Incorporate patient-generated health data into the EMR

A few months ago, I spent 15 minutes filling out a detailed health data form at the doctor’s office. The paper form contained multiple questions about my health, family history, medications and basic demographic information. I assumed that an administrative specialist would code it into the practice’s electronic medical record (EMR) to be put to use. So it came as a surprise when I spent another 5 minutes reviewing the form with my physician, who then proceeded to type this information into the EMR herself. I’m confident neither my physician nor I felt enabled by the experience.

Countless people have had a similar experience — or worse, filled out a form with no sign that any clinician ever saw the information. Though the industry has made outstanding progress in adopting EMRs, the practice of data acquisition from patients remains cloudy. Patient-generated health data (PGHD), a term encompassing all forms of data that patients provide on their own, is a relatively new concept in health care. It falls into two broad groups: historical data and biometric data.

Historical data is the type that clinicians are familiar with obtaining from patients: It includes the patient’s medical history, allergies, medications, family history and lifestyle features. Biometric data, little used at present, are health data gathered by consumer medical devices, such as blood glucose meters and fitness trackers.

Scenarios like mine above underscore how ill-prepared the health care infrastructure is for the sharp rise in both opportunities and requirements for PGHD, and the challenges of wrestling these data into the workflow of clinicians. Now, however, an important milestone is on the horizon. A recommendation from the HITSC Meaningful Use Workgroup would require practices with electronic health records (EHRs) to allow 10 percent of patients to report PGHD electronically. If approved in meaningful use stage 3, the final stage of HealthIT.gov’s EHR incentive program, it could push hospitals to incorporate patient-generated data.

This requirement may seem like a relatively simple intervention, but the ramifications are quite significant. If clinical decision-making is made on the basis of data supplied by patients and documented in the EMR, how can clinicians be sure that such data is complete, correct and valid? And will clinicians like me learn to rely on it, or will we disregard it due to concerns about its validity or barriers to integrating it into care flow? Furthermore, if a patient is in control of her health data entry, who is ultimately responsible for its completeness and accuracy — the patient or the clinician?

Incorporating biometric data into the EMR, an exciting prospect, is even more complex. Though clinicians are quite familiar with data entry from FDA-approved medical devices such as blood glucose meters, pacemakers and pulmonary function units, data from a myriad of consumer-driven health devices (Fitbit and others) will soon seek to flex their way into EMRs. Patients clearly value these data; a recent Pew Research report noted that 60 percent of adults claim to track their exercise routine, weight or diet, meaning providers have some catch-up to do in order to meet patients halfway. Some health systems, such as Partners HealthCare, have already been experimenting with the incorporation of PGHD from remote devices into the EMR, and other institutions should follow. Consumer health data devices are moving ahead at a staggering pace, and while the health care system can’t quite keep up, strategic planning should be happening now.

Meanwhile, patients are flocking to sites like PatientsLikeMe and 23andMe to compare and track health data, symptoms and treatment results. Though the connection between clinical medicine and these services is still quite murky, the data show that large contingents of our patients value the notion of comparing and visualizing their health data.

Despite the challenges, incorporating PGHD is a necessary evolutionary step for health care. Intelligently designed, well-executed systems that fully incorporate and display PGHD in a meaningful way will improve shared decision-making and enable patients as active care partners. Keen clinicians and patients will stay closely tuned to the numerous transformations to come.

Israel Green-Hopkins is a pediatric emergency medicine fellow who blogs at Vector, the Boston Children’s Hospital science and clinical innovation blog.

Comments are moderated before they are published. Please read the comment policy.

  • Peter Elias

    Very nice and timely post. Thank you. Some thoughts for discussion…

    ” If clinical decision-making is made on the basis of data supplied by patients and documented in the EMR, how can clinicians be sure that such data is complete, correct and valid? ”

    Let me turn that around and frame it as a patient-centric issue rather than a clinician-centric issue. If clinical decision-making is made on the basis of data documented in the EMR, how can patients be sure that such data is complete, correct and valid? Unless they see it and can comment on it, they can’t. They are trusting us. There is some evidence that this trust is unjustified. I currently give patients a copy of the OV note at the end of the visit, a process which has dramatically improved the quality and accuracy of my notes. I regularly get suggestions and corrections that are useful and valid.

    Health care is about the health of the patient. They have much more at stake than clinicians. It seems just plain wrong to me that we expect them to trust us to be accurate and complete while we resist the idea of letting them collaborate with us on THEIR health care because WE don’t trust THEM.

    The conversation should not be about whether or not they can contribute information to THEIR health record. It should be about how to do it in a way that ALL the information (not just what they contribute) is accurate and useful.

    Peter Elias, MD

  • http://onhealthtech.blogspot.com Margalit Gur-Arie

    I am aware of EHRs that allow patients to input/edit their histories and demographics from at least 5-6 years ago. Lat time I saw one, wasn’t very long ago and it was a makeshift check-in laptop that people got to use instead of a clipboard in the waiting room (one of the most popular EHRs). So the technology is there, it’s one of the few truly time-saving functions associated with EHRs, but few doctors use it. Something I could never figure out, and would love to understand.
    BTW, malpractice is not the answer here, since you (or staff) get to review the information before it is accepted into the EHR.

    As to all the Fitbit and diet and exercise data, I am not sure why that would need to go into an EHR…

    • Peter Elias

      The technology is certainly available. Just not the culture or the business model.

      As a primary care physician, I would love to have the Fitbit, diet and exercise data flowing into the EHR. Then the patient and I can talk about how this impacts health, we can plan interventions, we can track changes, we can see patterns…

      • http://onhealthtech.blogspot.com Margalit Gur-Arie

        Dr. Elias, why does this have to be in the EHR? The patient is there in the room. He or she can log into their Fitbit or whatever they are using, and you can both look at everything and do exactly what you describe.

        I guess my question is about what an EHR is supposed to be. Should it be a medical software package for use to record clinical transactions in a clinical setting, or should it be a health and lifestyle surveillance tool, of which clinical activities are just one component?

        • Peter Elias

          I believe it should be a tool that the patient and the clinician use collaboratively to collect and manage health related information. If it is the patient’s record, the patient needs to have some say in what they find useful and wish to have included.

          If the information remains in their Fitbit, or on their laptop, or on 3×5 index cards on their desk, my partners do not have access to it when they see the patient in my absence. And the information we used to make a decision is not included – unless I wish to become an overpaid data entry technician. This makes it much harder to review.

          Why not have their medication list also outside the record? Because the information is useful and used to make (collaborative) decisions. To the extent that diet or exercise or home weights or steps/day are useful and used to make medical decisions, they should be in the record. And the best way to get them there is directly.

          As long as it is framed as a ‘medical software package to record clinical transactions in a clinical setting’ it runs the risk of not being a patient-centered health record, but a clinician-centered record of billable activity. Perhaps that battle has already been lost.

          • http://onhealthtech.blogspot.com Margalit Gur-Arie

            Perhaps, but the bank record is also a bank-centered record in that respect. I do understand the desire to have an all-inclusive patient centered and patient controlled record, but I don’t believe the EMR can or should be such record, and I don’t believe the bank analogy is indicative of the existence of such records in other industries.

          • Peter Elias

            The bank record is finance-centered, not bank centered, in that it is not designed to record the bank’s intentions or how the bank uses the money in my account.

            And I didn’t say or mean patient-controlled. When I say ‘patient’s health record’ I mean record of the patient’s health. As opposed to a record of the clinician’s activities. I see it as a collaborative process and stewardship rather than a control issue and ownership.

        • http://www.mightycasey.com/ MightyCasey

          Why the [insert expletive here] does a patient have to be IN THE ROOM with a clinician for data to flow into or out of an EHR? Really, why?

          • Peter Elias

            The answer to that is simple but sad. The health care system is not selling health care, it is selling visits. Underlying the resistance to active patient participation is the awareness that this will disrupt the current business model. My office should have a diminished role in health care. Think about reading material. I can still go to a brick-and-mortar book store to find and purchase a book. I can also buy a hard cover book on line and have it shipped. Or I can download it while sitting on my back deck. Same with movies.

            There are things I (a family physician) can offer that cannot be replicated by the electronic transfer of information. If I no longer have to waste my time and the patient’s time and money making them come in, sit in a waiting room full of ill patients, so that I can review their blood sugars or Fitbit data or the family history they have spent time collecting, then I can spend time doing what I do best – helping them understand the medical aspects of their information.

            As long as we are going to insist that the patient and I be in the same room to transmit data to me, perhaps we should make the same rule for the radiologist, pathologist and consultant. Perhaps they should have to come to visit me and I should type the information into the record.

            I have a hard time treating as credible anyone who feels that the radiologist should be able to enter information into the patient chart but the patient should not.

          • southerndoc1

            “I have a hard time treating as credible anyone who feels that the radiologist should be able to enter information into the patient chart but the patient should not.”

            Nothing from the radiologist goes in the charts of my patients unless I have reviewed it and agreed to put it in the chart. Similarly, when patients bring in data that is important, I review it and agree to add it to the chart. I’m liable for everything that is in the chart and I treat all incoming information equally.

          • Peter Elias

            Our circumstances are different. Electronic stuff from the radiologist and lab and neurologist and the case manager and all manner of other folks flow into the chart. My options are to append and sign them or sign them without appending. I review everything.

            Other stuff that comes as paper (brought or sent by patients, letters from outside consultants, info from work-based wellness programs) I review and decide whether or not to scan and then flow into the chart.

            In my setting, I have little ability to keep something out of the chart.

            I absolutely agree with your concern about accountability for everything that finds its way in. For this reason, when a patient from away joins the practice and I get an old record, I review it and select what I want included in the EHR, and then the old record is returned to the patient with the advice to hang on to it in case we later want to look something up. I do not want to be held responsible for information I didn’t generate buried in an old record.

          • southerndoc1

            We handle old records the same way.
            I have no problem in theory with patients adding to, amending, etc. their records, but for that to happen would take a complete revision of our medical-legal system, and I don’t see that happening anytime soon.

          • http://onhealthtech.blogspot.com Margalit Gur-Arie

            I’m sorry, but we seem to be mixing things up here. If we are talking about virtual visits, then fine, I don’t have an objection to people swiping a card at the grocery store and having some clinician pop up on the screen, or doing that from home, or whatever.
            If we are talking about having a conversation with a doctor about the information, then I presume there are two people in that conversation, regardless of physical location.
            If we are talking about patients ability to edit/upload pertinent information (keyword being pertinent) to the EMR, then I am all for that (as I wrote above or below, depending on how you sort).
            If we are talking about redefining the EMR into a lifestyle management tool, then I don’t think this is appropriate, since the medical record, as its name implies, serves a different purpose.

      • Israel Green-Hopkins

        Interesting comments, all – thanks for the discussion points. There is certainly hesitation on behalf of providers and patients with respect to sharing biometric data and one can see the multiple sides of this. For providers, many are already under significant constraints with respect to documentation and quality benchmarks – forcing one more batch of elements into the workflow (such as biometric data like FitBit, Jawbone or other) may cause further workloads despite unproven ROI. On the other hand, patients may want and should deserve positive feedback and reinforcement for their goal achievements and if their biometric devices are the most feasible/reasonable for them to convey this, then the health care community should embrace systems that can support it. Indeed, Peter is correct, there is a large cultural evolution that will have to affect all of us providers to further enable acceptance and utility of various forms of PGHD.

        • Peter Elias

          As I said, good discussion. I think the cultural evolution is well under way. I’m not sure providers or institutions will be able to keep up. Change happens incredibly slowly until it happens incredibly fast. Hang on!

  • http://www.mightycasey.com/ MightyCasey

    As someone who has had to fight – in some cases, literally – to have errors in my medical records, and in my family’s, corrected by the clinicians/providers who entered that wrong data, I’d have to say that I’d trust my Fitbit right now a lot more than I would the fat-fingering potential of whoever is entering the data I scribble on a form into [whomever's] EHR.

    “Wrong” data is rampant, since there are so many cooks in the data soup. The patient scribbles on a form (welcome to the mid-20th century), which form is then partially entered by an administration-level human. The full history is then entered by RNs, PAs, MDs, and other alphabetically-enabled folks. Why the h-e-double-hockey-sticks isn’t it yet possible for a patient to enter all necessary information into their record when they register with a primary care practice? A few practices allow for that, the vast majority do not.

    As lovely as it would be for my PCP to accept my Fitbit’s and other PGHD, the industry seems unable to efficiently capture the data collected within its own walls. If my supermarket and my bank can tell me all sorts of stuff about myself – how often I buy celery, or where I spend most of my disposable income – why the heck can’t my medical providers enable me to fully track my health online WITH their participation? Medicine is firmly planted in the 19th/20th centuries still, when it comes to interacting with its customers (patients). Are they any better at capturing and reporting care to the payers? I have to say I’m afraid they might not be, which might explain mucho about our still-broken medical system …

    • http://onhealthtech.blogspot.com Margalit Gur-Arie

      It may be very useful to notice that neither your bank, nor your supermarket allow you to input any data into their systems. They track everything you do in their system and they use the information to make you buy more stuff from them and their, unbeknownst to you, partners. I am very confident that your medical records will eventually be used for the same purpose. It’s just a matter of time now.

      • Peter Elias

        I’m not sure I agree, though I do understand your points. When I use my credit card at a restaurant, that transaction is automatically included in my bank’s system. I do not have to go to the bank with a receipt from the restaurant and hope that the teller enters it correctly.

        I see PGHD as somewhat analogous. Why shouldn’t the readings from my peak flow meter or glucometer automatically enter my health record? It is becoming generally accepted that home blood pressures are a better tool for assessing risk and response to treatment than BPs recorded in my clinician’s office. Why shouldn’t these be recorded? Since Mediterranean diet and aerobic exercise may be more powerful interventions than medication to prevent heart attack and stroke in the setting of mild-moderate hypertension or diabetes or hyperlipidemia, why shouldn’t they be recorded? (Currently, the patient describes their diet and exercise and the clinician turns that into one or two phrases. It is hard to imagine clinical data LESS reliable than this.)

        • http://onhealthtech.blogspot.com Margalit Gur-Arie

          The analogy of entering the receipt from the restaurant into the bank system is to entering lab orders/results or prescriptions automatically into the EHR. And frankly not even that, because the bank enters a number of dollars, not an itemized receipt in its system. It’s very easy to track one, and only one, arithmetic variable.

          Full blown PGHD would be analogous to automatically entering the itemized restaurant meal into some system that tracks your dietary situation.
          Of course, to be accurate, the exact number of calories in the mashed potatoes would need to be calculated, and something would need to first measure exactly what was in your plate at a molecular level and then track if you cleaned the plate or not, if you added salt or not, if you shared some with a friend or not, etc., so basically it would need to track the molecular structure of what went into your mouth and any eating disorders that may affect that data…..

          The sheer level of monitoring for these things to be of diagnostic quality is not conducive to human dignity.

          Now, I am not opposing glucose and BP equipment interfacing with EHRs (and they have those), or people keeping electronic diet or exercise logs accessible to their physicians, but let’s not get carried away to the point where there really is no need to even talk to a patient because the “data” is there.

          • southerndoc1

            Thank you.

          • Peter Elias

            I se a straw man, here: “Full blown PGHD would be analogous to automatically entering the itemized restaurant meal into some system that tracks your dietary situation.”

            I am not suggesting that, any more than I think the EHR should record every word that was said and every move made by everyone present in the OR during surgery.

            There’s plenty of room here for a middle ground.

            The patient comes in with a recorded chronology which they have taken the time to prepare. Why should he and I discuss his piece of paper and then either send it home with the patient (the data is not in the record for me to review later) or have me pay office staff to scan it in. Why not have the patient enter it directly and then have the patient and I edit it as a collaborative document?

          • http://onhealthtech.blogspot.com Margalit Gur-Arie

            I don’t see any reason why you should not be able to that, and some Portals will allow patients to upload documents (and structured data too).
            I do see a problem with a patient uploading/editing such chronology independent of your looking at it or even being aware that it is there. For better or worse, the medical record is a legal document. And if we decide that the EMR is to be patient controlled, then you will need another tool to maintain your own documentation in.
            This, by the way, is the same reason why you cannot “balance your checkbook” online on your own.

  • Adrian Gropper

    The all-seeing, all-knowing EHR has reached the limits of the Peter Principle. I want my physician to be part of my care team and deal with me and the data I generate or bring into the consultation. Otherwise, I’m being tracked like a tagged animal and second opinions are neutered.

    In the old days, paper records were signed by the lab or the physician and could be moved around under patient control with patient redaction. Involuntary and often inaccessible health records signed by some hospital are an invention of businesses that want to control the physician and the physician-patient relationship.

    It’s time to move on to technology that puts the patient ahead of the institution. Blue Button Plus is a good start.

  • southerndoc1

    Your banks let you enter data into your EFR (electronic financial record)?

    • http://www.mightycasey.com/ MightyCasey

      I can schedule transfers of data (money) between both the financial institutions I do business with – one a credit union, one a commercial bank. I can allow access to external applications (i.e. PayPal, Square, Mint) to either deposit or withdraw funds on a regularly scheduled or ad hoc basis. I can up/download information in PDF or MS Office formats for a variety of uses, including tax reporting.

      So, yeah, I can enter data in my EFR. I can interact with data in my EFR. I can productively use data in my EFR.

      My EHR? Not at all. Not once.

      • southerndoc1

        Not the same thing at all.

        What you’re doing is, as Ms. Gur-Arie says, comparable to getting test results or script refills: using the data that’s already there.

        What you can’t do is go in and say that your balance is $5,000, not $500, or report about a stock you sold seven years ago through a broker on the other side of the country, or give details of your bankruptcy 15 years ago, AND expect your personal banker to review all your additions to your EFR, and be legally liable for acting on all the data that you have entered.

    • Peter Elias

      Yes. I can set up automatic transfers. I can go directly into my account and transfer money from one account to another or pay bills. I can move money into my retirement account. In addition, I can set up a budget (e.g., revenue and expense goals) and see whenever I wish if I am on target with those goals and where and by how much.

      Obviously there are – and need to be – limitations. I cannot add a $5000 credit to my balance. (Though I can use an ATM and deposit $5000 without having to make an appointment with my bank.)

  • Israel Green-Hopkins

    Spot on Nancy – it’s very possible to have an evolution of very divided patients – those who are tech-savvy and desire the participation and those who are not/do not – and in between lies plenty of room for mistakes, inaccuracies and challenges for both patients and providers. It will be an evolving area and good to see that some providers are beginning to push boundaries in ways that promote safety and innovative practices. Thanks for your thoughtful comment.