What if medical records worked like Wikipedia?

I’ve been thinking about EMRs, electronic medical records, lately. It’s a subject, despite some professional experience, I don’t feel particularly close to. In fact, if anything, they are a source of consternation. As an industry insider, I see them as an expensive albatross around our collective neck. As a human centered design adviser, I see them as an encumbrance for both providers and patients. And, as a patient I see them largely as an opaque blob of data about me with a placating window in the form of a portal.

Which makes me wonder, am I obsessed with EMRs lately?

One of the reasons is certainly my personal interest in technology. And, while I don’t work in health IT, it’s natural to draw some connections. For instance, Wikipedia is consistently in among the top 10 most visited internet sites (it is currently number 6). And, say what you will about citing Wikipedia, but a 2010 study found it as accurate as Britannica. Google trusts Wikipedia enough to use it as the primary source for its knowledge graph cards; and we’ve all settled a bar bet by finding some fact where a Wikipedia article is the canonical answer.

The secret sauce for Wikipedia is in it’s roots. Literally, the root of its name, wiki, describes the underlying structure. Wikis were the internet’a solution to knowledge bases – large repositories of information about a process or thing. Companies had been using knowledge base software for years. Traditionally, a central maintainer, often a sort of corporate librarian, curated information, such as common answers to customer questions, so customer service reps could find it quickly.

Wikis democratize the knowledge base by allowing anyone to edit an entry. If you work for a company which sells widgets and you discover a new way to service the widget, you simply amend or append to the record in the corporate wiki. But what about the corporate librarian, they all cried. Except, no body cried.

It turns out, the network effect and the wisdom of crowds produce richer, more accurate databases of knowledge when the literal barrier to entity is removed. Make it easy for anyone to input knowledge, and the database and its accuracy grow. And so it came to be, since anyone can edit almost any entry in the largest encyclopedia the world has ever known, Wikipedia is remarkably current and accurate.

So I wonder … what if medical records worked like Wikipedia?

What if, my record lived on some commonly accessible platform; not open to anyone, but accessible by my providers and I? Maybe we have to do some kind of online handshake to mutually access it.

What if we could both edit the record, at the same time? My doctors could put in their notes and I could add my own. Or I could edit theirs. And they could edit mine.

Some readers may have concerns about the records’ integrity but as patient advocacy expert Trisha Torrey points out reviewing our own medical records can help spot and fix errors. And, as we know from Wikipedia, more eyes and contributors on a record increase its accuracy and reliability.

Another important lesson from Wikipedia is the idea of revision log, which Wikipedia calls page history. Any registered user can make edits to almost any record in Wikipedia’s vast online encyclopedia. Every time an edit is made the changes are logged, including the name of the user who made them. Anyone can review the changes and roll back some or all of them, or make additional changes of their own.

Imagine a medical record platform where patients can review the entries made by a doctor, and if appropriate make additions at it or even changes. For instance, after reviewing notes from my last physical, I discovered a small unimportant inaccuracy in my record. I take vitamin D supplements, and in the record, it was noted that I take vitamin E. Big deal? Probably not, but what if it was related to a prescription medicine? Providers are human and, as we know, to err is human, but by allowing patients to review and edit their own records, they would be able to fix errors.

A Wikipedia-style EMR would also better allow for patient-contributed data. There are often symptoms, observations or measurements which patients observe outside of the timeframe of a visit with their doctor: a week of poor sleep, a month of improving blood pressure measurements, an off-again, on-again skin rash. These kinds of things may not even warrant a phone call, but wouldn’t it be nice to log them directly?

Finally, and this may perhaps be my strongest argument for a Wikipedia-style EMR, we’ve got to do something about data exchange. Color me cynical, but I’m not convinced the health information exchanges (HIEs), offered by the major EMR vendors as well as technology giants such as Oracle, are the answer. Each EMR vendor has a financial incentive to keep their data in a proprietary format. Further, their customers are, by definition, the providers, not patients.

No, what we need instead is a common, centrally accessible platform where patients and providers have parity, equal footing. No one party’s observations, notes, measurements, or data trumps the other. A common platform would make it easier for different providers to openly collaborate, in front of the patient, virtually, in a common record. Your specialist could be literally updating the same records which you, the patient, are adding to while your primary care doctor is also reviewing and making edits. 

And there’s an extra credit reason we need a Wikipedia-style EMR. It doesn’t just promote or enable patient empowerment, it demands it. Owning our own data requires responsibility. It becomes the patient’s garden to tend. And its our right to tend those gardens. Stephen Ross and Chen-Tan Lin, writing in JAMIA, concur:

Overall, studies suggest the potential for modest benefits (for instance, in enhancing doctor-patient communication). Risks (for instance, increasing patient worry or confusion) appear to be minimal in medical patients.

This doesn’t have to be a pie-in-the-sky dream either. Someone could build a WikiEMR today. The platform which runs Wikipedia is called Wikimedia. In fact, it would likely meet all of the meaningful use stage 1 requirements … except one, and could be regarded as HITECH-compliant:

  • Anyone can download it, or install it on a hosted server. It can use the same strong SSL encryption which protects Epic, Allscripts, Athena and McKesson platforms.
  • It is free (a substantial discount off the price tags stuck to the EMR giants).
  • It provides user access audits and record edit history.
  • It is accessible via mobile and desktop.
  • It can use multi-factor authentication.
  • Wikimarkup, the simple language used on Wikimedia sites, supports mathmatical calculations so a WikiEMR could do unit conversions, Boolean checks, and data aggregation and reporting (including graphing).
  • It can generate reports for the MU core measures including abstracting 14 core objectives, 5 out of 10 of the menu objectives, etc.

What one, small, requirement is missing? As far as I know, Wikimedia is not a certified EMR. Anyone want to start a fund drive?

Nick Dawson is principal, better.  He blogs at NickDawson.net, where this article originally appeared.

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

  • Patient Kit

    If both doctors and patients ever have equal access to the whole patient’s medical record, in any format, it has to remain extremely clear which notes are the doctor’s and which notes are the patient’s. Adding new notes, including corrective notes, is very different than editing each other’s notes. Changing a medical record could cause many potential serious problems, both medical and legal.. The complete history must remain crystal clear and easily accessible.

    • JR

      I think this is one place where wiki software excels. There is a complete record of every change along with a record of who made the change. It’s easy to revert to an old version. Click on “view history” to see the list of changes over time, compare two versions…

      I do think there should be records only the physician can modify, such as tests (the patient maybe can comment, but not change the results), or physician’s findings on an exam, etc.

      • guest

        So, in order for someone to have a comprehensive idea of a patient’s medical record, he or she would have to pull up all revisions and go through each one to get some sense of what had been added, what had been removed, who it was doing the additions or removals, and put it all together into an accurate formulation of the patient’s history? Here’s a news flash: most of us barely have time to pull up and review old records on a patient. Having to review revisions would be impossible. Lack of time to document adequately is one of the reasons we have inaccurate records in the first place. Layering in additional levels of tasks/complexity will make things worse, not better. Although I will concede there are some patients who might feel better about their treatment if they could contribute to their medical record.

        • JR

          How often does a physician go through 50 years of a patients medical history?

          • guest

            When we get new admissions we review their records going back as far as our system allows, which is into the 1990′s. And then we also request and review records from any other facility we know that the patient has been admitted to. Our state hospital is particularly good about sending records over, but I have gotten terrifically helpful records from a number of out of state facilities.

          • JR

            That’s surprising. I’ve never had a doctor request any of my previous records before.

          • guest

            That’s because most docs carry a much bigger patient load than I do. I am capped at 10 inpatients per day and have a full-time social worker whose job it is to handle administrative tasks and being first contact for family members. It frees up a lot of my time so I can do my job the right way.

          • Patient Kit

            Ah! You must be “psychiatrist guest”! ;-)

          • ninguem

            Oh, when I’m in the nursing home, I know I need fast access to that note on the quality of Grandma’s bowel movement on Easter 2011.

            I can usually find that easier than the medication list.

        • http://www.nickdawson.net Nick

          “Here’s a news flash: most of us barely have time to pull up and review old records on a patient.”

          And that’s certainly a huge problem too…but arguably its a separate problem (albeit one which needs critical and earnest attention).

          Though, one may make the argument, if patients are allowed into the walled kingdom they may be increasingly self-supporting and ‘empowered’ and thus need less time from —or make better use of the time with —their providers. Win/win?

          • guest

            Again, I suggest that you experiment with shared work meeting notes with your clients and let us know your thoughts after that on the possible advantages and disadvantages of a shared medical record.

            If it’s a win-win for you, perhaps it could be managed to be a win-win for doctors and their patients.

            Personally, however, I am already contending with a workday that has expanded by about an hour due to the addition of EPIC to my workflow. I am not readily inclined to be enthusiastic about yet another intrusion of technology, designed and promoted by non-physicians, into the work I do.

    • http://www.nickdawson.net Nick

      Great point – in a collaborative platform, the UI should probably delineate who made the contributions. This would also include data contributions from connected devices, non-traditional providers, lab results, etc. On the other hand, does knowing the source allow for too much attribution bias? Might providers discount something relevant because it was a patient contribution?

      • Patient Kit

        Medical records should “probably” delineate who made the contributions? I have to disagree and say that medical records should “definitely” delineate who made the contributions. I’m all for creating good doctor-patient relationships but I see no upside to medical records that do not clearly identify the sources. Sourceless medical records sound like a recipe for disaster to me.

        I mean, I consider myself to be a “good”, proactive, educated patient. I consider myself to be my doc’s equal as a human being but I don’t consider myself to be equal to being a doctor. I may know myself, my experience and my situation better than my doc does but I don’t have a medical education. We bring different things to the table in the relationship and it needs to be very clear who brought what info.

        I’m trying to wrap my shuddering mind around what my elderly mother’s medical record would look like if she was allowed to edit it anonymously and we couldn’t tell what info was from her and what was from her doc.

        This might — and that is a huge maybe — work for some select patients, but not for most for a variety of reasons. Nope. I can’t even imagine any good that would come of unsourced medical records.

        Why would you not want to know what info came from you and what info came from your doc?

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

    Well, let’s see… Can Wikimedia support electronic data exchange for lab, medications, and imaging orders and results? Can it incorporate First Databank, SNOMED, ICD-10, CPT, LOINC and a variety of vocabularies to be used automatically by those edits? Can it incorporate structured clinical decision support, and immunizations, bio-surveillance, and cancer reporting interfaces? Can it support X12 EDI for claims, payments and real time eligibility checks? Can it support interfaces to equipment, such as BP, EKG, spirometry, and uploads from home monitoring devices? And if it can be programmed to do all of the above, what will be the final price for such product, including maintenance and support for all interfaces, and timely updates to evidence-based guidelines, formularies, medical necessity rules and fee schedules? Can its group edited records be acceptable in a court of law as the authoritative medical record, or in an ER before dispensing narcotics?

    Bottom line, we are not looking to replace a piece of paper anymore….

    • http://www.nickdawson.net Nick

      The short version is: yes, the wiki platform can do those things. But the longer answer is that this is an allegory and not a direct proposal to implement wikimedia as an EMR platform. The point is that today, its extremely expensive and difficult to get even limited interoperability between the same EMR platform across two different organizations, let alone any user/patient contributed data. And that’s absurdly antiquated considering that, yes, a free open source knowledge base has solved that problem for 10+ years. If we desire empowerment, we cannot build and maintain a walled kingdom.
      And, frankly, the VA has done this with Vista, it’s open source, free EMR….pretty good model, and they are incorporating the OpenNotes standard to allow shared record editing. So yeah, it can be done in a cost effective, product way.

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

        Nick, Vista and the allegorical Wikimedia have as much in common as Epic and Wikimedia. The fact that one is open source and the other isn’t does not make much difference.
        As far as technology is concerned, interoperability does not pose any difficulties, and it is rampant wherever there is a business need for it and an ROI. It is a myth that “EMRs can’t talk to each other”. They can, but they choose not to (just like B of A chooses not to share client information with Citigroup).
        Opening the notes for patients to see and even edit (addenda), takes the proverbial five minutes of programming in most EMRs. Should everybody do it then? Yeah, why not… Will it make a difference? I seriously doubt it….

  • QQQ

    “What if medical records worked like Wikipedia?”

    Wikipedia isn’t a reliable source for information. You want that type of Wikipedia confidence in your medical records???

    • Brad White

      Wikipedia is as reliable as Britannica. Next time read the article before you comment.

    • http://www.nickdawson.net Nick

      Sure it is!
      And, it’s frequently one of the top three results patients are likely to find with researching symptoms or a diagnosis. Arguably, its already a significant part of our medical culture, even if we don’t acknowledge it as such.

  • guest

    Every day I have patients provide me with information which is objectively inaccurate, when I check it against their written records or even against objective observation. I think it’s a mistake to assume that the patient is the ultimate source of all accurate information regarding his or her condition and treatment. Where do you draw the line between the doctor’s perception, and the patient’s perception of a symptom or a sign? The patient’s perception is not always necessarily what is objectively accruate. I have had numerous patients report to me that they were experiencing intolerable anxiety or depression or pain but objective 24 hour observation of their behavior in an inpatient setting revealed them to be sleeping well, socializing comfortably and generally enjoying themselves. Likewise, I have had other patients stoutly deny experiencing auditory hallucinations when it was clear to everyone around them that they were hearing things that the rest of us could not hear. Personally, I believe that allowing some patients to correct their medical records, Wiki-style, would result in a far less accurate record.

    • guest

      I would be more inclined to believe the patient on this one. Often times the medical records are updated on multiple patients in one sitting, and staff are not going to know who had anxiety and who didn’t unless the patient is making a big deal about it. Also, just because a nurse peeks her head in the door and sees a patient lying there with their back to the door doesn’t mean the patient is sleeping. They may very well be wide awake.

      • rbthe4th2

        When you’ve suffered because a doctor was wrong and although you made additions to the records, they only believe the doctor, what would you think?

        • guest

          Doctors are sometimes wrong, and I get your point. I think the patient should be able to restrict who sees what, and that way if the doctor is wrong or writes inflammatory information, the patient can restrict other doctors from seeing it. I think that would make more sense then editing someone else’s opinion.

          • rbthe4th2

            That is an excellent idea. The only problem was the surgeon who I spoke of causing me the problems, stated, ‘I’ll put it in your record so other doctors can see it and wonder why”. I would think other doctors would wonder why you wouldn’t let them see it.

      • http://www.nickdawson.net Nick

        “Still, I don’t think patients should be able to edit medical staff notes, no matter how inaccurate the notes may be.”

        What if it is a non-destructive edit? Meaning, what if the original note remains and can be reverted? And/or, the patient’s edits and additions are store separately, so neither the patient nor provider can clobber each other’s contributions to the records?

        • guest

          I think (I could be wrong on this) that under HIPAA patients can ask for an addendum to the medical record to correct inaccurate information. That’s okay, in my opinion, but altering what someone else has written would be a mess. What if the patient makes up stuff? Yikes.

  • http://www.ronsmithmd.com/ Ron Smith

    Hi, Nick.

    “What if we could both edit the record, at the same time? My doctors could put in their notes and I could add my own. Or I could edit theirs. And they could edit mine.”

    This would be a medico-legal disaster unless the simple act of allowing this kind of access, automatically excludes patients from making any legal claims whatsoever.

    In a world of perfect hearts, where all could be trusted, this would work. But healthcare issues would be foreign to that world, and this discussion all would be moot.

    Warmest regards,

    Ron Smith, MD
    www (adot) ronsmithmd (adot) com

    • http://www.nickdawson.net Nick

      “This would be a medico-legal disaster unless the simple act of allowing this kind of access, automatically excludes patients from making any legal claims whatsoever.”

      That’s where an audit logging and revision history feature is critical, right?

      • http://www.ronsmithmd.com/ Ron Smith

        The problem is that if patients have this kind of access the legal validity of a claim based on them seems highly questionable.

        It is no different than reference Wikipedia in a scholarly article as something that carries the weight of scholarly expertise. It is well documented how that edits are made to Wikipedia articles by those with agendas and which will lie in print to propagate those agendas.

        Just having a patient’s comments about the medical record be considered as part of the medical record is not good for patients or providers.

        Warmest regards,

        Ron Smith, MD
        www (adot) ronsmithmd (adot) com

  • azmd

    Well, it’s an interesting idea. I wonder if it would work in other industries for example healthcare consulting? Do you think it would be useful, when you have a meeting with a client, for them to have access to your notes from that meeting, and the ability to go in after the meeting and change the notes you have taken? Do you routinely send clients a copy of your notes and then spend time with them going over the notes and making corrections as they request? Why not?

    I guess I would say that if the Wiki approach is being used in other fields to create a professional’s working notes of other types of meetings, then we should look at those other systems and learn from them. If it isn’t, why not, and why would we start experimenting with this approach with something as medically and legally sensitive as a medical record?

    • http://www.nickdawson.net Nick

      “I wonder if it would work in other industries for example healthcare consulting? ”

      I’m not a consultant, per se, but I love this idea! Certainly makes sense to me to share and co-edit a work product one is paid to advise on.

      “If it isn’t, why not, and why would we start experimenting with this approach with something as medically and legally sensitive as a medical record?”

      Certainly, its not new for workgroups to share common documents everyone can edit – look at the up take of Google Docs and shared editing platforms like Office 360. And a lot of sophisticated project management tools, like Base Camp, allow vendors to invite clients into a shared space. I’d argue this is pretty much the de facto way to operate on most work projects today. And I love your idea of looking at what works in those platforms and what we might learn from collaborative editing suites!

      • guest

        O.K., well you start doing it, and let us know how it all works out and then we can think about applying what you (and others) have learned to trying out a shared medical record.

  • wheatthins

    For all the naysayers: the concept of the Wiki EMR isn’t that different than the comment section here on Kevin MD. We read the article from the primary source, then scan the comments. Even if the comments are erroneous, we get a baseline perspective of the arguments.