Ensuring Investment in Healthcare Information Technology Does Not Flatline
Given the $47 billion awarded in stimulus funding, it's clear the government's assumption is that healthcare information technology (IT) will deliver better care at lower cost. The IT industry and all the healthcare IT mavens are waving the flags and beating the drums.
But can current IT deliver?
I wrote a white paper for Microsoft in 2005 that suggested healthcare IT was not delivering on its promise then Microsoft Paper. And in January, The National Academy of Sciences' National Research Council (NRC) published a report that states it's still not delivering today NRC report..
The NRC report is a sobering breath of fresh air in a discussion that has grown stale and over-heated with the self-serving marketing efforts of the IT industry and chest-thumping, one dimensional reports of success. The report tells a far different story: current healthcare IT and all the billions of dollars we have invested in it "fall far short ... of what is needed to support the ... vision of quality health care." The caution is that the huge investment in healthcare IT is in danger of flatlining, if it's not already dead on arrival.
The problem is not a lack of money, interest or hard working, intelligent, compassionate people. The root cause of our dilemma is the deeply embedded conviction that we have to solve big complex, expensive problems with big, complex, expensive solutions.
What's wrong our with our current IT solution? The NRC report spells it out. It describes monolithic, expensive systems that are difficult to change - in fact, implementing and improving them can take years or even decades. Further complicating the issue is the fact that these systems' designs are tied to automation of current best practices - which often aren't "best" or even "good" and certainly aren't the future. They lack support or even understanding of the cognitive functions and needs of clinicians and staff. Finally, the report describes how poor designs can increase the chance for error, add to rather than reduce work, and compound the frustrations of executing required tasks without an effective way to rapidly problem solve and improve them.
In my work I have seen far too many examples of such problems. One health system "successfully" automated physician histories and physicals in such a way that for months 50 percent of the histories and physicals failed to arrive in time for the patients' surgery. This multimillion-dollar system was perfectly designed to not work.
On the other hand, I've learned that it is possible to improve healthcare IT to effectively deliver better care at lower costs. Recently brainstorming with a visionary technology entrepreneur, we came up with an idea for a healthcare IT system that a patient would wear on their wrist like a watch. The system would create continuity between every healthcare provider the patient encountered. Similarly, my soon-to-be-published book, Designed to Adapt: Leading Healthcare in Challenging Times (Second River Healthcare Press, 2009), describes a simple "Ideal Health Card" that would permanently solve the hassles and frustrations of registration and medication reconciliation.
I believe the healthcare IT problem relates to my frustrating experience as a healthcare executive searching for more data, better metrics and going to endless meetings to solve the big problems - quality, safety, profitability, compliance, paperless-records, etc. - while physicians and staff struggled with the multitude of small problems that eventually created all those larger issues.
As a Visiting Scholar at Harvard Business School, I studied those few companies that successfully managed complex, dynamic, unpredictable work and found three common principles that apply directly to our current IT dilemma. The successful few were extremely good at:
1. Developing their people to solve small problems as close to the work as possible
2. Using that problem solving to create a responsive, learning organization
3. Accelerating learning with simple, flexible, locally responsive IT
In other words, technology was not the solution; technology was a flexible, improvable tool that was used to accelerate the solution.
My twelve years of experience in testing, validating and improving real time problem solving capabilities in healthcare makes it very clear what delivering on the IT promise will look like. Systems that will deliver will be less costly, modular, fast, flexible, friendly, and responsive. Think smart phones, distributed networks, intraoperability and locally improvable.
Systems that won't deliver are the opposite. Think desk tops, laptops, expensive, centralized, massive roll-outs that can't be changed or improved until the next version is designed, built, purchased and implemented (again).
Unfortunately, the latter is what the NRC report says we are getting. The big IT train is already out of the station. We won't derail it, but we can accelerate work on the alternative that will deliver on the promise. We won't buy this solution. We'll make it, through disruptive innovation.
Fortunately, all the pieces to build fast, flexible, friendly, effective IT are out there waiting to be assembled, tested and improved. Mandl and Kohane's editorial, "No Small Change for the Health Information Economy" in New England Journal of Medicine ( NEJM editorial) is a good summary. I know small companies with great, patient-focused IT solutions in important areas like diabetes that are in development now. Government can help by creating safe harbors for development and directing funding into non-traditional, disruptive opportunities.
For disruption to succeed, it is essential for a few, visionary IT companies and healthcare providers to link together to create the new systems that will make current systems obsolete. We must start with the patient. Development must be embedded at the point-of-care to allow real time learning. The focus must be getting patients exactly what they need at continually lower cost. That's the way to fix healthcare.
We can't afford for our IT investments to flatline.
Dr. John Kenagy is a former Visiting Scholar at Harvard Business School and the author of the forthcoming book Designed to Adapt: Leading Healthcare in Challenging Times (Second River Healthcare Press, 2009).
Loading comments…
Want to reply to a comment? Hint: Click "Reply" at the bottom of the comment; after being approved your comment will appear directly underneath the comment you replied to
If any one wants to learn more about workflow management technology in ambulatory care electronic health records, an excellent source is:
charleswebster.com
Charles Webster, MD, MSIE, MSIS is probably the nation's foremost expert in this area.
As you noted, technology needs to be driven from the bottom up rather than the top-down, with the focus on technology being an enabler rather than an end-solution. I agree with your sentiment, yet there is a catch 22 in this logic, from a technology perspective. For most of this type of technology to work, there needs to be digitized records (appointment schedules, lab results, images, etc) from which to build off of. We are in need of a basic electronic health foundation that will allow us to build the time and cost saving services we all crave.
You mentioned that the government should support small innovative companies. The best thing the government can do, IMHO, is to push its health services (Medicare, Medicaid, the VA, etc) to promote an open standard electronic health backbone, as hinted at by the NRC report you highlighted. The US government is the largest employer and spender of healthcare dollars in the world; use that power, and other financial and time incentives, to push for the sort of low-level backbone its own advisers are saying is needed.
Thanks, for your comment, Adam. Now we are getting somewhere. Instead of designing and implementing "the solution," as I read your note, you are recommending government can help create the foundation upon which a multitude of new solutions can continually evolve.
That's very interesting. What do other readers think? Instead of designing and implementing "the fix," what are the keys to creating the platform on which simple, fast, friendly, effective healthcare IT can continually evolve?
My work suggests that the evolution should be intimately tied to rapid incremental change focused on the patient at the point of care. How can that happen? In the words of the NRC report, how can we make incremental change in IT rapid and radical? To those characteristics I would add safe, low cost and customized to be focused on the patient. Who else has views on this subject?
I still feel my design below addresses most of your issues. The problem with applications is not updating the application, Microsoft replaces them all the time with updates, but rather the underlying database. One way to solve that is to store the information in the Xform itself. And save the Xform with it's XML Schema. That way as the system learns or evolves new XForms and schemas can be created, but the older ones can still be processed. I still think the tele-coms are better able to hold the forms due to the way the budgetary process works. But I do feel the Government needs to be in charge of the "Best Medical Practices" workbooks. The user interface to the doctor would then be a web page or pdf document.
Dr. Kenagy---I am part of a team at a Midwestern hospital that just started working with a couple of your "disciples" so I was floored to see your post today. We just finished our first week of observations. Very interesting to have what I'm learning about put into larger context by the man himself. On huffingtonpost, no less! Keep em coming!
Thanks! You are working with some great people. I guarantee you will learn a lot. And, from what i have heard, your team is doing well. The good news is, if you keep problem solving, you and your organization will never stop learning.
The challenge will be for you to help transition the managerial role in your hospital from gathering data, going to meetings and implementing projects to developing people with real time problem solving in the work place. That's the real opportunity.
I am on Huff Post regularly, so they'll keep coming. Search john kenagy for previous posts. They all have a relationship to Adaptive Design. I'll be interested in your future comments and insights! Tell some stories about your and your colleagues work when you see a chance.
thanks for a well written piece. At a major Boston cancer treatment/research organization, I saw them go from Meditech to PeopleSoft, at a multi-million dollar pricetag. Meditech was better. The implementation was rushed and (working in HR) I saw immediate problems, as eligible staff literally could not be entered to PeopleSoft to enroll in the pension plan!
Now here's some IT heresy: I think a good paper system beats an inefficient computer software system ANY DAY. "Upgrade" are not necessariy better than the systems they replace, either.
No question that IT is important to health care, but nowhere as important as providing quality care, universally.
Thanks, Mark. Great example of the problem. I have seen this story repeated over and over again. But it is a story no one wants to hear.
And your comment on paper is insightful. Our problem is not paper records. the problem is managing knowledge and information effectively. Toyota coordinates the activity of 4000 people putting together 12,000-14,000 parts to make a new car every 30 seconds with paper as the dominant means of communication.
On the other side of the coin, a large health system I have worked with instituted a "paperless medical record" and discovered the use of paper in the system increased! Why - all the workarounds to the new system used lots more paper.
And your final comment is right on target, let's get people the care they need and use technology as a tool to get it to happen, not a solution.
I like the Toyota example, but Toyota uses structured XML documents to produce the paperwork. The assembly line also has a QA department that does random sampling and physical tool settings used together they produce high-quality cars.
To replicate that at the doctor's office I would use DITA (used for IBM's technical reference manuals) with Xforms and XML Schema for validation. The schema would represent the QA department doing 100% sampling.
Part 3
Step 3 - Wrap the hospitals and Health care insurers in a Services-Oriented Architecture (SOA). Think of going on an internet shopping spree and all the forms you've filled out. Now envision where there is no user interface and your shopping trip could be completed and validated in the blink of an eye.
Pros - Because the medical community has its own very precise terminology, what’s covered, partially, wholly or not at all by an insurer can be checked using a webservice in real-time. This would allow the patient and doctor to choose the right treatment according to effectivety and cost. The use of webservices at the hospitals is to allow the use of workbooks without having to replace any of their current systems because it can be placed on top or as a step in a workbook/workflow.
Cons - A lot of layoffs at both the Health care insurers and the doctor's offices.
Since the technologies I proposed using are at least five-years old and use XML and XSD (XML Schema for validation), the health care savings and costs can be measured accurately. Director Orsag's 700 billion dollars by using "Best Medical Practices", Senator Sanders 400 billion dollars in administration fees, malpractice insurance rates, medical errors and prescription drug costs are all addressed, if this system were adopted. But what a whirlwind of change it would be.
The next multi-agency SOA architecture that is successfully and cost-effectively deployed will be the first one successfully and cost-effectively deployed. For that matter, how many true CORBA (or similar OO solution) architectures have been cost-effectively implemented to serve multiple organizations?
After all, SOA is basically just a re-marketing of the same old ideas that sounded absolutely awesome in doctoral dissertations and self-adulating IT academics. But when it comes time to implement outside a lab, these architectures become immensely complicated to manage, spawning cottage industries (like SOA's wonderful three-headed spawn, UDDI - how much does a commercial UDDI v3 server license cost, hmmm?), and forcing programmer retraining (ah hell, let's just off-shore it instead!) to deal with a maze of concepts that are really just the same as older concepts, but have to be renamed so that there is an appearance of some sort of improvement from the older concepts.
Please, let's step away from the altar of fad technology and just put together something simple that works.
www.ussailing.org uses webservices to access both an old and new database until the last of the asp pages can be replaced with the new asp.net pages and the old database can be retired. Apache.org has it as freewear.
Part 2
Step 2 - Create electronic medical workbooks maintained by the Government and Health care Industry that would represent "Best Medical Practices" for all the different treatments using IBM's open-source Darwin Informational Typing Architecture (DITA) for views and workbooks.
Pros - Because the workbooks are XML-based each treatment can be checked for accuracy and completeness in real-time. Medical errors would head lower because the workbooks can be designed to check that all the necessary steps were taken for the treatment. In addition, because "Best Medical Practices" were followed and the workbook stored at a third-party punitive damage awards should also head to zero. Another plus is that distribution of new or updated workbooks is done at the telecom-level, it could be achieved in a matter of seconds.
Cons - None that I see, except maybe lawsuits in the battle for the best effectivety rating amoung pharmas.
Part 1
I think the issue is not a large or small system but rather to setup the engine of Health care IT with checks and balances.
Envision an electronic loose-leaf notebook (not computer) that has pages, workbooks, outcomes that represents your medical history. Where the pages and workbooks can be checked for accuracy and completeness. And then we need to complete the following three steps:
Step 1 - Store all doctor's office patient records following HIPAA regulations at the local-telecoms. No name and address. Hospitals could use cloud computing with a locally located third-party controlled server (SaaS.)
Pros - The telecoms are big enough, have the redundancy, the geographical reach, the competition, the bandwidth, on-call staff and their business model is based on providing reliable services at a low cost. They can provide better security for the patient's records than at a doctor's office. And because the records are more centralized it would allow real-time studies of the effectivity of treatments and "Best Medical Practices".
Cons - People may not want their records open to research. This would be a voluntary effort, otherwise the record would be encoded using the doctor's electronic certificate of use, but I would hope that a statistically significant portion of people would allow their records to be used for research. Another con would be the layoffs at the doctor's office as the telecoms offer more and more services like billing, reminder calls, the list is endless.
If you are really interested in healthcare IT that works, you will check out EncounterPRO which has won the Davies Award multiple times. It was developed by family with no capital and no marketing budget.
This product is not an EHR in the traditiona sense. It is a workflow management tool that allows a lay user to build an EHR to the specifications of each office and each doctor in the office. Setting up the system to match the desired workflows of the office takes a day or two. Changing the workflow for a particular office process takes minutes and with training, even clinical office personnel can effect these changes. The workflow engine is so sophisticated, it can be set up to redefine itself in real time based on clinical data that becomes available during the course of the visit. And the system can be set up to automatically report data to reporting agencies such as vaccine registries.
The adoption success rate for this software is over 95% and the time that it takes for the office to return to prior levels of productivity is weeks, not 6 - 18 months that is typical in the industry. Currently, the EncounterPRO system is serving about 5 million patients in ambulatory care and its users have about 50 million years of digitized health records. And, I'll bet you have never heard of this system.
Thanks for letting me know about EncounterPro. It would seem a good example of flexible, friendly and powerful. And restoring productivity quickly is a key attribute I did not mention in my article.
what barriers does the software encounter? Is it too different from more traditional IT systems? Does it not meet IT buyers' expectations? what are your thoughts? John
The Company is under-capitalized in part because investors have no concept of how revolutionary the technology is. Doctors who are not users also do not understand that a true workflow management system is VERY different from all other ambulatory care EHRs that are advertised as "workflow" systems. Other vendors say they have a "workflow" system because the user can navigate to various screens. It is essentially a meaningless statement because all products with more than two screens have "workflow." A workflow management system allows the user to determine in advance which screens appear for what people at what time for what tasks. Every encounter is defined as a succession of screens with user-input options and each task or treatment within each encounter is further defined as a succession of screens and operations. So, for example, a CBC screen may automatically appear as part of the encounter workflow for the nurse when a female patient over a certain age is being seen for her annual checkup. The behavior of the CBC treatment is also defined by the workflow engine and can involve such steps as performing the CBC in-house, sending it to an outside lab, and notifying the physician when the CBC results are finalized.
The company was run by a middle-class family with a genius son named Mark Copenhaver and no connections to capital. The company got behind on its payments to the IRS and an over-zealous IRS agent who did not understand the software business drove the company into bankruptcy three years ago when the company finally had its first profitable quarter. (A $100,000 check hit the company's account for a hardware sale and an IRS agent thought she would rather have the check than have a payout of company profits and she levied the bank account.) The IRS agent then changed her mind and allowed the company to "unfile" the bankruptcy. The damage was done, however and the Company immediately lost 75% of its sales.
The product has more than met buyer expectations although the company has never been sufficiently capitalized to provide first-rate support.
A good contact at the company is Dr. Charles Webster, who is an expert in workflow technology. charles.webster@encounterpro.com.
Dr. Webster has, for all intents and purposes, donated years of his life to the product because there is nothing else like it in the world in ambulatory care.
I worked on a product for the Danish giant Novo-Nordisk - Diabetic Electronic Medical Record. This was 13 years back. We were deeply involved with the company and with their associating hospitals to understand the needs of that particular department alone. Spending the stimulus money on enabling IT alone will have a good impact on the economy. Today, we live in a far better connected world.
However, the medical fraternity are quite lethargic when it comes to adapting towards technology. IT can definitely enable the health care system to provide a better service.
Thanks for for the comment. the example you offer of obtaining a deep understanding of the needs of the work place is very important. what the NRC report documented and what my experience has shown me, is that current healthcare IT doesn't work that way.
IT can make a big difference in healthcare, I agree. But, in my experience, lethargy is not the problem. Healthcare avidly accepts new technology that works. The lethargic response is a symptom of the underlying problem that current IT does not fulfill it's promise. Fulfilling the promise of better care at lower cost will not engender lethargy - it will be both attractive and exciting. And that's a great business opportunity.
A great post. This article falls in line with a book I recently read: http://www.bustingloosefromthebusinessgame.com
Thanks! and I appreciate the lead on the book.
You must be logged in to comment. Log in or connect with
