How Long Will It Take to Adopt EMRs?

One story that got a lot of press last week was Prof. Aashish Jha's research update on the adoption of Electronic Medical Records. Prof. Jha has observed that only about 12% of US hospitals have a fully operational EMR at the end of 2009 and only 2% would have complied with Meaningful Use guidelines release this July that determine whether Providers can receive Federal incentives starting next year. 


Common sense suggests that electronic medical records should have much wider adoption, particularly since most sides agree that a properly implemented EMR improves patient care. After all, as one advertisement by United Healthcare recently noted, "your Pizza delivery guy stores your orders electronically. Why can't your doctor do the same?" The reality is that EMRs have not been widely adopted because there are many counter forces:

• Medical information like health status is not as determininistic as, say, financial data. There is no unique score, or set of scores that can uniquely pin point the state of your health in a way that two reasonable, independent observers would always agree. In banking, there is no ambiguity on your bank balance when it is viewed by two separate observers. Similarly, your Pizza preferences are, in general, mostly deterministic and can be described in a way that is interpreted in the same way. I buy this argument, but it doesn't explain it all.
• Some commentators have suggested that medicine is viewed as an art by an older generation of physicians and as long as that group makes decisions for the medical community, EMRs and other metrics-driven tools are likely to remain on the backburner. This is an artful argument, but again, it cannot explain the widespread inability of Providers to adopt EMRs. Some (more cynical?) observers have even claimed that the Providers don't want EMRs because it allows them to preserve inefficiencies and make more money.
• The most compelling reason for the lack of adoption is simply that the benefits of EMR accure to those that don't pay for it. Put in other words, those that have to pay for EMRs aren't the ones that benefit the most. So, cash strapped hospitals are not very likely to want to make a huge IT investment if they don't see direct returns of some form. The Federal goverment has recognized this misalignment of incentives and much of the HITECH funding through Meaningful Use attempts to address this deficiency.

So, if you believe this logic, it would stand to reason that EMRs will see a much more rapid pace of adoption as Federal incentives kick in. Prof. Jha's paper seems to make a similar conclusion.

For more on the topic, Prof. Eric Ford and others published a bass diffusion analysis applied to EMRs. This study was published in 2006 and concluded that EMR adoption would take a decade longer than the 2014 target set by the Bush Administration as part of E.O. 13335 that established the ONCHIT. I have yet to see an update on this study, post Meaningful Use incentive payments, but the paper is still worth a read for its analysis of buying behavior in the medical community. Specifically, the paper examines the similarity in buysing behavior between healthcare IT and consumer goods, given that both are influenced by the coefficient of imitation (doing what key opinion leaders do) than coefficient of innovation (buying the latest and greatest). Heady stuff.

Gold Rush Redux?

The Boston Globe's story this weekend on Athenahealth CEO Jonathan Bush confirms what is becoming increasingly apparent: the EMR revolution is indeed a gold rush and federal incentives will define the agenda in Healthcare IT for the foreseeable future. A survey of news items in HIStalk also lists several stories supporting this conclusion.

If you compare the current environment to the pre-bubble days of the internet, you could assume we are in 1994. Netscape has not been launched yet, but there's a wide expectation of a revolution to come. Web 2.0 proponents will claim that the first wave of the internet did not produce productivity improvements comparable to those from Web 2.0 technologies. Similarly, skeptics might believe this gold rush will not truly transform healthcare IT, but it will set the stage for an eventual revolution in technology, patient engagement, and evidence-based medicine.

Whichever way you look at it, participants in the market will adapt their strategies in order to profit from this environment. News satire site, The Onion had a memorable headline a few years ago:  "Recession Plagued Nation Demands New Bubble to Invest In." I think that call is being answered.

The Best Care Anywhere?

I'm reading the second edition of Philip Longman's classic: "Best Care Anywhere". The book is a fascinating read into the story behind the VA's transformation from a case study in Government mismanagement to a model of modern medicine with manageable costs and outstanding outcomes.

The book offers some surprises, and I need to research them more.

• One story concerns Beth Israel New York and Duke Medical giving up promising studies into better patient outcomes because the programs were not profitable. This highlights the disconnect between incentives and goals that everyone talks about. I imagine other hospitals aren't any different and these two should be lauded for at least trying outcomes-based programs.
• The other bigger story for me was the apparent success of the VistA EMR system which is Open Source. There are emerging companies like WorldVistA, vxVista, and others following a "Red Hat strategy" of wrapping maintenance, enhancements, and support around VistA. Could this be an alternate world to the commercial EMR vendors?

See also the RAND Corporation's research brief on quality of care at the VA from 2005. Apparently, the VA's success isn't new.

Analysis of the Final Rule on Meaningful Use

Inga from HIStalk provides a helpful summary of the changes between the preliminary and final rules for  Meaningful Use. Also John D Halamka has posted a very nice summary slide deck for those who need to present a synposis of the changes. Finally, Keith Boone's colleague, Dan Sepulvada helped compile a table summarizing the differences between the preliminary and final rules.

Two quick observations are below; look for a separate post in the coming days on how these new rules affect document management, particularly Release of Information.

• Across the board, the final rules are a lot less ambitious than the preliminary ones. This is mostly a good outcome because we still don't know the behavior these rules will motivate. More* on this at the end of the post. 
• The preliminary rules were written with an assumption that medical records will be mostly electronic by 2015. Several studies suggest this goal is not realistic, so it will be interesting to hear whether these rules will indeed prevail or whether there would be further changes down the road. A Bass Diffusion model applied to physician EMR systems by Prof. Eric W. Ford and others claims true adoption will take a decade longer than anticipated. And that's the optimistic scenario!

I look forward to hearing your thoughts in the comments section below.

*Why less ambitious rules may be a good thing: Rob Lamberts at The Health Care Blog draws an analogy between Meaningful Use rules and those in No Child Left Behind--another well intentioned legislation that introduced undesirable behavior on the part of the actors (the teachers). One would rather have rules that are realistic than those that are so strict everyone has to find workarounds.

Kodak-Medco Data Partnership

This week, Kodak announced a partnership with Medco Data to provide conversion of paper-based charts to digital files. There are a number of local vendors currently serving this space, so the entry of a major brand like Kodak appears to validate the belief that EMR adoption improves when there is continuity with historical records.

The press release states:

"Conversion of paper-based patient charts represents a critical consideration for practices implementing EMR/EHR systems. Scanning charts prior to the installation of an electronic records system can significantly improve the chance of a successful adoption."

The product is targeted towards physician practices and ambulatory care. It emphasizes the ability for existing staff to perform scanning work.

It will be interesting to see what physician practices and ambulatory care EMR systems, particularly AllScripts, do to better integrate paper and electronic medical records.  Stay tuned.

EMR or EHR?

EMR (Electronic Medical Records) and EHRs (Electronic Health Records) refer to two very distinct concepts. Unfortunately, government officials, vendors, respected journalists, and consultants have used these terms interchangeably in order to refer to EMRs.

HIMSS Analytics provides clarity on the meaning of these terms:

• Electronic Medical Record: An application environment composed of the clinical data repository, clinical decision support, controlled medical vocabulary, order entry, computerized provider order entry, pharmacy, and clinical documentation applications. This environment supports the patient s electronic medical record across inpatient and outpatient environments, and is used by healthcare practitioners to document, monitor, and manage health care delivery within a care delivery organization (CDO). The data in the EMR is the legal record of what happened to the patient during their encounter at the CDO and is owned by the CDO.

• Electronic Health Record: A subset of each care delivery organization’s EMR, presently assumed to be summaries like ASTM s Continuity of Care Record (CCR) or HL7 s Continuity of Care Document (CCD), is owned by the patient and has patient input and access that spans episodes of care across multiple CDOs within a community, region, or state (or in some countries, the entire country). The EHR in the US will ride on the proposed National Health Information Network (NHIN). The EHR can be established only if the electronic medical records of the various CDOs have evolved to a level that can create and support a robust exchange of information between stakeholders within a community or region. While some forms of early EHRs exist today in limited environments, it will be difficult to establish effective EHRs across the majority of the US market until we have established clinical information transaction standards that can be easily adopted by the different EMR application architectures now available.

John at "EMR and HIPAA" argues that this distinction is irrelevant and that these terms will always be used interchangeably. He writes:

"My favorite in the EMR world is when people go crazy if you use the term EMR and not EHR. Let’s just get over it. The doctors I talk to really are. They use them interchangeably to mean everything that you might technically consider an EHR."

What do you think?