Matthew Holt

EMR use: on the steep part of the S curve, or being replaced by a new idea?

Ten plus years ago, I was giving talks suggesting that at some point relatively soon the EMR was going to become a reality. In 1999, at Harris Interactive I actually got the chance to launch a study which I hoped was going to soon show a relatively steep growth in EMR use in physicians’ practices. (The study was called Computing in the Physician’s Practice). Sadly because the study wasn’t a huge financial success and because I wandered off to do other things, it was only fielded in late 1999 and early 2001.

I developed a very complex metric of what computing was used in the physicians office, and in particular what doctors actually did and planned to do electronically. So for example, did they order lab tests electronically, did they look up test results or PACS images online, did they take notes electronically, etc, etc. By the second fielding, the questionnaire changed a little and sadly Harris started asking doctors if they were using an EMR—a pretty useless question as they didn’t define what it meant.

The results at that stage weren’t too encouraging. Because new data is out today, I went back and took a look. Around 25% of physicians were looking at some individual patient data using some type of a computer back then, but only around 8% were taking any clinical notes using one (pens and micro-cassettes were the order of the day). The juicy data is on page 80 onwards of the slideshare deck. Obviously, the total percentage of doctors using an all singing-all dancing EMR was very low.

For those of you really interested in the archaeology of the EMR I’ve put the 1999 study results up on SlideShare (Yes, it was proprietary then, but I can’t imagine anyone cares much about that now!).

So nine years ago you had about a quarter of doctors who at least came within sniffing distance of something like an EMR. And that was already a low number in terms of international comparisons. The docs in the N countries (New Zealand, Netherlands and Norway) were already on their way to 100% penetration then.

Fast forward to this past Friday.

CDC (which for some reason has taken on the mantle to find out
what’s what in physician computing use) has done a study of about 1,200
docs this year—updating the one they did in 2006. Only preliminary top line data is out but it is encouraging:

In the 2008 mail survey, 38.4% of the physicians reported using
full or partial EMR systems, not including billing records, in their
office-based practices. About 20.4% reported using a system described
as minimally functional and including the following features: orders
for prescriptions, orders for tests, viewing laboratory or imaging
results, and clinical notes. Comparable figures for the 2006 NAMCS, the
latest available for the full survey, were 29.2% and 12.4%,
respectively.

So now I’m going to make a giant leap and say that my 1999
data—the 28% who said that they looked at some clinical data about an
individual patient on a computer—is very roughly equivalent to
the 29% in 2006 who said that they were using a full or partial EMR
system. That suggests that not much changed between 2000 and 2006. (Yes,
I know that I don’t know the definitions in the CDC survey, but I have
asked that they respond to this post….)

So now, two years after 2006 we’ve gone from 29% to
38%—a relative leap of nearly 40% and an absolute increase of 9%. This
suggests that just maybe with the growth in all those EMR vendors like
Epic, eClinicalworks and many many more, we are now in the steep part
of the “S” curve growth.

And we have yet to come the $40 billion more to be invested that the Prez2Be has promised (OK, I’m mostly joking here….)

There’s just one problem. We’re using the wrong words and talking about the wrong thing.

While the EMR diffusion has been waiting to happen, what’s
going on out there in the wider world is a rash of different
applications being developed for physicians, patients, consumers, and
everyone else in between. What we now need to realize (to paraphrase David Kibbe writing on THCB a few weeks back) is that the EMR is not the be all and end all.

In fact the EMR doesn’t exist. Nor does the PHR, the
integrated compiled mainframe, and who knows what else. Instead, like in the
rest of the tech world there are data and there are applications.

In health care those applications are starting to form
tools that help both to record what is happening, and help to
personalize and analyze information. They next then help support
decisions and enable transactions. And those tools need data both about
general issues (e.g what’s the right treatment for these symptoms) and
specific individuals (e.g. what drugs is this person taking, and what
symptoms do they have?). Those data are increasingly coming from a
multitude of sources, and more and more are being opened up.

So the correct question is not, “are you the physician
using an EMR?” In fact the TLAs of EMR, PHR and the rest should be
scrapped.

The correct question is now, “how are you receiving data
about your patients, and what tools are you and they using together to
improve their care?”

I didn’t know how to ask that in 1999, but since the
growth of Health 2.0, I now do. And hopefully we’ll see the number of
physicians and patients able to answer that hit the steep part of
the “S” curve in the next two to five years.

Categories: Matthew Holt

Tagged as: , , ,

16 replies »

  1. The growing demand by patients, government entities and private sector companies over the last several years is driving demand for Electronic Medical Record (EMR) systems. The complexity of information technology in general, combined with the additional intricacies of a patient’s interaction with the healthcare system, provides for significant challenges in effective integration and adoption of EMR systems.

  2. Technology truly has a big impact on health in the
    developing world and implementing EMR I think will definitely lead to better care..Advantages of electronic medical records include the ability to catch medication errors, check for adverse drug interactions, and track test results and schedule follow ups.

  3. There is no mention of WorldVista in this thread: tailored, adaptable, free.
    You guys spin your wheels while a good solution has been available for about 2 years.

  4. EMR Software and Dragon NaturallySpeaking are being utilized together in more practices and by more physicians every year as they move toward a paperless office. In light of that fact I would like to invite any forum members, lurkers, or passersby to come register and join the KnowBrainer Speech Recognition Forum http://www.knowbrainer.com/PubForum/ to learn about all the nuts and bolts of Medical Voice recognition. If you are interested in Medical speech recognition this is THE forum to either get the most accuracy and productivity out of your current system or do research to decide whether you might be interested in getting started with a new one.
    We have well over several thousand members including quite a few Doctors that ask questions and share information on a daily basis. There are a number of devoted senior members that have been using NaturallySpeaking since the birth of the product so no question ever goes unanswered. There is no way to put a value on free tech support.
    We also receive by far more traffic than any other speech recognition forum and our forum is usually the place where new speech recognition technologies and hardware advancements are announced.
    We have added three additional sub forums to the list on the front page. They are the Vista Speech forum moderated by Rob Chambers [MSFT] and the EMR/Medical Speech Recognition forum moderated by the creator of TexTALK EMR software and owner of Alma Information Systems. The Last is the Computer Issues / Recommendations forum moderated by Chuck Runquist
    (Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS).

  5. EMR Software and Dragon NaturallySpeaking are being utilized together in more practices and by more physicians every year as they move toward a paperless office. In light of that fact I would like to invite any forum members, lurkers, or passersby to come register and join the KnowBrainer Speech Recognition Forum http://www.knowbrainer.com/PubForum/ to learn about all the nuts and bolts of Medical Voice recognition. If you are interested in Medical speech recognition this is THE forum to either get the most accuracy and productivity out of your current system or do research to decide whether you might be interested in getting started with a new one.
    We have well over several thousand members including quite a few Doctors that ask questions and share information on a daily basis. There are a number of devoted senior members that have been using NaturallySpeaking since the birth of the product so no question ever goes unanswered. There is no way to put a value on free tech support.
    We also receive by far more traffic than any other speech recognition forum and our forum is usually the place where new speech recognition technologies and hardware advancements are announced.
    We have added three additional sub forums to the list on the front page. They are the Vista Speech forum moderated by Rob Chambers [MSFT] and the EMR/Medical Speech Recognition forum moderated by the creator of TexTALK EMR software and owner of Alma Information Systems. The Last is the Computer Issues / Recommendations forum moderated by Chuck Runquist
    (Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS).

  6. Suzanne Houck, you wrote, “The good news is that ever so slowly, progress is being made. Payers are beginning to see the failure of pay for volume vs. quality…” This good news has not yet sifted down to us in the Great Plains. Can you tell me what signs you are seeing?

  7. matthew
    I whole-heartedly agree with your post- especially the part about the correct question to ask is what tools are you and patients using to improve their care. this post has generated a lot of good conversation. hopefully leaders in the industry will keep the conversation going and find solitions so that all constituents of healthcare will be better off.

  8. Great, thoughtful piece. Your question “how are you receiving data about your patients, and what tools are you and they using together to improve their care?” appropriately refocuses on the goal of improving care delivery. I would expand the question from the process goal of how are you improving care to the more explicit outcome goals of, “to heal, enable, and delight patients at the lowest possible cost?”
    Healthcare encounters from self care to open heart surgery include four key steps: information gathering, analysis, decision making, and action (including treatment). The right “vehicle” would be a virtual activity hub where data and analysis is easily entered and accessed. Decision making as well as action/treatment would also be enabled for patient customers and providers alike. For example, patients with symptoms could enter data and access analysis regarding appropriate action summarized from best practices. Based on data analysis, recommended actions or treatment might include self-care, a virtual visit, phone care, group visit, or office visit with physician or mid-level provider. If advised, an actual office visit, phone or virtual appointment-not just a request could be made during the same transaction.
    Folks rely on application terms like EMR and PHR to understand and organize functionality of the four steps information gathering, etc. Unfortunately, most existing applications provide unsatisfying user experiences built on outdated, siloed cultural and financial relationships that put providers and payers vs. patients at the center of encounter activities.
    The good news is that ever so slowly, progress is being made. Payers are beginning to see the failure of pay for volume vs. quality, providers are beginning to understand the importance of accountability for measurable quality outcomes, large employers are getting behind vehicles that put patients in the drivers seat for data collection and control.

  9. Great post, Matthew.
    My favorite part is your close:
    So the correct question is not, “are you the physician using an EMR?” In fact the TLAs of EMR, PHR and the rest should be scrapped.
    The correct question is now, “how are you receiving data about your patients, and what tools are you and they using together to improve their care?”
    I couldn’t agree more. EMRs or Clincian Data Tools will help digitize patient data. Let’s hope Data Portability will put data in the hands of patients. Ultimately, it will be up to the patients to promote the use of their data by their docs to drive better care.

  10. As Janet points out the current crop of EMRs attempt to be too general. When you try to write software that does everything for everybody, it’s hard not to make it cumbersome. To date medical software hasn’t attracted any insightful designers. I’m interested in seeing what the Behemoth of Redmond will produce from their medical records project. They aren’t known for their brilliant design, either though.

  11. My husband is a Family Medicine doc and one of his partners took a stab at EMR about five years back. And it was disastrous. They were cumbersome and not at all time saving.
    Since the rubric in Family Medicine (as I see it) is to see a lot of people very fast in order to make money AND to document excellently computers would seem to have a role. However, my husband has an older patient population where complex multiple problems are the rule. (These patients are preferentially off-loaded by younger physicians who understand my first sentence well and understand that those two things are nearly impossible to achieve with high complexity patients.)
    A friend of ours managed a cardiology group and he said their transition to EMR was flawless–“a home run”. But I would suspect that the cardiologist has a much narrower profile of types of visits.
    Until EMRs truly understand the variety of visits in primary care and can shape themselves to that, I think their relevance will be questionable. But at the same time, I think that with thoughtfulness, it can be done. But since primary care docs still have to deal with what is commonly known as the door knob question–that is the doc is poised to leave the room with his hand on the door knob–and the patient says, “Oh by the way…” The “by the way” can be any thing from anterior chest pain (you cannot put them off until another day) to imminent divorce to a skin lesion (which can be big or little) and suddenly the visit morphs just when the doc though he/she was out of the room.
    Not long ago, my husband,his younger partner and I were talking about the dilemma of documentation and his partner very acutely said how he often looks at the documentation of docs who are attempting to streamline their documentation and he said, “their records are so _empty_” And he’s right. As a school nurse, I routinely receive physical forms from a local pediatric office who uses electronic records and there is literally nothing to be gleaned–except “normal exam”. In my older paper format, I used to see that pediatricians would express a speech concern (and then I can ask the school SLP to do an observation) or I would notice a history of prematurity and lung problems or a history of a host of other ailments that can impact a school OR tip me off to risks in learning.
    I am willing to understand EMR as in their infancy but right now, they are not serving the health consumer or the physicians themselves who use records to look back and see what has already been tried, discussed or work-ed up (or why it has not been worked up).

  12. EMR’s were supposed to make it simpler to practice medicine by freeing caregivers from the burdens of charting, paper orders, missing information, missed connections and lack of relevant clinical evidence. Had this been the case, there would have been a significant productivity gain, which would, in turn, have translated into more income (either realized in cash or shorter work days for the same amount of clinical effort). See Kibbe’s earlier post (Confessions of a Physician EMR Champion) for a far more detailed and thoughtful analysis of this problem.
    Thought the toolset has advanced dramatically in the past decade, EMRs are still far too difficult to use, and the communications links I talk about (MD to MD and MD to patient) don’t work very well, if at all. The lack of interoperability is not the main problem; it is their failure to foster teamwork and seamless communication. And they are so expensive to acquire and operate that there has been no apparent institutional gain from their use for most large systems for at least a decade.
    We are only now seeing in places like the Geisinger Clinic and the VA, who have been at it for more than a decade, the changes both in clinical practice and cost that justify the investment. (And the VA is just at the beginning of a costly and complex system modernization written ably about in this blog).
    The Obama Administration is going to have a rude awakening when they begin looking for the actual payoff from past investments in this technology. I think there is far more immediate social gain from simplifying and completing the automation of payment processes than from the EMR itself.
    It’s tempting to blame the vendors, but as I argued in my book “Digital Medicine: Implications for Healthcare Leaders”, it is the fragmentation of the clinical workplace, bureaucratic culture and lack of leadership on the provider side that is the real culprit. The toolset is not inherently transformative; only with, particularly, clinical leadership comes the gains that justify the investment.

  13. In all my readings on, and about, EMR I have never clearly appreciated who gains – is it the patient, the physician, or some payer? The inability of EMR to gain traction, I believe, can be sourced to this fundamental problem. Where is value generated and whom is it appropriated by? There have been no clear answers to this question. The “S” curve reasoning may be applicable if we understand that to mean the 29% or so of physicians who have embraced EMR are the early adopters and the individual gains – one presumes, monetary, since there has been a concomitant investment by the parties concerned – are likely to see a ramp up by other new users. For that, we need compelling ROI evidence.
    In this context, I have to mildly disagree with Jeff Goldsmith who argues in favor of interoperability as the hurdle that needs to be crossed. That is a very tech perspective. While interoperability is important, this has to be grounded on a sound economic rationale for the different constituencies affected by it. EMR today may just not be solving a business case at the individual level.

  14. Good post, Matthew. And Jeff’s term “clinical groupware” is terrific. That’s the transition you’re describing in a nutshell, from stand-alone client-server EMR software application from a single vendor, to “clinical groupware” (web-based) capable of assembling relevant patient data, sometimes from multiple sources, and communicating between and among care providers, patients, and appropriate others. The outputs are coordination and continuity.
    What we’re seeing in the “EMR space” was predicted by Clay Christensen according to his “law of the conservation of modularity.” Namely, that EMR components or modules would emerge as the products improved, creating options for medical practices to purchase these parts at much lower prices and to mix and match, plug and play. E-prescribing and web portals are prime examples of this trend. And, overall, it’s good for the physicians and good for their patients. Less risk, more and better defined benefits, and generally increasing use of health IT for more affordable care.
    Regards, DCK

  15. Matthew, this is a superb post! Good on ya, mate! As a long time EMR enthusiast, I’m convinced you’re asking exactly the right question, to be followed by “are you using digital means of communicating with your patients, specialty colleagues and the hospital at which you practice (if any)?”
    As Bob Wachter and others have communicated in this space, the present EMR’s principal functions have been: documentation (handled at essentially a Windows 95 level of functionality) and ordering/results reporting (ditto). Wachter’s plaintive post of a few months ago (Why Cannot EMR’s Be More Like Facebook?) goes to the heart of the problem: they OUGHT to be GROUPWARE, that enables the clinical team to communicate better around a common core of knowledge about the patient and the relevant evidence about their condition.
    This is why the continuing largely fruitless struggle to achieve interoperability is so important. Until that happens, the EMR will not fulfill its most important function as a communications platform for knitting the clinical team, patient and hospitals together. Presently, the only way to achieve this interoperability is inside a single vendor’s “walled garden”.
    This focus on communications/co-ordination is also the difference between VISICU’s eICU product and most commercially available EMR’s- the eICU was intended first and foremost to be a communications/care management tool.
    The evolution toward clinical groupware isn’t happening anywhere fast enough, and until it does, a lot of the shining promise of the EMR will remain just that- potential unfulfilled. And we’ll see minimal actual improvement in clinical practice or outcomes for our thusfar massive societal investment.