Why Facebook should be a template for electronic medical records

The Internet beta 0.5 version was nothing like it is today.  Back in the ’80s and the early ’90s the most common way to access the outside world was to use a phone modem to dial the number of a remote computer. These primitive servers usually ran DOS based software called a Bulletin Board System (BBS) which allowed users to post messages to each other. For the most part, these servers were isolated and did not communicate directly with each other. If you wanted to connect to a different BBS you literally had to hang up on one and call another one. There were usually no centralized servers that could link and share information among multiple BBSs. This sharing of information among multiple interconnected computers, servers, and networks is essentially what the modern internet is and once the world wide web started to proliferate in the mid to late ’90s the old BBSs became extinct. Facebook is a perfect example of modern Internet use. It’s essentially a vast central database with millions of users who can easily access information using multiple devices, upload and download data in multiple formats (text, links, pictures, video) from multiple sources and all from a single internet connection.

It sounds like Facebook would make a great starting template for a vast interconnected medical records system. But the reality is that the electronic medical record (EMR) industry is still stuck in the era of the BBS.

The similarities between modern EMRs and the BBS system are striking. Like many old Bulletin Board Systems the vast majority of EMR systems do not communicate with each other (nor even the outside world). Not only are they often incapable of communicating with another EMR or computer but even in 2012 most new EMRs don’t even have an option for sharing information with other systems! This is one of the biggest paradoxes and failures of almost all EMRs. Designed for an industry where the sharing of medical information among different facilities and health care providers is critical to the timely, effective, and safe delivery of medical care, the majority of these systems are designed to share information only within the limited confines of the specific facility or health care system that they serve. EMRs are essentially information islands cut off almost completely from direct contact with the rest of the interconnected world.

The system at the hospital where I work is a perfect example of this isolationist mentality.  In its current form, the hospital EMR cannot send or receive information from doctor’s offices, labs, or imaging centers outside of the actual facility. Acquiring old documentation still requires one or more phone calls, several human intermediates, a fax machine (40 year old technology) and open business hours (no luck if after office hours, on weekends, or holidays). Even worse is that the system can’t even communicate with older electronic systems within the same facility and has no capability to input and store faxes or scans in a format such as PDF for internal viewing. This means that we are still stuck with a hybrid EMR-paper chart system that is often more cumbersome and inefficient than using either system alone.

And how did it get to this pathetic point where a guy in rural Brazil can upload a picture of his strange rash to someone in China in real time but I can’t get critical medical information on a patient found unconscious until their doctor’s office opens the next business day? Given all of the concern about online privacy and hacked Facebook accounts you might think that the health care industry is cautiously avoiding systems where private information could be compromised (never mind that there has not been a massive collapse of the credit markets from widespread identity theft and fraud from 100s of million of Facebook users). The motivation of the health care industry in avoiding interconnected medical record systems is not to protect patients but rather, to protect itself.

Specifically, they want to protect themselves from competition. The majority of health care facilities and health care providers still think about medical records the same way they did 100 years ago – as property or proprietary information. A physical medical chart is considered to be the property of the facility or provider who generated the chart even though the patient is considered to be the “owner” of the information contained within the chart. A Paper chart is usually stored at the same facility where it was created and patients often prefer to return to the same facility or provider if for no other reason then because “that’s where my records are.” Changing facilities or providers can often be a problem for patients who must go through the bureaucratic hassles of making a formal request for their records and then having to pick them up and hand-deliver them to a new office or provider. This system is effectively a disincentive for patients who wish to change providers or health care facilities and is, ironically, a barrier to the sharing of information.

Not surprisingly, the electronic version of the medical record system continues this same isolationist mentality. Even though digital information replaces paper folders and charts, electronic records are usually restricted to the system that generated them. Most EMRs are designed to be run on a server or computer that physically exists within the same office or facility and without any direct connection to the outside world. The only way to share information is the old fashioned way – i.e. the electronic record must be printed out and either faxed or hand-delivered to another office or facility. In this way, most EMRs are little more than a hard drive replacing a filing cabinet. But wait, that’s not all. Most EMRs are themselves proprietary software. Most of the expensive systems in particular are not compatible with transferring information to other EMRs without extensive and expensive modifications. If all EMRs could communicate with each other equally then providers would just purchase the cheapest ones thus damaging the finances of the industry. Right?

Wrong. Business is ripe with examples where standardization and compatibility among different products has lead to expanded market size and share even among competitors. The electronics industry is just one example. Another example is the Internet and web browsers. If Internet Explorer were only able to access web sites running Windows server software, then the massive growth and success of the Internet would have been severely blunted. Instead, the industry got together and decided on a common language (HTML) that is usable by any browser. Obviously industry standards can be voluntary or by government mandate. So the question is; why not make all EMRs capable of communicating and sharing information with each other? The Federal government recently had the chance to do so and severely blew it.

The Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 provided financial incentives for early adopters of electronic health records and even provided definitions of “meaningful use” and what information an EMR should be able to record and store. But the single biggest failure of HITECH was that it did not define standards of compatibility nor even require compatibility among any of the thousands of different electronic medical record systems. Hundreds of millions if not billions of tax dollars will now go towards further entrenchment of the current isolationist proprietary EMR industry.  And this is not good for patients or consumers either. In theory, if all electronic medical records were freely and easily transferable among different providers and facilities, the artificial barriers for patients to changing providers and facilities wound be gone and providers and facilities would have to compete based solely on the quality and efficiency of their services rather then relying on holding medical records hostage in order to generate return business.

However, improved competition among health care providers is not the primary aim of open and efficient transfer of health related information. Accurate and up to date health care records that are easily and universally accessible have the potential to improve the safety, quality, and efficiency of health care delivery. Costs can be decreased by reducing the need for repeat labs and tests. Older patients in particular would benefit from accurate medication lists and lists of prior drug reactions and allergies to reduce the possibility that providers will prescribe medications that could interact with a patient’s other medications or be contraindicated for any one of their medical conditions.  Ensuring that open and free communication of medical information exists between EMRs appears to be beyond the scope of the current records industry. This is something that only the Federal government will be able to do when and if it decides to get serious about modernizing health informatics.

Chris Rangel is an internal medicine physician who blogs at RangelMD.com.

Submit a guest post and be heard on social media’s leading physician voice.

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

  • Chris OhMD

    Great article. I’m running into same problems using two EMRs on staff at 2 hospitals – eCW and EPIC – both great products being used by two competitive hospital systems – both completely ignoring the other exists and trying to exert its own IT influence as far as possible geographically. I’ve always thought that healthcare IT needs to learn from SoME concepts outlined above but I think it really comes down to security and the (un)willingness to share – i.e. political reasons. It is very easy to share data – XML, HL7 – standards exist already but let’s take this example. Let’s say an eCW doc wants to share patient data with EPIC – how would that occur? This is different from email where hotmail sends and email to gmail – in the case of med rec the exchange of data has to occur in a “secure” environment – i.e. a portal. But then who’s portal? eCW’s portal or EPIC portal? No one agrees because they both want their own portal. The problem is not technicals but political will and the (un)willingness to share among competing hospital systems.

  • http://twitter.com/Somnonaut Claude Albertario

    This is all because there is a profit motive in all of medicine. Take that out at the source (insurance companies) and you will change the whole scenario. We cannot break free of the stranglehold the insurance industry has over our lives. We must kill the insurance companies.

  • Fred Dempster

    Thanks for addressing a few questions I have had. I have still stuck in my head the concept that HIE solves for this, but as I got involved with EMR/EHR I saw the complexities. NOTE: Do not follow the new Facebook model – the timeline – as for the average user they have added complexity and reduced performance to something that was simple to use.

  • Maged N. Kamel Boulos

    In fact the Timeline model lends itself very well to health records; chronological view is very common in the latter to follow the history of a patient/disease condition. Timeline has been used by companies on Facebook to tell their stories since they were founded and their corporate milestone. Very clever! (I hope my other comment to this kevinmd.com page gets approved and not censored/deleted :-)

  • Chris OhMD

    I agree that the timeline feature is the best facebook feature – when I open my eCW or EPIC chart – I need to refresh my patients course of events for each disease: e.g. DM when? LDL 150 when? started simva when? f/u Lipid/LFT when? etc. EMRs should definitely integrate timeline feature.

    When I get a test result now my workflow is as follows: go back, look at past note to remind myself what I told pt – when was follow up? then I call pt about test result – remind him,/her about follow up etc, document I spoket to pt – so many window opening and closing for one task which should be done in one timeline window – I recently spoke to eCW about this cumbersome feature in eCW. I don’t think any other EMR (including EPIC) is better.

  • http://twitter.com/Serrenity Drew McAdams

    See, I’m not sure how I feel about this. I’m an EMR consultant and I work almost exclusively with Epic (disclaimer).  I feel like the assertion that no one wants to play with anyone else is made in a bubble without really understanding a lot of the issues.  Interfacing information between systems using HL7 is a huge part of any implementation — and the sharing of information is paramount for any organization.  While I can’t speak for every organization, I do know that in my experience, Epic has never refused to interface with another system, or done anything to minimize the amount of information that can be passed back and forth. 

    9 times out of 10, there are technical limitations or considerations that play into when information can be sent between systems and how.  Part of the IT team’s role (which often plays with all system) is to determine which systems need to talk to each and with what information based on the requirements for patient care.  The rub of the situation is the different software products work differently, store data differently, and just generally function differently.  Using a metaphor like connection to different BBSs is flawed because it implies that the underlying technology between different EMRs is the same — and it’s not.  The challenge is not the same. 

    It’s better represented by thinking of different EMRs as different cars.  A car is a car, right?  By the logic put forth here, I should be able to take one part from my Honda and plug it, without issue, into my Chevy.  That happens, right?  Except it doesn’t work like that.  True, both vehicles have the same end result, but they go about getting to that result in subtley different ways.  So to get two systems to talk to each, I have to create an HL7 interface. 

    The HL7 interface is the equivalent of a translator.  Between two different systems.  So we can think about it as a real time modification piece for the parts leaving our honda so they can be used in our Chevy.  If I just send them across without making adjustments, my Honda parts won’t fit my Chevy car — so I have to put them true a translation table to make minor tweaks to my parts to fit the Chevy.  Often times, we have to make permanent changes to the Honda and the Chevy themselves to allow parts to be used between both cars. 

    I hope that metaphor helped a bit.  That was a pretty basic scenario in terms of making two systems talk to each other.  It’s not that any one system doesn’t want to share information, its that it’s technically challenging to do so in some cases.  Additionally, you aren’t factoring in provider requirements on how things have to work between different organizations (Because believe me, it does vary … a lot), nor are you taking into account HIPAA legislation regulating who can see my medical information when.  What would be a workflow for a patient to grant access to their “Clinical” timelines? 

    I agree thats there’s a ton of room for improvement and innovation within the EMR world, but saying that all EMRs want to be the only kid of hte block and that they all money hungry is short-sighted, and leaves out an entire aspect of this problem that influences more than what you perceive it to. 

    • http://profile.yahoo.com/6VZNUBBOZ44BFKI357VDKLB3I4 Mike

      I like your analogy but this is why I have said repeatedly that EMR is not ready for primetime. The key benefit, interoperability, is not there. Now if your organization wants to use it, that’s fine. But forcing all hospitals to adopt one with the premise that we will all be able to talk to another is another example of the waste in our healthcare resources. 

  • http://www.facebook.com/profile.php?id=1592238231 Scot Silverstein

    “And how did it get to this pathetic point where a guy in rural
    Brazil can upload a picture of his strange rash to someone in China in
    real time but I can’t get critical medical information on a patient
    found unconscious until their doctor’s office opens the next business
    day?”

    This is what is known as the false dilemma fallacy:

    1. Uploading and sharing an image is a triviality compared to a set of relational database records of complex structure and varied content. By way of irrelevant comparisons, buying a toaster at $50 of course, is a bargain compared to buying a Ferrari at $150,000… but so what?

    2. “Patient found unconscious” … without ID and alone is as rare as hen’s teeth.

    3. “Patient found unconscious” –> E.D. —> most likely even
    if the patient is unidentified and unaccompanied, doctors can use their
    clinical skills to evaluate and stabilize.

    4. If patient has ID and/or friends & relatives, that can be
    used to get in touch with their doctor(s) and/or answering service
    and/or hospitals where they’ve been, where more information will be
    available, if even by someone having to go to the doctor’s office and/or
    contacting the hospital(s) where patient has been and having records
    FAXed or just read to them.

    Spending $100 billion++ so that rare scenarios such as the above,
    that can be managed via traditional means, can be managed by cybernetics
    is, on its face, a very dubious justification for precious healthcare $ that can be spent elsewhere.

    • Kaushal Saxena

      Scot, my first disagreement is that the scenarios you have just mentioned are not rare, but quite frequent. In India, where I reside, it is not rare to find unconscious, sick, injured, and alone person. A large section of Indian population does not even own an ID, let alone carrying it with them all the time. I am pretty certain that many geographies including developed nations face such issues very frequently.
      Secondly, the whole philosophy behind accessibility of EMR/EHR is to improve the medical care and reduce the risk involved based on the history of treatment. The EMR/EHR system also quickens the pace at which care can be delivered by giving an objective assessment of reasons behind the patients’ failing condition. The evaluation and stabilization of patient, as you mentioned, becomes more effective, faster, and less risky.
      Thirdly, why would you not want a faster,easier EHR system. Suppose you were a security personal who checks identity of a person at an entrance. Would it be not faster and convenient if the person carries an electronic ID, which stored all background information?; or would it faster and convenient if that same person gives you a paper id which you need to validate by calling the ID issuing authority and validate all background information.

  • Stefani Daniels

    I don’t think its a question of interfaces availability or of the vendors willingness to build one and the clients willingness to pay for it.  More importantly, the issue I have with most vendor EHRs is that they’ve simply replicated the paper record.  Theres a section for doctors, a section for physical therapy, a section for nursing, etc etc.  the EHR vendors have perpetuated the silo functioning of the hospital.  As a physiciah, with just 4 – 6 clicks, I should get a FB ‘timeline’ of my patient – not only whats been happening with the medical treatment plan, but I should be able to see what social services is recommending, what nursing has discovered and is recommending, what my consultants are recommending, the outcome of my patients’ physical therapy session.  As it is now, I have to hunt around to ‘find’ each disciplines place in the EHR.  Do I do it….take a guess.

  • Kaushal Saxena

    Ownership of the medical records has been a key issue in EMR systems communicating with each other. In definition EMR is something which is owned by the care facility. In theory, a superset of EMR is EHR which is owned by the patient. A patient cannot easily switch provider unless EHRs are centralized and updated in a way similar to ‘Timelines’ on Facebook, in which multiple source update the profile information about a person. The information on timeline can be search and queried using important events and history very easily. Another feature is that records can find relevance to another and show up as a clubbed record. For example, If I were to have a throat infection twice in a two year period but no reason could have been determined immediately. This record was then reported to my timeline. Later, the throat infections records show up as clubbed results on my timeline, indicating that I happen to have a seasonal allergy when I am at a certain place. The intuitiveness of such system will be tremendous and very helpful to doctors.

  • Thomas Munghono

    Each time we technocrats refuse to adjust and adapt to the contemporary context, we risk becoming irrelevant ie, salt with savour; tree without fruit.

  • KarenConway

    Thank you for articulating the danger of repeating the integration mistakes of the past.  This issue extends beyond EMRS to how we might use or misuse the potential of cloud technology going forward. Whether it’s the federal government making policy or healthcare delivery organizations looking at how to best work together to be accountable for the care of specific populations, we need to coordinate how we think about current and future needs. If we do not, we are destined to recreate in the cloud the same integration problems we experience with hardware and software deployed on premise. The problem is, we still treat technology implementations in healthcare as a project, not a strategy.  As a result, we implement one system to address one problem, another to address another need…whether it is clinical, business or operational…without considering the ability of those systems to communicate with one another.  As we move to a healthcare system that has to deliver VALUE, which is measured by the ability to provide quality care in an efficient and cost-effective manner, we need more than EMRs to talk to one another.  We need to truly understand and report on what contributes to both quality and cost.  That data exists in many systems – clinical, supply chain, financial – and with many organizations.  The good news is, better coordinated care almost always delivers better quality at lower cost.  The use of cloud services can help increase that coordination and sharing of data, once we get over the security fears, which I believe, as the good doctor states, are more a function of old beliefs that data is property and proprietary and needs to be protected on premise.  But if you look at the facts, more security breaches occur when data is stored on hardware, on premise, than in the cloud. 

    If others agree, we are still faced with a bit of a “Prisoner’s Dilemma.”  Hospitals, physician practices and other care delivery organizations, even medical device manufacturers and payers, recognize that working more closely together can result in better patient care, but it’s hard to collaborate on technology strategy and take a more long-term, holistic approach when so many are focused on just trying to comply with meaningful use and other regulations or just keeping their doors open.  But sometimes, it might make sense to look up once and awhile and do a little cloud gazing with your colleagues.