Monthly Archives: April 2010

Cathedral & Bazaar Revisited: Healthcare Information

One of the problems we face when we talk about electronic medical records, personal health records, etc. arises because we think and talk about them as systems. A recent post on LinkedIn says:

What was apparent from the first 23 comments is that opinions and perspectives are all over the place. It’s not clear folks are all of one mind with respect to WHO constitutes the users, WHAT constitutes an EHR, what constitutes usability, and/or HOW one should be assessing usability. Many look at usability only from, ultimately, the safety perspective (decreasing medical errors), but how about efficiency, including impact on overall workflow? User acceptance/satisfaction? What are the appropriate usability measures by which to evaluate EHRs?

These are systems related questions. Systems lend themselves to being defined, developed, deployed and used. That is not the case with healthcare information. The information requirements of healthcare providers, patients, and the supporting infrastructure are evolving rapidly as we learn what works and what doesn’t. Supporting technologies – iPhone and iPad to name just two – are evolving and enabling new cost effective services to be provided. The economics of health care are changing because the current medical business model doesn’t fit any economic model that makes sense and the economics of health care are what will provide much of the funding for information solutions. The problems and opportunities of health care information don’t lend themselves to the discipline required for traditional systems.

In 1997 Eric S. Raymond published an essay called The Cathedral and the Bazaar that described the traditional system development process as centrally managed and built to last like a cathedral as contrast to a bazaar that is constantly being modified by its users to meet their evolving needs—in his essay: Linux. The article was sold and is now copyrighted and available only for a fee.[1]

Eric now points to In Praise of Evolvable Systems by Clay Shirky which points in the same direction. I think Shirky’s definition of Evolvable Systems provides and apt description of what is required to realize the benefits promised by improvements in health information:


Evolvable systems — those that proceed not under the sole direction of one centralized design authority but by being adapted and extended in a thousand small ways in a thousand places at once — have three main characteristics that are germane to their eventual victories over strong, centrally designed protocols.

  Only solutions that produce partial results when partially implemented can succeed. The network is littered with ideas that would have worked had everybody adopted them. Evolvable systems begin partially working right away and then grow, rather than needing to be perfected and frozen. …

  What is, is wrong. Because evolvable systems have always been adapted to earlier conditions and are always being further adapted to present conditions, they are always behind the times. No evolving protocol is ever perfectly in sync with the challenges it faces.

  Finally, Orgel’s Rule, named for the evolutionary biologist Leslie Orgel — Evolution is cleverer than you are. …  it is easy to point out what is wrong with any evolvable system at any point in its life. … However, the ability to understand what is missing at any given moment does not mean that one person or a small central group can design a better system in the long haul.

Evolution is messy, brilliant ideas don’t work, money is wasted, efforts are duplicated, but the Internet has shown that the process is capable linking growing requirements with expanding capabilities to produce solutions to problems we don’t even know we have. The Internet is a better model than traditional, cathedral like systems for what we are dealing with when we talk about converting and sharing medical information in an electronic format.

It’s dinosaurs vs. mammals, and the mammals win every time. … Infrastructure built on evolvable protocols will always be partially incomplete, partially wrong and ultimately better designed than its competition.

There will be some large systems to deal with complex environments. Small systems to deal with special needs. “Apps” to deal with general needs, and forms we haven’t imagined to deal with opportunities undreamed of. We need to recognize that lack of clarity and structure is just part of the process. We need to learn to live with it and occasionally laugh at it, curse it, and celebrate it

[1] Eric S. Raymond (1999). The Cathedral & the Bazaar. O’Reilly. ISBN 1-56592-724-9.

Short link:


Safe Patients, Smart Hospitals

If you are concerned about the quality of health care and rising costs, read this book.

If you are responsible for change management in a corporation, read this book.

If your are working on an electronic medical records system, read this book.

Peter Pronovost, M.D. and Ph.D. took on change management as a process to reduce the costs, and more important, the risks of serious illness or death associated with the practice of medicine. This book is not just about a good idea. It is about having a good idea and then doing what needs to be done to realize the benefits of that idea.

“We knew our success would depend on three key elements: Developing and unambiguous checklist that would encapsulate as much evidence or knowledge as we could capture on a particular procedure; changing the culture and associated broken systems to remove any barriers to implementing that checklist; and measuring results so we could gauge the checklist’s efficacy and provide feedback to provide whatever changes are necessary to improve it.”

“What was striking was that nobody debated the evidence, nobody challenged the items on the checklist, and nobody questioned whether we should do them. But everyone objected to the change in culture. … Perhaps the most radical step we took was to promote the checklist was we gave it to family members of patients. We explained how it worked and why it worked and encouraged these nonclinicians to observe and ask nurses and doctors if the procedures were not being followed. … Now that family members were aware of the checklist, they were asking useful questions that actually helped the nurses and doctors do their job.”

Great move, but one that requires clear support from the people responsible for performance in accordance with the checklist – doctors and nurses – and senior management. For a number of reasons that are covered in the book Dr. Pronovost and his various teams had outstanding support from the senior management of John Hopkins Hospital.

As a business analyst for process improvement and systems development, I find that good ideas are relatively common. Particularly those that include the results of shared experience and thoughtful consideration. “… changing the culture and associate broken systems. …” is the hard part and the part that is least understood by most champions of change. The way Dr. Pronovost and his various teams dealt with those issues is some of the most interesting material in the book.

By the way, very well written.

Safe Patients, Smart Hospitals
By Peter Pronovost and Eric Vohr
Hudson Street Press, 282 pages, $25.95

Short link:

EMRs & “The Cloud”

I ran across a Webinar from March 4, 2010 the other day about “the cloud.” The presenter was Steve Ballmer, CEO of Microsoft. It changed my understanding of where the cloud is and where we are going. He didn’t address EMRs specifically, but it easy to see how the cloud will be part of the solution to networking EMR and health information to make it useful, reduce costs, and improve outcomes. Excerpts from the transcription:

CloudSo, I’m going to give a little bit of perspective on the cloud really from the standpoint of people who get a chance to use it, to drive it, to shape it, to make something of it. …

First principle, the cloud creates opportunities and responsibilities.

The amount of invention that needs to happen is high. The world is still not a perfect place in terms of the commercial infrastructure. Yes, you can create a Web page and put on an AdSense ad. But, we certainly haven’t fulfilled the sense, the opportunities to create technology that empowers the creator. …

Immediately people get nervous, particularly when you talk about advertising. They get nervous, what about my privacy. And that’s why I think we have to talk about the opportunities and the responsibilities. The responsibilities for creators, for business people to respect the consumer, to build technologies that really do allow the user to be in control. … And yet I think we have a responsibility, all of us, not just to socially respect the user, but to build the technology that will protect the anonymity, the privacy, the security of what I say, who I say it to, where I go, what’s important to me.

Second dimension of the cloud: The cloud learns and helps you learn, decide and take action

The world is a large, complicated place. So, the first thing that got built to help people navigate was essentially directory services, search services. People built tools to help you navigate and find information, pull it all together, et cetera. And yet, we’ve got to go further than that. The cloud needs to learn about you and it needs to keep learning and figuring things out about the world that has been described virtually.

It’s great to know about 83 million Web sites on the planet, but if you’re actually trying to find something specific … I’ll put my hand up, as part of the U.S. healthcare debate I decided I should actually understand what we spend money on as a society. Try that one out for size. Pick any search engine you like and go give it a whirl. You’ll get a bunch of links, you’ll find a bunch of data, you’ll probably try to cut it, copy it, paste it, but you won’t be able to just sort of describe maybe like a simple, little chart that you would like to see populated. How much money do we spend on healthcare, how much of it gets spent on older people, younger people, poorer people, richer people, people in the last year of their life.

It’s only about eight numbers, there happen to be eight numbers that you can’t learn by following the public debate. But, there were eight numbers that I felt as a citizen I ought to know. But, the ability of the cloud to actually learn from all of the data that’s out there, and the ability of the cloud to learn from me what I’m interested in is not what it will be two, three, four, five years from now.

I happen to be a numbers thinking guy, I would create that little healthcare thing as a little spreadsheet. I would want Excel to just go get that stuff from the cloud. And so this notion of learning, learning about me, learning about the world, making conclusions, and then helping me to decide and take action, I think is a very big idea.

The cloud itself needs to learn. It’s got to collect new data. It’s got to sense new data. It’s got to represent the real world, and keep getting smarter and better, so that it can help me learn. … [demonstration using maps and photos to bring information from multiple sources together, including real time sources, e.g hospital, doctors, labs, pharmacies, personal health records.]

I hope the demonstration does a couple of things. Number one, I hope it kind of wets your whistle for some of the kinds of things that can be done. And number two, it really helps bring together this notion about learning about the world, how do we learn from others, how do we pool the data that’s available on the Web to learn about the world, and then map it and make it of interest to somebody in real life.

Third dimension. The cloud enhances your social and professional interactions …

Dimension No. 4, the cloud wants smarter devices.

The way in which we can learn about you, the sensors, the cameras, the voice, the gestures, today. This year, we’ll get about 10 billion utterances, speech utterances, submitted to us in the cloud through something called our TellMe Service, which handles call centers, and Bing kind of phone voice response searches, and the like. And so, the ability for the device to participate in connecting to the user, providing a richer interface, to get data back from sensors, and use that to improve the cloud experience on behalf of the users is really quite strong.

Later this year, we’ll ship a thing that we call Project Natal. It’s a camera that comes with the Xbox, and it recognizes you, and your voice, and your gestures. … The great smart device hardware is going to bring together the best of what we think of today as rich clients, and the best of browsers, and the best of a next generation of natural user interface, voice, touch, speech, et cetera, all in one unit. [EMRs and the user interfaces.]

Dimension No, 5, the cloud drives server advances that, in turn, drive the cloud.

Cloud Computing Remarks by Microsoft CEO Steve Ballmer
Paul G. Allen Center for Computer Science & Engineering, University of Washington Seattle, Wash. March 4, 2010



If I Were A Doctor … Impatient Patients

There have been a number of articles recently about the changing relationship between doctors and their patients, or what might be called their “impatient patients.”

As an impatient patient, why should I have be at your office for my first visit 15 to 30 minutes early to deal with paperwork that is essentially the same as I have filled out for every doctor I have ever visited. Give me a standard set of forms. Post them on your Web site and let me download them so I can do this one more time and then just update as necessary for future doctors. You and future doctors are asking for information about me; give me a tool and I will take care of that.

Better yet, help me find them in a place like Google Health or Microsoft Health Vault or somewhere similar. After I make the appointment, I can review the forms, fill them out and send them to you. Have a general medical set, insurance set, HIPAA set, specialty set, etc. In the future I may have to fill out a new form for a new specialist and update a couple of fields but that is easier than filling out every form for every doctor. In any case, let me do it on my free time rather than as part of my appointment.

It could also include a list of recent tests and a history of standard things like my weight, blood pressure, etc. that could be accessed in advance. If my standard form shows that I had a PSA test in 2009 and the reason I am coming for this appointment might make my PSA level of interest, your system (or assistant) could have that information in the office when I arrive.

One of these forms should also give you a list of my other providers in case you need to communicate with any of them before my appointment or want to send them a follow up report. Or better yet, let me know that you can use the list already in my Google Health file which includes specialty, addresses, and phone numbers.

It should include a list of my meds from Google Health or a similar service. This will give you accurate data from my pharmacist that is date stamped so you can see what I have taken and am taking. Almost certainly better than the records I bring with me today. Saves me some more time and you get better data. That will also give you information about my pharmacy of choice.

Send my prescription to my pharmacy of choice. Why should I have to make two trips to the pharmacy? One to drop it off and one to pick it up or wait 30 minutes. I want to make just one stop and not have to wait to pick it up.

We all know doctors run late. You truly have life threatening emergencies. But that doesn’t mean I should sit in your office if I have something better to do with my time — I almost always do even if it is to go someplace private and make some phone calls. (Don’t you hate it when people make cell phone calls in your office?). If you are going to run 30 minutes late, have your appointment system or assistant call me as soon as you know. Ask me if I would like to be notified if you are delayed further, push “1” for yes or “2” for no.

If you get lucky and make up some time and get to your office before I do, take care of the little old lady who was scheduled after me but has been sitting in your office for more than an hour because she doesn’t have a cell phone. She will appreciate it and I will appreciate the reduced impact of your unavoidable lateness.

Stay open to iPhone apps and other tools that will help both of us make better use of our time. Not for all of your patients — that day is probably somewhere in the future. But start using it now. You and I other adventuresome folks can experiment knowing that we will encounter a learning curve to be climbed.

I know you have better things to do than play on the bleeding edge of technology. On the other hand, if I work through the learning curve with one of my other doctors, I will expect you to have it right the first time we use it. And, if I start with you, I will be trained to do it your way. If you are reading this, I would rather work with you.