I’m a big fan of the online world. I love the ease of online banking, the efficiency of Zappos shoe shopping, and the simplicity of reading The Drudge Report for all the latest news. Someday I may also be a huge enthusiast for online patient portals, but that’s not quite the case today.
During the workday I rarely think about mundane tasks such as scheduling physicals or calling the eye doctor to order new contacts. I am more likely to recall that my daughter needs a follow-up appointment with the ENT when I notice her taking off her hearing aid for the night. Or, I’ll remember it’s time for a mammogram while sharing a bottle of wine with girlfriends and someone mentions the joys of her most recent scan. That last one happens a lot, actually.
I’d like to think I am the quintessential candidate for online patient portals: busy single mom who works full-time and is tech-savvy. I have little patience for being placed on hold for 10 minutes while listening to an endlessly looping recording about the importance of my call. I’d much rather schedule a doctor’s appointment with a few clicks on my keyboard while sipping my first cup of coffee. I get annoyed when my only communication option is to wait until the office opens at 9:00 a.m., navigate the automated phone system, listen to on-hold messages, and finally exchange forced pleasantries with a multi-tasking receptionist.
Recently I had a very positive experience using my primary care physician’s patient portal. One of my specialists requested a copy of my PCP’s referral form in order to schedule a new appointment. I accessed the PCP’s patient portal and in about two minutes found the referral and requested a copy to be forwarded to the specialist. The next day the specialist’s office called to say they had the referral in hand.
Other recent patient portal attempts have been a bit less successful. Typically if I need to schedule any type of medical appointment, I first go to the practice’s website and determine if they have an online scheduling option. That’s what I did a couple of months ago to schedule an appointment for my daughter and the whole process worked beautifully: the system asked for my preferred days and times; the next day I had an email informing me to check to practice’s portal for a message; the message informed me of the appointment time, which I then confirmed.
Unfortunately, a couple of days later my daughter reminded me of a conflict. So, back to the portal I went to send a new message requesting a reschedule. After several days I realized no one had responded to my message. I sent a second message. Again, no response. I ended up having to call the office, navigate the automated phone system, listen to on-hold messages, and finally exchange forced pleasantries with a multi-tasking receptionist.
Another one of my physicians uses a patient portal but its functionality is limited. For example, I am able to request an appointment with preferred dates and times, but rather than having an automated response, someone calls me back to finalize the appointment time. It beats having to call the office and being placed on hold, but if I miss the call or am driving, it’s back to the old-fashioned telephone method.
I often hear providers complain about the Stage 2 Meaningful Use requirement that at least five percent of patients view or download their personal health information via an online portal. Many argue the threshold is too high because many patients lack Internet access or computer expertise, or simply prefer communicating with a live person. However, I’d contend that providers are not doing themselves any favors by implementing poorly designed portals with limited functionality. As a patient, I wonder why I should use a portal if it doesn’t eliminate having to call the practice. I worry that my messages are getting “lost” – either due to technical glitches or office workflow issues. I get frustrated with confusing navigation and functionality that can’t hold a candle to what my veterinarian offers.
In a world where we can spend 10 minutes online and pay a month’s worth of bills, buy a pair of shoes, and read the day’s headlines, why is the healthcare industry so far behind in its efforts to provide patients with a consistently efficient online experience?
Any dancer or doctor knows full well what an incredibly expressive device your body is. 300 joints! 600 muscles! Hundreds of degrees of freedom!
The next time you make breakfast, pay attention to the exquisitely intricate choreography of opening cupboards and pouring the milk — notice how your limbs move in space, how effortlessly you use your weight and balance. The only reason your mind doesn’t explode every morning from the sheer awesomeness of your balletic achievement is that everyone else in the world can do this as well.
With an entire body at your command, do you seriously think the future of interaction should be a single finger? – Bret Victor
The USC Institute for Creative Technologies is a pioneer in Virtual Human (VH) technology. ICT’s work with virtual humans creates digital characters that look, sound, and behave like real people.
Understanding the human face is an especially complex process. The face contains 43 muscles, and it takes five muscles to display whether we are happy, sad, afraid, angry, disgusted or surprised. But understanding and sensing emotions in real humans is key to making virtual characters more realistic.
VH technology is currently being used to help clinicians better interact with patients.
Virtual Human technology is used in role-playing and training to help clinicians improve their interactions with patients. But new research by ICT has netted some surprising results.
New research finds patients are more likely to respond honestly to personal questions when talking to a Virtual Human.
Originally, ICT began training clinicians by having them interact with a Virtual Human patient. In the new research, the tables were turned – patients interacted with Virtual Human interviewees asking questions a physician might normally ask. The process started with general getting-to-know-you types of questions gradually leading to more personal and revealing questions like, “How close are you to your family?”
“Half of the participants were told that their conversation was entirely computer-driven and not being observed. The others were informed they were being watched by a person in another room who was also manipulating the machine to ask certain questions. In all cases, video images of their faces were recorded and later analyzed to gauge their level of emotional expression.” – Tom Jacobs, “I’ll Never Admit that to My Doctor“
Surprisingly, Virtual Humans were able to extract better patient data. In discussing private matters with the computer-generated entities, patients disclosed more information. Why? According to Gale Lucas, who led the study for ICT, participants did not feel like they were being observed or judged. They also reported “significantly lower fear of self-disclosure.”
You can read more about the study in the journal Computers in Human Behavior.
Across the pond, researchers in England are using Virtual Physiological Humans “to engineer a simulation of the body so true to life, any data could be potentially input to create a personalized health plan, and predictions for any future patient.”
According to Marco Viceconti, Director of the Insigneo Institute at the University of Sheffield,
“If I now feed to my simulations the data related to a particular individual, that simulation will make health predictions about the status of that individual. This is not personalized medicine, this is individualized medicine, we can finally say something about you not because you are about the same age and sex and disease as another thousand people, but because you are you with your condition and your history.”
In a recent opinion piece for CIO, Brian Eastwood writes that wearable tech’s dilemma is too much data, and not enough insight. He explains that even though he runs marathons and writes about healthcare IT, he still does not have a fitness tracker.
I started thinking about how Virtual Human technology could combine with wearable devices. Although speech recognition technology is already used with Google Glass, it is not at the level of sophistication of VH. Imagine your own Virtual Human personal trainer who would have an understanding of your emotions and behaviors, and your personal weaknesses and motivators. Interacting with your VH through speech-recognition technology would minimize the need to display lots of data on a small screen. Your VH-enabled wearable device could know just the right words and cues to promote healthy behaviors, and maximize your personal wellness.
There will be no distinction, post-Singularity, between human and machine and between physical and virtual reality. – Ray Kurzweil
If you’re not already familiar, the blockchain is the real breakthrough in technology that underlies bitcoin. Pay attention because the implications are much, much larger than bitcoin and crypto currencies. The implications affect trading almost any kind of information or asset, digital rights, and trust. It’s big. And don’t just take my word for it.
Venture Capitalist Fred Wilson at Union Square Ventures says the block chain will be the next big investment cycle following mobile and social.
Mark Andreesen, Netscape founder and now VC at Andreesen Horowitz, says “when we’re sitting here in 20 years, we’ll be talking about Bitcoin the way we talk about the Internet today.”
Chris Dixon, a partner at Andreeson Horowitz (via WSJ) expands on this idea, “people would have perhaps thought in 1993 that the new network of interconnected computers would have allowed more efficient file sharing and messaging, “but no one imagined Twitter, or Wikipedia, or YouTube, or all those amazing inventions that happened over the next 10 to 20 years… That’s because extensible software platforms [like bitcoin] that anyone can build on top of are incredibly powerful and have all these unexpected uses.“
Big time investors are or will be betting their dollars that this will go much further than an online currency.
What’s the magic behind the block chain and how could it be applied to healthcare?
It comes to a concept of distributed trust. Rather than a third party verifying credit, verifying the deed to property or verifying an identity and all the messy fraud that can go with it, the blockchain network can perform this kind of verification.
The pieces that pull this all together are:
This architecture can be used to verify payments, as is done with bitcoin, solving what’s known as the double spending problem (having your money and spending it, too). It’s a method to ensure that the same dollar can be spent only once. In other words, once you’ve spent it, it’s gone from your account and can’t be spent (by you) again.
A potentially even bigger breakthrough is that the same concept could be used to verify any information exchange in a distributed trust network. In a sense, it opens up the potential that any piece of information that could go onto the blockchain is unique and verifiable, more like things in the real world that occupy a certain place and time. In essence, the same problem that was solved for not having real world dollars and coins for digital currency. Theoretically could be any kind of digital item for which you wanted to establish a true or original copy.
Unlike health information exchange organizations that verify members’ trustworthiness (e.g., HIE organizations, DirectTrust, etc.), the block chain uses hashing, which is creating a unique code (hash) for each collection of content, a process that can’t go in reverse. If you have the content and a key, you can verify that the content existed before by seeing that the same hash is generated. The hash is calculated locally, and you can’t recreate the content from the hash.
In short, the use of hashing alleviates the time and many of the resources needed to verify that a party has done what they say they have done at a certain point in time because all of the information necessary to see if something has changed is published on the network.
Proof of Existence is an application of the block chain as a virtual notary, and perhaps hints at solutions to come. You can prove, mathematically, that you had a certain set of information, in this case a document, at a certain point in time by putting it on the blockchain, which acts like a time stamp and a ledger of activity. If you make changes to it, a hash representing the content will no longer be valid. Artists are using it to establish proof, and possibly ownership, of their work.
To play around with this technology, I encourage you to check out Proof of Existence to see the potential.
What Proof of Existence is used for currently:
The rather small caveat to all of this is that the network has to be large enough and incentivized to check the work on the block chain, but for even small problems, that shouldn’t be a problem. Compared to traditional verification methods, it’s pennies or less on the dollar and more reliable.
These elements of the block chain perhaps could solve some critical issues in healthcare, such as the seemingly opposing forces of privacy and sharing. I’ve often written about the potential for a health information economy once we have a trusted system and people don’t fear their own information could be used against them. What if we could freely share information without fear it could be used without permission and that it was verified?
This is, at least in theory, what the blockchain can offer, but laws will have to catch up. We haven’t tested ownership or use of the blockchain as proof of existence or a time stamp, but that day is surely coming.
There are lots of ideas beyond bitcoin and proof of existence on how the blockchain might be used to verify uses and rights. Here’s a list of 80+ possibilities collected by Ledra Capital. Some I’m sure will pan out, others probably won’t, but it should provide a flavor for some of the potential.
A forward-looking part of blockchain in health care is the ability to build assurance contracts on it.
Kickstarter is an example of one type of assurance contract. If a certain threshold is reached, the kickstarter agrees to build a “public good”. As one bitcoin and Google developer points out in a “future of money” talk, these arrangements are messy for banks. They can too easily turn sour, so the overhead is too great. But with bitcoin, these contracts are relatively straightforward, they’re built into the protocol, with the potential to be automated.
It’s all very early, but there may be an opportunity to write many different kinds of lightweight assurance contracts around health. The opportunities to solve some major issues around identity, ownership, and contracts in healthcare, and elsewhere, seem endless.
This is all very new, how might we protect privacy while improving healthcare services by verifying information and financing activities with blockchain technologies? Why will or won’t any of this work?
Data-filled electronic health records hold the promise of predictive, personalized medicine and improved population health. Up until now, much of the data retrieved from records is structured data, information recorded in specific data fields. However, the total amount of structured data accounts for as little as 20% of the EHR with as much as 80% of EHR data unstructured. For example, all MD and other health care provider notes anhd all written reports accompanying examinations (e.g., radiologic data) is unstructured data.
Barriers to retrieving and analyzing unstructured data are many. (Quoting from @Clinithink) It’s messy, dispersed/disconnected, complex, non uniform/non standardized, varied, and inconsistent — all features that make retrieval, filtering, and analysis difficult. Historically, this data has been accessed manually — an inefficient, time consuming process.
Technological projects to get at this elusive data are, however, ongoing and increasing in number, with partnerships between various data analytic firms and hospital systems. A few examples include:
Additionally, unstructured data analysis has been employed by health plans and ACOs to calculate participant profiles to better anticipate and accurately assess risk, and by researchers to help identify clinical trial candidates.
Let’s discuss the topic of unstructured data during this week’s #HITsm chat. Below are the topics. If you are unable to participate in the chat, please leave your thoughts on the topics in the comments!
A few weeks ago, I read the New Yorker piece “Stepping Out” by David Sedaris on his obsession with his FitBit. Since getting the device, Sedaris says he’s increased his daily steps taken to 65,000, far above the recommended 10,000 steps a day for good health, thanks to the FitBit’s constant documentation of his progress, reminders, and rewards in the form of badges.
Perhaps it’s because I’m a fan of Sedaris, but after reading the essay I briefly thought I should get a FitBit. If it motivated him, wouldn’t it motivate me? Then I remembered the drawer full of pedometers and Weight Watchers tracking devices I already own and the BodyBugg activity tracking and weight loss arm band I had given away because I never used it. Yes, I was the type of person Sedaris was referring to when he wrote:
“Others I met weren’t quite so taken. These were people who had worn one until the battery died. Then, rather than recharging it, which couldn’t be simpler, they’d stuck it in a drawer, most likely with all the other devices they’d lost interest in over the years.”
Before the quantified self movement (using technology and data to gather information about your daily life) had an official name, I dabbled in its lower tech precursors like old school wearable pedometers and tracking my food intake and weight loss on paper or via computer. But even when I got results, I could never keep it up and I was never sure why.
Well, once again I have been thrust into daily data tracking, this time unwillingly thanks to a recent diagnosis of gestational diabetes. My first pregnancy three years ago was full of complications, but diabetes was not one of them. This time I failed both my one hour and three hour glucose tests, earning myself a trip (along with follow-ups) to a dietitian and diabetes nurse educator.
When the diabetes educator asked me how I felt about being there I had one word, “Annoyed.” With my first baby I had issues with my blood pressure that turned into preeclampsia, a condition that is the leading cause of maternal death globally. This meant bed rest, home blood pressure monitoring, visits to my doctor twice a week, weekly non-stress tests, and biweekly ultrasounds. My blood pressure readings just kept getting worse and worse until I ended up having an emergency c-section. So needless to say I was hoping for no complications with this one, hence my annoyance with the diabetes diagnosis.
The educator reassured me that by keeping close track of my blood sugar and following the diet prescribed to me I could minimize complications. At least there are ways to keep this issue in check, I thought, unlike the preeclampsia, which takes whatever course it wants.
I knew from prior experience that collecting data didn’t always sit well with me. Sedaris mentioned that he thought the FitBit worked so well for him because he’s obsessive to begin with. The FitBit’s constant data collection and feedback inspire him to do better. I can also become obsessed with data, but in a negative way.
After my first pregnancy, I followed up with my primary care physician because my blood pressure had not come down to normal levels. When I showed him my blood pressure log, he took a deep breath and gently recommended I only take it twice a day. I had become so fixated on my blood pressure readings that I was checking them about six times a day. And when they were bad, that would cause me stress, which would cause them to rise, which would cause more stress, which would cause them to rise, etc.
Now I am checking my blood sugar four times a day, and again, keeping close track of those darn numbers. I also have to keep track of how many grams of carbohydrates I eat at each meal and what I ate.
After being given this task, I have come to some realization about why daily life data tracking has not worked for me in the past. Instead of seeing data and numbers as motivation, I see them as a source of stress. If my numbers are “bad,” I often want to give up and quit whatever life improvement project I’m on (except you can’t quit preeclampsia or diabetes like you can with walking or weight loss). Numbers cause all-or-nothing thinking for me.
Instead of looking at the total amount of weight lost or steps taken, I look at how I failed because I did not lose weight that week or reach my activity goals. Now I am wondering how I will feel if my blood sugar is too high at a reading. Will that set off the intense stress that my poor blood pressure readings did? And what will that stress do to my blood sugar levels? It could easily have a snowball effect and raise them along with my stress levels.
I’ve also realized there’s more to making lifestyle changes than just data. It requires time and effort and planning. Instead of just grabbing a snack when I feel like it, I now have to plan my snacks and check my blood sugar two hours after eating in addition to upon waking. I also have to keep a food and carbohydrate log. Data determines when and what I eat. It’s a lot more time and thought than just eating whatever, whenever. It requires being more rigid, which is hard for someone like me who takes more of a easy, breezy approach to daily scheduling.
So far, thanks to all my tracking and obsessing, my blood sugar numbers have been good. You’d think that would make me happy or willing to declare that all this recording of data is worth it. Instead, it makes me wonder, Do I really have gestational diabetes? Do I really have to do all this planning, poking of my finger to get glucose levels, and recording? I still do it because I know I have to for my health and the baby’s. That doesn’t mean I have to like it.
If only someone sold a wearable device that could check my blood sugar without puncturing my finger four times a day and plan all my meals for me, I would probably buy it.