way diabetes data is used.  I’ve noticed several reoccurring themes in the last year that came up again yesterday:

1) People are ready for patient empowerment.  This sounds obvious, but that is just the point!  This is no longer an upstart idea or something that should be argued about, but rather a given, a precondition to any new system or idea.  People want to own their health data and it is going to happen, whether or not the medical establishment is ready or desiring.

2) There is tremendous excitement and energy being put into technology in healthcare.  There is a seemingly endless number of people with remarkable ideas, passion… and a startup.

3) There is nearly as much trepidation and concern about HIPAA and regulation by the FDA.  Though nobody will quibble with taking patient privacy and patient safety seriously, the pendulum has clearly swung too far in one direction and is slowing innovation.

4) Nobody can seem to believe the fact that “this hasn’t happened yet.”  To a culture accustomed to upgrading from the iPhone 4 to the 4S to the 5 in the course of 18 months, it is impossible to imagine that a similar pace of innovation has not also happened in healthcare.  There is a mixture of surprise, curiosity, and sometimes sad resignation at this fact.  It is my belief that this great surge of innovation that I mentioned above in bullet 2 is building up behind a dam, waiting for a few structural changes in healthcare policy and economics to occur before it bursts through.


(On a personal side-note, my apologies for my long absence from posting on the blog.  I was off getting married and on my honeymoon!)

A recent reminder of the limits of diabetes technology

As excited as I am, and many people are, about making major strides in improving the capabilities of diabetes technology, we have to keep in mind the limitations of technology.  There is often a time and a place where innovations can improve the daily life and “workflow” of a person with diabetes, but there are situations where face-to-face human encounters will never be replaced.  I had a recent reminder of this.

At one of the locations where I work, I do regular video-chat appointments with people with diabetes who live in remote areas.  At one such visit recently, I spoke with a patient, Gloria, who was using hundreds of units of basal insulin a day and hundreds of units of bolus insulin a day.  Gloria lamented the fact that this amount used to control her diabetes well but no longer was doing the trick and her hemoglobin A1c was now over 14%.  There are many considerations in a situation like this, including ruling out an infection as the cause of her high insulin doses “no longer working,” and after thinking this through, I was tempted to proclaim her situation due to worsening diabetes and insulin resistance.  I was ready to switch Gloria over to U500 insulin (five times the concentration of “normal” U100 insulin… think about the new forms of laundry detergent where you need less fluid to do the job).  This is a somewhat drastic step because of the risks of confusion on the part of the patient but especially the rest of the medical community over her insulin doses, since U500 is relatively infrequently used.

Before switching Gloria over to U500, I asked her to make the multi-hour drive to come see us in clinic, in person, face-to-face.  I wanted to make sure she was actually using her insulin properly before ramping up her dose.  She met with our excellent diabetes educator, who discovered that, in fact, Gloria was making several errors in how she was injecting her insulin.  Over the next two weeks, without any changes in insulin dose, Gloria’s blood sugars came down from 300s and 400s to 100s and 200s, and we avoided needing to switch to U500.

Sitting together with the patient and diabetes educator, watching the patient go through the end-to-end process of injecting her insulin was not something we could do over a video-chat from hundreds of miles away.  There was nothing that could have replaced this in-person, and personal, interaction.  Even as a technology proponent, I smile at that simple fact, because it is the human interaction that led me and many others to become physicians.  When designing new technologies, or thinking about how to implement existing ones into practice, it is important to remember that human interaction cannot be replaced in every situation, nor should we strive to do so.  Technology should serve to augment and enhance these interactions.

Data Design Diabetes Demo Day (whew!)

On Wednesday, I was in New York along with other members of GreenDot to make our presentation to the judges and audience as semi-finalists in the Sanofi Data Design Diabetes Innovation Challenge.  Our mission at GreenDot is to collect diabetes related data from all sources into one platform and make it more accessible, intuitive, and actionable.

The energy in the room during the five semi-finalist presentations and afterwards was phenomenal, and really exciting to be part of.  The people in the room, both semi-finalists and attendees, all have incredible energy and passion about innovations in health care, and there is no doubt in my mind that many major improvements will be forthcoming from everyone who was there.  I’m really happy to be a part of this competition and have the chance to meet so many wonderful people.

For the first time, I even had the exciting experience of meeting someone who recognized me because of this blog!  (Thank you for reading, Anna!)


To vote for GreenDot, click here.  We’ll find out on May 24th whether or not we move on in the competition to the final two.


A few photos from New York and Demo Day:

FitBit activity monitor: I’m a believer

While eyeing things like the Jawbone Up and FitBit for the last year or so, I’ve stayed out of the fray until recently.  As part of my mission to develop a better sense of empathy and understanding for what it is like to have to monitor oneself all the time, I finally started using a FitBit Ultra recently.

I was skeptical, now I believe.  These little things really do have the power to change behavior.  It is not going to overhaul your life, but merely nudge you in the right direction.  But those nudges add up to real change.

I’ll admit, I was formerly a bit lazy and usually took the elevator at work, even to go up a few flights.  I looked for the closest parking spot I could find to where I was going.  Wearing the FitBit the last few weeks, I find myself taking the stairs at every opportunity.  Where I used to take the elevator from the 5th floor of the hospital to the 11th floor in-between our Wednesday conferences, I now take the stairs.  I park my car further away.  Perhaps embarrassingly, I even pace up and down my halls at night while brushing my teeth if I’m a few hundred steps short of a badge, trying to reach that next level.

But how do these seemingly small changes in behavior translate to real-life changes?  I run and I consider myself generally to be in good shape, but in the past, when walking up to our endocrinology clinic on the 5th floor, I would get slightly winded between floors 3 and 5.  This morning, I noticed that I bounded straight up to 5 without much of an issue.  Even more impressive is my weight change.  My medical residency and its incumbent poor diet and sleep deprivation put me up to 180 lbs.  Since residency, with improvements in diet, sleep, and exercise, I have held steady at between 167-171 lbs for the last 18 months, not once wavering out of that range.  However, this morning, I weighed in at 164.5 lbs, more than 2 full lbs lower than I’ve been previously.

Obviously, this is very anecdotal and does not guarantee that everyone using a FitBit will see a result like this.  Nor does it yet mean that my improvements will be sustained.  However, it does mean that a small device that you clip on to your waist can nudge you to make small changes in what you do minute to minute and those small changes can add up to make real differences.

I’m a believer.

FitBit daily dashboard:

FitBit steps log: