Wednesday, July 23, 2014

Patient Generated Data Goes Mainstream

Since 1999, Patientsite, the BIDMC  shared record between doctor and patient,  has enabled patients to track blood pressure, glucometer readings, activities of daily living, mood, pain etc. but few patients have used those features.

Why?  Because it requires time and energy to maintain that data.

What if data gathering was entirely passive?

Today, I own a Withings bathroom scale and Withings Pulse O2 wearable monitor.    When I get up each morning, my weight, body mass index, and fat percentage is measured by my scale and wirelessly sent to the Withings cloud where it is routed to my Microsoft HealthVault account and my iPhone Withings app.

As I go through each day, my Pulse O2 device tracks my sleep pattern, my activity level (distance and elevation), my heart rate, and my pulse ox measurement.    Using Bluetooth Low Energy (BTLE), all the data is instantly synched to my iPhone and viewable by hour, by day and by week.

There is nothing I have to do.   All of this just happens as part of my activities of daily living.

There are three factors that are combining to create a perfect storm for patient generated data to enter the mainstream

1.  The devices, standards (content, vocabulary, transport), and usability are good enough.  The total expense of acquiring/using them is cheap enough

2.  ACOs are beginning to accept the fact that home care devices such as pulse oximeters, blood pressure cuffs, electronic  scales, glucometers, and sleep monitors will be essential to care delivery between episodic visits with clinicians.

3.  It’s likely that Meaningful Use and other government programs will offer stimulus (or penalties for non-compliance) for incorporation of patient generated data into the electronic health record.

Every year I take on a personal project and do my best to roll it out in production.  In the past that has included administrative simplification, healthcare information exchange, patient/family engagement, Google Glass, mobile support for our consumer/clinician facing applications, shared care plan creation etc.

Patient generated healthcare data is the next key frontier in care coordination, population health, and clinical research.  Although Beth Israel Deaconess has invested significantly in home care, care management, and telemedicine, it has not yet made the commitment to be a leader in patient generated healthcare data.

I’ve used the pulse oximeter on Mt. Fuji to track my performance and I’ve learned that I walk an average of 20,000 steps per day during my life as a healthcare CIO and farmer.   I’ve learned that sleep pattern is instant deep sleep for an hour, punctuated by light sleep/deep sleep intervals for 30 minutes, followed by waking in 4 hours.    I’ve watched variation of my weight - weekends have more activity and less time eating, so by Sunday night I’m at the lowest point of the week.   During the weekday meetings and office time, I eat more and exercise less.    The good news is that over the course of each month, my weight peaks and troughs cancel each other.

For the next year I’ll be exploring patient generated data, both in devices I use myself, and in the creation of novel applications that enable such data to be incorporated in ACO and clinician workflow without creating data overload for any stakeholder.

I’ll document all my lessons learned along the journey.


Anonymous said...

Flashback: "That's cool!" :)

Jonathan Ross said...

I'd be interested if your definition of applications is to have/create a open platform, not just limited to BIDMC.

As a healthcare technologist as well, with family/friends with chronic diseases, I have thought about ways to create passive and active methods for interacting with patients.

E.g. Building a communications platform to reduce the paper needed for Chronic Disease patients who have multiple doctors, across multiple states, and potentially secondary diseases. As well as leveraging the patient home devices providing information to be tracked by healthcare professionals (e.g. India) to track status/alerts related to their disease, and provide a place for care-givers or POA to access/track their loved ones progress, to help make more informed decisions.

Ideally the EMRs & Labs would deliver this information to the patient as well, to visualize their progress.

Everyone likes to see progress, even if a disease is degenerative, sometimes having a symptom, then seeing bloodwork /basic health data can provide better context for patients and illicit proactive behaviours for those with the disease and/or their care-givers.

Obviously, most ideas come from a void found in your own life.

If you ever get funding within or outside of BIDMC around this topic, I would jump at the chance to see it come together.

Bernard said...


As these devices get better and more people use them won't the perceived value of the data rise? Because right now 'data' from patient seems to require some test to validate its accuracy, even when it's in our own interest to be truthful.

Other question. As this sea of data increases, can you see sites with APIs that will analyze data sets to offer suggestions and/or notice patterns.

Anonymous said...

As part of the use of this data, I suggest a "plug-in cloud suite" be made available from manufacturers to hospitals. I might trust BIDMC with my medical data, but I certainly don't trust Withings. Thus I would allow my wearable device to transmit data *only* to a cloud wholly owned and privately maintained by BIDMC (e.g.). The legal protections, and the ethical ones, for PatientSite data simply aren't available with commercial apps. In fact, they're often "stood on their head" given all the releases and disclaimers one accepts with that little click-thru check box...

Susannah Fox said...

I can't wait to see what you are able to create and implement!

I've collected a few resources for inspiration and information if you (or your readers) are interested.

For example, the talks from the Quantified Self Public Health symposium are worth a listen:

People are crowdsourcing a wish list of patient data tools here:

I think Americans are more ready for the personal data revolution than people might think, based on a national survey the Pew Research Center conducted in 2012. Yes, a lot of the tracking that people are doing is "in their heads" or on paper, but just think about how that could change with a shift toward wearables. See:

And I'm intrigued by the idea of measuring social determinants of health: