EHRs for a Small Planet

EHRs for a Small Planet


Right now, American health care information technology is undergoing two enormous leaps. First, it is moving onto Web-based and mobile platforms – which are less expensive and facilitate information exchange – and away from client-server enterprise-centric technologies, which are more expensive and have limited interoperability. In addition, more EHR development activity is headed into the cloud, driven by large consumer-based firms with the technological depth to take it there. Both these trends will facilitate greater openness, lower user cost, improved ease of use, and faster adoption of EHRs.

But they could also impact the shape of EHR technologies in another profoundly important way. What is often lost in our discussions about electronic health record technology in the US is the relationship these tools have to our health and health care problems…globally. We could be designing our health IT in ways that are good for the health of people both here and around the world, not simply to enhance care in the US.

Designing health data and management tools only for the particular operational needs of the current US health system may be doubly wrongheaded: It risks locking us into outdated technology and an expensive, dead-end path, while, at the same time, it could restrict collaborative exchanges of ideas and innovations that could improve health care here and abroad through better designed information technology.

Perhaps we should design EHRs for a small planet.

Rene Dubos (1901-1982) was a microbiologist who produced the first commercially marketed antibiotic. He also wrote widely about the relationship of humans with their environment, notably in So Human an Animal (1968), which won a Pullitzer Prize. In 1972, with economist Barbara Ward, he co-authored Only One Earth: The Care and Maintenance of a Small Planet, which set the issues and tone for the first major international conference on the environment. Dubos also first used the term “think globally, act locally,” advice to consider the widest possible consequences of our behaviors, but to take action in our own communities.

What would our EHR technology design efforts in the US look like if we incorporated Dubos’ more expansive framework? What principles might shift our thinking about EHRs away from America’s failing health system paradigm — with its illusion of unlimited resources, delivered by a fixed and ritualized set of professionals and institutions, and costs that double with each passing decade — towards a vision in which EHRs promote sustainable efforts in disease prevention, health improvement, social responsibility, and environmental protection? How might we think about EHRs globally while acting locally?

Principle 1: Define success with local health and health care problems in mind.

Defining EHR success is important, partly because US federal policy for EHR adoption is currently so dynamic. It would be easy to simply define success in terms of physicians’ short term acquisition of today’s EHRs, and the economic boost that might result from new government IT spending (e.g., IT jobs and EHR vendor profits). But Dubos might argue that successful EHR adoption should require measurable social and ecological benefits in the communities where the technologies are deployed, after consideration of the ‘big picture’ in which health spending is one among many societal priorities competing for limited societal resources, and therefore ought to be conservative.

The US’ current EHR adoption strategy channels money directly to doctors and hospitals, among the most privileged professional groups in any community. It could, instead, send those funds directly into the communities served, focusing on the local circumstances that result in fragmented, disorganized, and inconsistent health care delivery within driving distances of its citizens. EHR technologies could address communities’ continuity and access-to-care problems, and relate these to major preventative and chronic illness management challenges, e.g. vaccinations, obesity, and risks of heart disease. More and more people in adjoining communities could be reached by building on successes. Lowering health costs nationally is an important goal, to be sure. Maybe the best way to get there is to stimulate uses of health IT to improve individual and community health through local action. (It goes without saying that the system’s financial incentives would also have to be re-aligned.)

Thinking globally and acting locally would require us to study and plan how EHRs might benefit different communities, as unique populations with particular health risks, public health problems, and care delivery challenges. We would have to study those risks and challenges in each community, or in groups of neighboring communities. This is not easy, and it can be time consuming.

But the alternative, which seems to be to spend huge amounts of state, federal, and local dollars on one-size-fits-all health IT projects, top-down EHR systems that work for the VA or DOD but probably nowhere else, or data exchange efforts that may not be capable of solving, or even suitable to, the problems most at hand in that locale, could be simply disastrously wasteful by comparison. What works in central Indiana, quite honestly, may not be the right thing for Green Bay, Wisconsin, Helena, Montana, or Pamlico County, North Carolina.

Principle 2: Make the best possible use of existing IT resources before building or installing expensive new EHR systems.

Rather than ask “What could we do if everyone had computer systems like the most advanced large groups, e.g. Kaiser or the VA, let’s ask “What could we accomplish if we utilize the computers everyone already has?”

Experience has shown that it is not wise to expect big and complicated things to somehow become small and simple. For one thing, costs don’t necessarily scale. In contrast, though, the evidence is now overwhelming that with browser-based software running on personal computers and cell phones, and small applications running on hand-held devices, like the iPhone, consumer use can grow at extremely rapid rates and lead to complex social networks, rapid communications and feedback loops, and massive search and data analysis capability.

Examples abound of the kinds of resources available through inexpensive personal computers connected to the Internet, cell phones, and the newer smart phone technologies. Skype, the Internet-based voice communications company, has over 500 million registered users world-wide, which would make it the largest telecom carrier, if it were one. The top 25 wireless providers globally already service over 3 billion registered customers. The iPhone, introduced in 2008, has more than 57 million users, the fastest user growth in consumer technology in history, many times faster than the earlier rapid growth in PCs or the Apple iPod. Facebook – the social network platform where people send email, chat, share photos, and share interests – now has 350 million users and is growing at 660,000 per day! Lest we forget, these ubiquitous technologies are not just used for fun and games: massive amounts of data are being exchanged as well. And they are getting cheaper to own and operate all the time.

And yet they are for the most part useful only at the margins of health care, an industry that has somehow walled itself off from IT modernity. We certainly have not yet capitalized on the health and medical uses of the extraordinary networked computing resources available now in almost every home and work site in this country. EHRs for a small planet need not cost $54,000 per physician, which is the current estimate used by ONC and HHS.

It would be a critical mistake to waste our resources, time, and effort building new specialized state or regional data centers requiring complex and proprietary identity management technology for access, and to train a generation of IT professionals how to manage these expensive centers and the technology deployed there, when better design and efficiency could be obtained by use of the existing “off the shelf” general- and multi-purpose data highways, application platforms, and end-user computing capacity now available for health data exchange.

Principle 3: Design EHRs for the smallest unit of care delivery, with a focus on connectivity and communications.

Connectable EHRs can be designed for small medical practices and clinics in primary care, where the great majority of care is delivered, and for patients’ themselves — in their homes and places of work. Designed from the local, grassroots perspective, EHR technologies would also focus on affordability, ease-of-use, and especially on connectivity and continuity of information across those units in a given community, using existing computers, cell phones, the Internet, and yes, even fax machines.

Our current approach to health care IT, in contrast, is biased towards the needs of a handfull of professionals working in a relatively small number of large enterprises, such as hospital systems, and in large multi-specialty practices. These large units typically represent the most complex “use cases” for EHRs, based on the needs of the most complicated and sickest patients, requiring the most intensive usage of drugs and pharmaceuticals, and at the far end of the spectrum in terms of complicated ancillary medical devices, such as MRIs, medicated stents and proton accelerators.

These large health care units are often fiercely competitive and have little use for data exchange with competitors, and even less interest in using computing resources to reach across the communities they serve. As a result, they may be among the least appropriate and least competent stewards of community-based health IT resources. And yet their representatives dominate the steering committees and governance boards for the nation’s health information exchanges (HIEs) and regional health information organizations (RHIOs), where a big chunk of the federal funding is now going.

If waste is the failure of design, then designing EHRs for a small planet would avoid lengthy and disruptive installations and long training cycles involving expert consultants. Instead, they would favor modular, browser-based EHR software that are familiar to physicians, their staffs, and their patients, and that can be navigated simply.

Implicit in this design priniciple is a requirement for minimal training that focuses on how to use the software to best improve care, rather than on which buttons to push in which sequence to optimize fee-for-service reimbursement. EHR software that looks more like Facebook and less like a database manager’s tool kit, that can work through web browsers and mobile devices, and that can be incrementally expanded as new uses arise, is not only likely to be more adoptable than today’s EHRs, but also less expensive to own and operate.

Principle 4: Recognize that what sustains most information technologies is people’s desire to connect with one another.

Email is the “killer app” of the Internet. Facebook and Twitter have become the amazingly fast growing online social networks. Human beings seek connection at nearly every opportunity. Technologies that facilitate that connectedness and then provide key utilities are most likely to succeed.

Maintaining and restoring health, preventing disease, and the act of caring for others who are in need due to problems of the body and mind: these are among the most basic social activities of human beings, our communities, and our cultures. And yet, for complex reasons associated with money and power, our health system and the care it delivers is too often fragmented, dis-connected, and isolated. And its technological disconnection is both a symptom and a substrate of this phenomenon. Physicians and nurses face many barriers in communicating amongst themselves, with their patients and with their patients’ caregivers. The current crop of EHR products do virtually nothing to address this problem. In fact, EHRs in the US may have exacerbated our health care dis-connectedness.

EHRs that can share data, information, and connect the experience of patients, caregivers and doctors more directly are much more likely to be utilized at the community level than EHRs that in essence capture and remove data, isolating them and their potential social uses in faraway databases that no one can get into.

The huge success of health-related social websites – like, DiabeticConnect,com and – are testament to the desire that many people have to close what Adam Bosworth has called the “collaboration gap” that stands between the limitations of the legacy health care system and the almost infinite benefits that arise from participating in self-help and online socializing activities. People who share their experiences – and data about themselves – know that this is helping them close the collaboration gap. But this gap is being perpetuated by EHRs that are organization- and enterprise-centered, and can only be substantially closed if physicians and medical groups in communities around the country use EHR technology to leapfrog over the communications and socialization barriers inherent in their older technologies. This will require new forms of EHR technology capable of socialization, which we have described elsewhere as Clinical Groupware.

Principle 5: Separate data from the applications and from the transport layer.

It is a stunningly simple yet powerful feature of the most familiar and widely-used information technologies that data – the message – is deliverable regardless of the sending or receiving applications, and independent of the network or transport layer that carries it. Email messages can be sent and received via many hundreds of client applications (what you and your computer use to compose the email or to display a received email.) Email and messaging services can carry many dozens of different kinds of attachments, e.g. pdf documents, across both open and secure networks, and networks with different kinds and levels of security protection in place.

This is a small planet idea that is the direct consequence of the openness of Internet protocols, but one that has not yet become incorporated in US health care, where data messages, applications, and network transport protocols remain unendingly, even stupifyingly, proprietary. Not only do these approaches perpetuate “walled gardens” – hospitals using one EHR system can’t send a simple electronic medical summary to another hospital using another EHR system across the street — but it also is a barrier to the innovators who would design, build and implement new, low cost applications like modular EHRs.

Clay Shirky makes this point in a blog post recently:

Thus the question for broad participation… is not: “What will the most complete system look like for the richest and most technically adept institutions?” Rather, it is: “What’s the simplest and most low cost way for a small vendor or new market entrant to get a small practice tied in?”

…Here’s what a workable set of transport standards will not do: It will not assume to know what kind of applications any given network participant is running locally. Once the data are delivered, it should be usable by everything from the simplest to the most complex application, since the recipient of the data will have the best understanding of what works in their local context.

This ability to separate data from transport and applications from data is the essential pre-condition for innovation — a group that has a valuable new idea for presentation of data for clinical use should not also be forced to think about the data encoding or the way the data are transported. Groups working on new data encodings should not be tied to a pre-existing suite of potential applications, nor should they have to change anything in the transport layer to send the new data out, and so on.

Patients and doctors in offices, homes, laboratories and pharmacies most often need information, and most often they need it in the form of small amounts of summary data such as a medication or problem/diagnosis list, a specific allergy, a limited number of recent or historically important lab tests or images. Where there is continuity of care and information flow, especially, there is rarely the need to access the complete or comprehensive medical record or its full contents.

For most ambulatory and outpatient clinical care needs, simple dashboard and summary health “EHR light” products may be sufficient, and there is a logical progression towards more complex health IT as the acuity of care increases. Modular design of EHR technologies may help to bridge this gap without creating large discontinuities of user interfaces and may also keep prices for health IT in the community setting at a lower point than otherwise.


In the U.S., many of our health problems result from the growing burden of chronic diseases occasioned both by an aging population and our sedentary lifestyles. In much of the developing world, by contrast, the local health problems – pandemics like HIV/AIDs, malaria, and drug-resistant tuberculosis – result from poverty and a lack of basic public health resources. However, similar EHR technology in each of these settings can provide efficient health data exchange and information management. Both individual and population health status could be improved with medical records that are inexpensive, simple to use, and capable of network exchange

To this point, each of the above principles for small planet health IT is already being put in place effectively in many developing countries, where cell phones are used to remind patients of their medication regimens and are the vehicle for relaying laboratory test results and vaccination information from provider to provider in sparsely populated and very resource-limited communities. As part of the Millenium Villages Project in Ghana, for example, cell phones are part of a program that is dramatically improving the chances of survival for pregnant women and their newborns.

Our brethren in other countries, developed and developing, face many of the same challenges obtaining good quality health care that we do here in the United States, including realizing the promise and hope offered by health IT. If we persist in federal EHR policies that “over-serve” local US communities’ needs by developing complex and expensive systems of health IT, we may not only be missing the mark at home. We might also be missing the opportunity of helping the other inhabitants of this small planet.

David C. Kibbe MD, MBA and Brian Klepper, PhD write together about health care technology, market dynamics and reform. Their collected writings can be found here.

Leave a Reply

22 Comments on "EHRs for a Small Planet"

Vikram C
Jan 5, 2010

Determining business objectives for IT implementation is a very important aspect. Here are some, looking at the wide landscape.
1. Disease outbreak control
2. Data bank build up for health actuarial activities
3. Geographical medical services adequacy verification
4. Medical supply /procedure surge verification
5. Disease research & treatment effectiveness research
6. Provider selection tool
7. Provider productivity
8. Enhanced clinical decisions
9. Lower medical services utilization
Now onto to basic premise of IT implemntation. All human bodies and medical transactions provide opportunity to learn and then later re-use the knowledge. As of know we are throwing away valuable opportunity to learn and improve.
Now if you look at the list of business goals, the most valuable returns are obtained population level data aggregation. However we then run into jaded political quagmire about who should be the custodian of the data and the role of government.
The future of medicine is customized treatment. Consequentially the future of data also has to be focussed on smalles unit- each individual. A lot of user data is locked. Further the investor in health IT is not the prime beneficiary reducing the motivation to spend on technology. As such the role of government agency is inevitable.

Jan 5, 2010

A lot of lofty goals for EMR, obscured by the mumo-jumbo of technology enthusiasm.
“focusing on the local circumstances that result in fragmented, disorganized, and inconsistent health care delivery within driving distances of its citizens. EHR technologies could address communities’ continuity and access-to-care problems, and relate these to major preventative and chronic illness management challenges, e.g. vaccinations, obesity, and risks of heart disease.”
A. So fragmented, inconsistent care and access issues can be magically improved by just having the EMR “focus” on it? How so?
B. We already know a great deal about risk factors and chronic illness management, sometimes maybe more than we can use (how many of us do know that Parkinson’s disease is negatively associated with smoking?). Do we really need an EMR to figure out that many patients are obese, lead a “sedentary lifestyle” (as later mentioned) and get type 2 diabetes with all its complications, or that others are inconsistent with medical care for their hypertension?
Moreover, general problems with data collection were omitted (e.g. the question about the actual quality of collected data, or the fact that, if data is not perfectly organized, the overwhelming flood of information can obscure relevant information).
If one really wants to connect the health care “fragmented gardens”, one needs either central data collection (e.g. the VA system for everyone, or something like a “medical history facebook” page, which of course needs to be near mandatory in order to be useful) or perfect interoperability (i.e. ability to access all patient data with a single query). After one decade of using 3 different EMR, I am not sure whether the latter is technically feasible with a realistic effort.

Vikram C
Jan 5, 2010

rbar- It’s not that we know everything. Not about autism, not about alzheimers etc. Now the things that we know, such as Parkinson’s disease management’s negative relation with smoking, my question is how many years did it take and how much did it take to get to that information?
Also this thing about us knowing all we have to-drug companies making products and throwing darts out there hoping some are benefited. There is no way they can make a drug that will have consistent results for all. We just dont know enough.
I concur centralized data collection would be required. More than that I would say we will have to take control of destiny collectively (yes that means Government will be involved) and do things which are important to society and not just wait on business until it becomes profitable opportunity. We already have a public system as far as germ sharing is concerned. We share syphilis, AIDS, malaria, cough, cold what not. I wouldn’t mind sharing my health information, DNAs etc whatever it will take to fight back ever mutating virus and bacteria.
We will get there eventually, but in meanwhile the more the digital records we have the better. Hopefully we will have situation like where Google will set about business of assimilating all health records available.

Vikram C
Jan 5, 2010

rbar- Dr. Lippiton mentioned 80% of PCP visits are due to hypertension. It surprises me. I would like to know that. EMRs could have helped in that regard.

Jan 5, 2010

While well written, the report is frightening, and the comments thus far are horrifying. Until the HIT becomes meaningfully useful to the users, namely the doctors and nurses, enabling efficient care of patients with complex illness, the best strategy is do not buy, do not buy, do not be duped by the “free government money”.
And Dr, Kibbe, will you please explain your arrangement with HIMSS, being a keynote speaker at HIMSS Annual in Atlanta? Are you paying your own way or are you being wined and dined?

Jan 5, 2010

I think we would all agree that we are never going back to the old system and that EHR is the future. If that is the case, then I agree with the author that simplifying the design and delivery of EHR technology would benefit us all. It does not make sense to have large, cumbersome, enterprise encompassing systems rolled out for all physicians in all practices. A light, modular web based system makes more sense. This would allow each physician to tailor the EMR to his practice specifics by adding modules as necessary. This is not unlike the application concept with the Iphone. The key will be to develop and “operating system” that allows the data to flow between these modules. We all know this technology already exists in just about every other enterprise on the planet…why not medicine.

Jan 6, 2010

Hoorah, hoorah, hoorah! This is the first totally “disruptive technology” manifesto I’ve seen on this blog. And it’s very well done. (There may have been others that I haven’t seen.)
What’s disruptive about it? Most important, it completely disregards the establishment. It has no interest at all in tying into the establishment. This is a source of so much freedom that those tied to the establishment can’t begin to imagine it. And the freedom allows innovation in radically unrestricted (and unpredictable) ways.
(I say that as someone who saw his previous industry destroyed by disruption. I marketed typesetting machines. When desktop publishing came along we in the establishment said “You dorks, you don’t have X and Y and Z features. And you have no IDEA how hard those are to develop.” But the Pagemakers of the world started with simple things that end users found valuable – newsletters, at first. And four years later they DID have X and Y and Z – and their users were completely free of the establishment.)
Y’all in the big-iron world, you can disparage me all you want, but I *am* you when you do that, and I know how the movie ends.
The other radically disruptive aspect is that it’s all about giving end users (patients) what they want, solely for the purpose of getting people to use it. And here this industry’s very different from typesetting, because we already have tens of millions of Americans who’ve been priced out of the market and are fending for themselves on a shoestring. (Yes, the third world has come to America. By the millions.) Give these people something valuable, for cheap, and it’s hotcake time.
2009’s events in Washington convinced me more than ever that the industry is never going to give in, and is too massive to be stopped. The only salvation will be to disregard it and go around it. And just as with our typesetting machines, the industry will have no idea how to compete.

Nurse Missy
Jan 6, 2010

“2009’s events in Washington convinced me more than ever that the industry is never going to give in, and is too massive to be stopped.” _E-patient Dave here
The fraud being exposed at i-soft as reported at e-health insider may be the tip of the iceberg and serve as reason for the above observation. The adverse events we experience are covered up as are isnsider trading deals and kickbacks. That is how the industry works. Do not expect savings or improved safety any time soon. Sad.

Vikram C
Jan 6, 2010

IT itself isn’t the one which is valuable. It is the data which is valuable. As long as data entry cost can be minimized, rest of the software features such as reports, alerts, search are easily constructed.
The longer you keep doctors and nurses away from IT the more you harden their stand against IT. The biggest problem of IT is not the quality, not the cost but the user adoption rate.
As an example look at Salesforce from Siebel. Entering leads data is least useful exercise for sales rep and they are most reluctant to enter the data because at the end of the year they will be evaluated for sales they bring in and not data entered into the system. Yet, smart organizations know that once a sales rep goes, he/she taks along with the contact data and the sales history, which they can ill afford to lose.
Going back to your advise on -don’t take IT until proven 100% useful, I would say give it a chance. Bumrungrad in Thailand prides itself on having a great HIT. Why should it be any different elsewhere?

Jan 6, 2010

You correctly describe healthcare as “an industry that has somehow walled itself off from IT modernity”.
That “somehow” is being attacked and disrupted by what Clay Shirky refers to as the convergence of social networks and electronic networks.
In the short run, I’d continue to bet on the existing medical-industrial complex (successfully) to resist.
But in the long run you’d have to believe that healthcare can continue to exist indefinitely in a paper/phone/fax world to keep the walls of isolation propped up.
Kibbe/Klepper — continue your insistence to “tear down those walls.”


With the recent HHS standards proposal, it’s great to to see David and Brian open up the conversation on EHRs. You’re right to look at standards from both global and individual perspectives.
From a global perspective, I’ll merely note that a major reason the conversion from document filing to data filing for public company financial statements worked well in 2009 was because the U.S. worked with Europe and Japan. Even though Europe and Japan use different accounting standards, all picked the same technology standard — XBRL — to apply to their different accounting standards. The ability of each jurisdiction, and others, to leverage the global standard made the process faster, cheaper, easier, and better for every jurisdiction.
From an individual perspective, the quote from Clay Shirky is a good starting point. As ePatient Dave knows well, this is not a data issue but a people issue.
Resolving the data silo problem by focusing on the data in the silos will simply perpetuate the EHR problem. Moving data from analog (i.e. paper) filing cabinets to digital filing cabinets simply relocates the problem. Quite simply, the Internet is broken. Securing user data scattered among countless web silos is complex and consumes huge amounts of our most valuable resource: time. New identity standards or fancy new EHRs, alone, aren’t going to fix things. A full paradigm shift is needed, a shift that focuses on patients in health care — and on people in every domain in which making data more useful can improve outcomes.
CLOUD’s emerging technology standard is that shift. It goes far beyond identity, to empower Internet users to control precisely how their information is used. Think of it as privacy and authenticity standards that work — not a confusing Web-based control panel, but standards to let anyone — user or service provider — develop tools that are simultaneously more sophisticated and easier to use. Doing this requires a shift in thinking equivalent to what HTML brought to the Internet 15 years ago: this time, though it is a mark-up language not for text, but for people. It’s a mark-up language that supports Internet connections that transcend the browser paradigm that’s consumed us since the 90s.
CLOUD believes the keys to adoption are, one, Local Ownership and Use of Data — hence the name of our Consortium. Two: to break down health, finance, education, and other silos to simply connect people, not industries. Just as people use one main standard to connect text on the Internet, we should use one main standard to connect with each other on the Internet. And three: empowering people to separate their identity from their data in every silo. Thus, privacy is ensured and the economic value of connections among people and their data grows. It’s ME 1.0, not just Web 2.0.
The following white papers show how this new approach not only improves things within health care (using clinical trials as an example):
but across domains (using immunizations as an example):
Since we use our school system to track immunizations, having this health record easily available for educators is vital. A new standard for EHRs or a new standard for SISs won’t work. The only thing in common between the health system and the education system, in this case, is the student/patient.
More info can be found at

Jan 7, 2010

Great blog!
I have been working on health information systems in developing countries for some time and it is interesting that a lot of your principles are the same as what I have been proposing to resource poor countries. The ideas of keeping things simple, incremental development, local use of data, minimal data set, appropriate technology and of course data standards and interoperability.
I like your Principle 4… that the desire to connect is what sustains systems. This is very insightful. One of the biggest problems we have is designing systems that will continue to be used and be effective after the initial funding runs down. All of your principles help in this regard but the most compelling is to make applications useful and the insight that it is the connections that make things useful is profound.
Thanks for this post.

David C. Kibbe, MD MBA
Jan 7, 2010

Dear MarkS: We’re really gratified that people such as yourself you who are working in health IT in developing countries find that our ideas in EHRs for a Small Planet are valid. It’s the sad disconnect between US-bound health IT and more global settings that we’re trying to address here. Your feedback is very important. Thanks so much. And good luck with your work. Examples and stories from your work would be welcome.
Regards, DCK

David C. Kibbe, MD MBA
Jan 7, 2010

Dear Gary: Many thanks for your comments. We’d love to hear more about how the use of XBRL by Japan and Europe evolved, and influenced behaviors in the US. Good white papers, too. Very thoughtful. Regards, DCK

Jan 7, 2010

Hey Dr. K:
Who is paying your way as the keynote speaker for HIMSS?
Are you under the financial influence of others?