Previous: Switzerland's Healthcare Explained!
Next: HCT News #1: Eat More Salt



View count:31,790
Last sync:2022-11-15 19:30
The news is full these days of advances in phones and personal devices that will allow us to track and share more and more information about our health. The latest big announcement came from Apple, not long ago, when they announced a partnership with Epic Systems, which is one of the biggest purveyors of electronic medical records around.

Each time I see one of these announcements, it's accompanied by a claim about how any day now, patients are going to be able to share all their amazing data with physicians, and how that's going to change everything. Except it's probable not going to. Why? Watch and learn.

Much of this is adapted from a piece Aaron wrote for the NYT. All of the links you might want for supporting data and evidence are there:

John Green -- Executive Producer
Stan Muller -- Director, Producer
Aaron Carroll -- Writer
Mark Olsen -- Graphics

If it's not clear from the fact that you're watching this on YouTube on the internet right now, I love technology. I'm an early adopter of everything. I've got tons of Apple products, including my iPhone, always sitting right here just off-screen. And I love collecting data! I'm wearing a Nike fuel band right now!

I've also been awarded many grants and published many papers on the use of mobile technology for chronic disease management. I've designed information systems for use in care, and I've studied them as tools to implement decision support. I'm as pro-IT and a pro-mobile technology as they come.

But I'm really pessimistic about any great leaps forward we might see from announcements like this [see annotation] in the near future.


The first big problem is one of communication. More specifically, it's a lack of communication between information systems. All over the place! It's shockingly common for doctors' offices, hospitals, and emergency rooms to be unable to share data because they don't use the exact same system.

Let's say Stan's doctor gets totally going with a full-bore EMR. But when he goes to the Emergency Department they have an EMR from a different company, and no way for the two systems to communicate easily. They discharge Stan home, but no data is transferred to his primary care doc. Later, when he follows up, his doctor sends him to a lab for tests. They too have a different system and it stores data in different fields than his doctor's. Turns out he needs a simple procedure at an ambulatory surgery center, with a fourth way of doing things.

Then [Stan's] primary care's EMR vendor goes out of business. They need to buy a new one but there's no good way to export all the data and import it into the new system, so they have to pay someone to manually enter all the data from one to the other.

This isn't all by chance. It's actually in companies' best interest that their systems don't talk to others' easily; that way when you need to expand you need to keep adding on pieces from them and not from other companies.

But if that's the case, where are the awesome cost-savings? Where are the efficiencies? Where are the reductions in medical errors?

All these new announcements do is throw more data in new formats into the mix. There's no reason to believe it will be any different this time around.

But the issues with technology and data standards are nothing compared to the issues surrounding people themselves. In 2009, Charles Freeman (a giant in the world of medical informatics) proposed a fundamental theorem of bio-medical informatics. It stated that a human working with medical technology is better than a human alone, but the first corollary was that people are more important than technology. So many people ignore this but you can't minimize it's importance.

Real change happens when people change their behavior, not when new technology appears.

And there are lots of issues with people here.
Let's start with patients. First of all some people are private; they don't want to share all that data with their physicians.

Look, some patients don't do what we tell them. For instance, research shows us that adherence to recommendations for insulin regiments is generally poor. Doctors know this. Patients know this. I've even done some of this work myself: we were interested in using mobile technology like cell phones to help adolescents with diabetes share their glucose values with their physicians. Did it work? To the research!

We interviewed adolescents with diabetes, and their parents. One of the greatest concerns was that they were skeptical of sharing too much information. [graphic ends] No one likes to be nagged. In fact, one of our biggest concerns in doing further research was that if the technology got patients "caught" for not doing what they were supposed to they'd stop using it. That would defeat the purpose.

Further we need to be realistic about who's going to use this technology and how much it can help them. The people least likely to use smart phone technology are the elderly and the very poor. These are two groups who are most likely to be in poor health and need the most assistance. We should give it to them! And there's little reason to believe that this will be the way to get it to them.

But the barrier even bigger than this is doctors. No endocrinologist, no matter how awesomely dedicated they are, wants to know every glucose value of every patient every day. No primary care physician wants to know every patient's daily weight. No one wants to know every patient's hourly blood pressure measurement. Physicians are totally stressed by their workloads already. Research shows that over a third of doctors have reported personally missing test results that led to care delays to their patients, because [the doctors are] already overwhelmed by alerts and data. More data isn't the answer.

We could be smarter about how we aggregate the data. We tried that in our own research where we created new systems that automatically monitored patients' data on glucose values. We then used the system to get a Nurse Practitioner to notify a patient if something looked wrong. The results were a modest success at best, and only with respect to short term outcomes-- believe me, I'm not in the business of underselling my own work. Creating systems like these are difficult for offices to do. They require a lot of time to contact patients, and they aren't cheap.

Doctors don't want to spend lots of time or money on something until lots of patients demand it. They aren't going to change the way their offices function for a few early adopters, and until they do so, the people using this products like this will have no one using the large amount of data they're collecting and sending.

It's so, so, so hard to change physician behavior. It's even harder to get them to use electronic medical records. Before the High Tech Act, back in the "dark ages" of 2006, only 11% of office space physicians had a EMR that would be classified as a basic system. After a ton of spending that's gone up to 48% in 2013. And that's a basic system, not one that's ready to implement data sharing from patients' iPhones!

Moreover, and this can't be minimized, one of the most significant factors of leading to physician dissatisfaction isn't reduced salaries or malpractice- it's interacting with an electronic health record.
They hate them, and they often have good reasons for doing so. EMRs often don't work as advertised, they're expensive, and they can be difficult to use and maintain, they can interfere with clinical practice, they can decrease efficiency, they can even get in the way of a good doctor-patient relationship.

I could spend hours talking about the studies that show this to be true and I encourage you to review the references below. Until all of these problems are fixed the likelihood of massive expansion into things like Apple's 'health kit sharing' seems small, no matter how excited those promoting them may be.

There's a paper in the Journal of Modern Healthcare that's entitled, 'Shared Information Can Revolutionize Healthcare'. I want to quote to you from the abstract: "Hospitals soon may learn the value of shared information technology. The idea of "community health information networks' is picking up steam as payers and policymakers look for ways to control costs and measure the quality of care."

That paper was published in 1992. Should happen any day now.