January 28th, 2018

Apple Wants to Build an Electronic Medical Record — Here’s Where They Should Start

WPA Poster 1941, Library of Congress.

If you want to get a bunch of clinicians riled up — make them really mad — ask them to describe the problems with their current electronic medical record.

So when a company like Apple announces it plans to introduce an electronic medical record of sorts, we should rejoice, right?

This is the company, after all, that leveraged the Think Different and It Just Works slogans to boast about how easy their products are to use.

However, before we can celebrate that a famous tech giant will solve the EMR quagmire, remember that this is the same battlefield that has so far flummoxed Google and, to some extent, Microsoft.

Both those tech giants have struggled to bring sense, accuracy, and usefulness to the chaos of our medical information. The million lines of code written by their software engineers may live somewhere on hard drives and servers, but they are little more than historical records of a failed enterprise.

Think the software equivalent of Napoleon’s invasion of Russia. Ouch.

Given these past lukewarm approaches (some would call them out-and-out failures), I’d recommend that Apple not be too ambitious.

Instead, they should focus on information that is simple, highly useful, and lends itself perfectly to a unified electronic medical record.

Not only that, it’s ID-related. And every clinician and patient will benefit.

Apple, please build us a Universal Immunization Record.

Here are the features of this must-have service.

  1. Every person would have access to their record through a user-friendly secure access. No complicated password requirements.
  2. Information would instantly synchronize across multiple locations, akin to what Dropbox has perfected (and has been widely adopted by other cloud systems).
  3. Patients would grant access to their clinicians or other vaccine stakeholders (schools, occupational health departments, immigration offices), who would be able instantly to review and to update the information.
  4. It will be entirely agnostic about operating system, platform, or device.

Just think how care would be transformed with a widely adopted Universal Immunization Record.

  • No more ambiguity about school entry requirements or hassles about forms.
  • Which pneumococcal or meningococcal vaccine has been administered? Sudden clarity.
  • The problem of employee influenza vaccination tracking — solved.
  • International immunization records will easily cross borders — no passport required!
  • Pharmacy or health department or employer gave your patient a vaccine? No problem, now you can find it.
  • Using antibody titers as a marker of prior immunization? A thing of the past — and good riddance, it’s fraught with false negatives.
  • Possible vaccine repetition, “just in case”? Whether on the inpatient service, in outpatient clinics, in travel clinics, or in emergency rooms, say goodbye to this wasteful practice.

Apple, you might be tempted to add a few bells and whistles to this straightforward immunization record — say a list of medications, or allergies, or lab results, or surgical procedures, or (heaven forbid), doctor notes.

But you need to walk before you can run, so I strongly suggest you keep things as simple as possible.

Start with the vaccines. Master that first.

You could even argue that, with vaccines so globally valued (minus the anti-vaxxer fringe), that this Universal Immunization Record will be your company’s thank you gift to humankind for buying so many of your products.

A little something you can do with that nearly 1 trillion dollars you have lying around for a rainy day.

18 Responses to “Apple Wants to Build an Electronic Medical Record — Here’s Where They Should Start”

  1. Max Voysey says:

    They’ll probably do what ever they want – it’s their trillion $, and we’ll probably buy it.

  2. Carlos del Rio says:

    Like the idea. One additional feature very useful in global heslth would be an iris scan recognition of the individual that syncs to their immunization record so when out in the middle of nowhere you can see a kid and know what vaccines they are missing.

  3. Benjamin Jolley says:

    Are you familiar with Vaxigo? It is a software product that does effectively exactly what you describe. Take a look. Vaxigo.com

  4. Colin Rose says:

    No doubt at least a $trillion has already been spent by various governments and private companies on the holy grail of a universal digital medical record. So far all we have is a digital nightmare of a Tower of Babel of myriad systems serving a small number of users that cannot talk to each other. Just in one city there can be hundreds of deaf and dumb EMRs.

    Apple could easily spend another $trillion and never see a return on its investment. Who will pay for it? At the population level money and lives would be saved but the advantage to an individual buyer would not be obvious. The problem is that unlike eBay, Amazon or your bank there is no financial incentive to develop a universal digital medical record, even as limited as an immunization record, accessible by doctors and patients anywhere in the world in any language.

    The WHO already has a framework for a universal EMR but it is ignored. No government want to give up its digital sovereignty or pay the immense cost to implement it for its population.

  5. Maxwell A. Helfgott, MD says:

    Wow!! Just what we need: another gigantic database that will store unambiguous, invariate, minimally time-sensitive (gee, will my kid get to go to camp?), as a totally useless, non-EMR. It will do NOTHING to promote interoperability, NOTHING to enhance PHYSICIAN productivity, NOTHING to improve population health (but it sure improves the productivity of bean counters everywhere), and NOTHING that addresses the fundamental problem with ALL current EHR models, i.e., that the practice of medicine is not a transactional but an investigative, iterative, and semantic management activity for all its participants, from patients, PCPs, specialists, sub-specialists, super-specialists, and clinical and basic science researchers. The evidence that our EHR-saturated health care system is SAFER is rather underwhelming. The evidence that our health care system is providing better care because of our use of EHRs is on the thin side. The evidence that the privacy and security of our health care system’s information is better is a joke. The evidence that our health care system is less costly is non-existent.
    The PRACTICE of medicine subsumes the BUSINESS of medicine, which is what basically all current EHRs are adept (in varying degrees) at doing. It has taken 5-10 years for the medical community to wake up to the profound failure of the EHR initiative and the activities of the ONC (Office of the National Coordinator), which has been not only a costly failure, but has probably dug an intellectual pit which will take another 5-10 years to get out of. The exploding costs and complexity of health care were clearly and fully predicted years ago, but the profound failure of leadership in the government, academia, and organized medicine has led us to pray for a barren, dead-end solution from a tech giant whose only goal is to keep up the intrusive, privacy invading practices of their tech colleagues. Maybe FaceTime could come up with something similar.

    • Maxwell A. Helfgott, MD says:

      Kate,
      Could you explain to this old guy the connection between a universal allergy record and anti-microbial stewardship, the latter which means to me using antibiotics judiciously to decrease the development of resistant strains of microbes. Again, such allergy information should have been automatically interoperable in ALL EHRs, because it’s simple, straightforward, usually unambiguous (though too often not really allergy), and not subject to much description, interpretation, or re-analysis, which occupies most of my time and knowledge management.
      The failure to insist on such a fundamental data interoperability is just one example of the failure of the ONC and the current EHR discussion.

  6. Kate says:

    Just think what a universal allergy record could do to support antimicrobial stewardship!

    A hopeful allergy fellow

  7. Catherine Ratliff says:

    I read a lot of medical records because I’m a disability lawyer. EMRs make me crazy, like drowning in spaghetti. EMRs should be data bases that can be sorted by different variables. Narrative has its place but should be managed some way, color-coded perhaps so if its a cut and paste job ( which has its legitimate place) one can easily see what’s new or different.

    • Maxwell A. Helfgott, MD says:

      The sorting of variables is fundamentally a lexical, or word-based function, i.e., looking up all mentions of something like “mass in the fundus”. Which fundus? Eye, stomach, gallbladder, uterus? Humans have evolved and are trained to interpret the semantic value of such variables by the context in which we find them, such as other words, or if one is at a GI or Ophthalmology convention. Cutting and pasting ambiguous information does have some problems for safety, quality, and efficiency of health care. A descriptive narrative in medicine is called an H & P or progress note, and is functionally and logically different from a lab result or other evidence or diagnosis which is a PROVISIONAL explanation of the patient’s condition, based on the collected evidence.

  8. Philip Saccoccia Jr says:

    Dear Paul,
    Nice thought good idea. Did you send it to Apple CEO?
    May as well give the yahoos a chance to think about it.
    Cheers

  9. These posts are so negative. The problem with EHR’s is that the user interfaces are archaic. Apple is the UI expert of the world. The nuances of medical practice can be preserved in the medical record. just like the nuances of our online lives can be preserved. if an intuitive and fast method to record relevant data is available. It is not clear to me that the whole backend database needs to be rewritten. Data can be presented in a contextual fashion similar to what already occurs online. The problem is that hospital administrations would not pay for such a system as they do not understand the value of improved physician efficiency and more clinically relevant records..

  10. Maxwell A. Helfgott, MD says:

    Databases have structure such as one-o-one or one-to-many relationships. The quintessential, paradigmatic database is the one that relates a product on a shelf, with a price, a supplier, a product number, its quantity, etc. All those variables are pretty unambiguous, and updating one doesn’t necessarily require re-jiggling all the rest. Medical analysis is many orders of magnitude more complex, more ambiguous, more variable, and requires a massive commitment that most institutions are not willing to make to change and update their EHRs as new data and concepts are introduced into the practice of medicine. I think a proper EHR should be able to deal with the ever-growing complexity of medicine in real time, not be restricted to native language, i.e., English, definitions so that we could access Hungarian, Indonesion, and other medical data in a standardized, interoperable way, and should enhance physician productivity, which is the ONLY way to decrease costs, outside of coercion.

  11. David K says:

    Unfortunately, the more access patients have to their raw results, the more frantic screaming midnight phone calls we will get, asking why their liver was called “unremarkable” or why their percent monocyte count is 0.1 outside of normal.

  12. Mike Stein says:

    I agree with Robert. The focus now needs to be on us, the clinicians, in designing UIs that make us more efficient at managing data. Apple has gained a lot of experience in producing UIs that enable our smart devices to help us with our daily lives. I welcome any company that can do the same for us with EMRs.

  13. Anna says:

    Just a very naive question? Who is going to enter the data? Who will be responsible for accuracy?

  14. michèle halpern says:

    Sounds too nice to be realistic. I doubt patients will want their immunization records to be available to whomever. Just imagine ICE peaking into these records…

  15. Clancy Hughes says:

    Please please, date each page on any subsequent print out and separate the doctor’s notes from everyone else’s entries.
    Do not use undated problem lists, and separate insurance diagnoses from working and established clinical diagnosis.
    Let the computer take dictated narrative, and let the physician do the thinking.
    The computer on the other hand is good at statistics. Let the computer collect real-time statistics on the association between, symptomology, genomics, laboratory data and outcomes.
    Let the computer list alternative currently recommended treatments, but not listed on printouts
    Let the computer focus on diagnosis more than treatment.
    The computer has a better memory but poor judgment; the physician has better judgment and less reliable memory, simple but forgotten in the rush to AI.

Leave a Reply

Note: This is a moderated forum. By clicking on the "Submit Comment" button below, you agree to abide by the NEJM Journal Watch Terms of Use.

Our physician bloggers cannot respond to requests for personal medical advice, and recommend patients discuss health issues with their individual physicians.

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.