The Limits of Open Source - A New Model for "Open" Source in Medicine

I'll use an example from the business world to show how a variation on the open source model could make a large positive difference where it currently does not. In some important ways, Medical office practice management is a good candidate for open source development. To some extent doctors are in competition with each other but generally any competent doctor has a reasonably full patient load. If a doctor doesn't have a full patient load, it's not because his or her office uses the wrong office / patient management software. Only the absolute worst billing nightmares or horrendous scheduling (long patient waits) problems will ever cost decent doctors their patients. The success of a medical practices depends the doctors' quality and how they deal with their patients.

It's the doctor and the doctor's staff that deal with the office / patient management system. Doctor's should ask themselves, how good is their current system, how much does it cost them, how many computer systems have they been through, and do they have, does it exist, or can they afford, the system they really want? From what I've observed while waiting to see doctors, I don't think most medical practice software is very good. For example, on my last doctor visit, they couldn't schedule my next visit because their Windows 2000 system was down. There was a technician working on the problem who was clearly stumped, as he was calling asking for advice what to do next.

The medical industry could do better because office management software is an area where medical practitioners generally share similar needs with most others. Under the initiative of the American Medical Association, a seed project could be formed where the AMA would bring in some of the leading developers in the field to create a complete new practice/patient management system but instead of the developers owning the project, they would be paid well for their time but the result would be an open source medical practice management system built in a modular fashion to cover the full range of needs including office visit, lab and therapy scheduling, billing, electronic patient record management, prescription tracking, conflict management and ordering, and whatever else was appropriate. The system would need to be configurable for a sole practice to a corporate practice with hundreds of doctors and handle GPs as well as any specialty.

For those who have never worked in the not for profit sector, it's important to understand some basics. Normal businesses can always come back to the bottom line and ask whether or not any initiative will increase profits. Not for profits have non financial goals but these are typically general enough that any of several competing goals may be competing for top priority. Elected leadership may change high priority projects in short order.

Still, all not for profits know they cannot successfully pursue any other goals without members and funds. For organizations with a political or social agenda, influence is likely to be directly proportional to the number of members and the size of the affiliated PAC. Besides dues revenues and member contributions, major sources of revenue for most professional associations are advertising revenues in the association's periodical publications and exhibitor revenues at meetings, conventions, and educational events. Often the largest component of this revenue stream is from vendors of vertical market software aimed at the association's members.

A number of developments in the online world are presenting serious challenges to the traditional methods of operation for professional associations and threaten both membership bases and revenue streams. Associations that do not adapt to online conditions face possible extinction but at the same time most associations have a high level of respect in their respective areas and thus have a natural advantage over potential online competitors, if they use their advantages wisely. Online sales represent less than 2% of retail sales and just as brick and mortar are hardly doomed, associations generally have time to figure out how best to fit into the online world.

A project where the AMA actively participated in the development of new practice management software would be highly political, and the AMA would lose a significant amount of revenue from their current vertical market software advertisers and exhibitors, who the AMA was in effect working to put out of business. Done right, they should be able to build a system that's better than any proprietary system available.

To mitigate the financial damage to the AMA, a standard open source license would need to be modified. Use of the system could be limited to AMA members. Except for this limitation the rest would be like open source so that any doctor who needed a special modification could hire a consultant to add it, and if generally useful, it would go back into the standard system. The license could even provide that anyone who used the software without formally joining AMA, would automatically become a member and be assessed past dues for the time they used the software. I know some open source purists will be highly offended by this suggestion, but I think it preserves the cooperative nature of open source while dealing with market realities.

www.gnu.org says "``Free software'' is a matter of liberty, not price." The modification to the model I'm proposing makes the software free by every definition except price. It's free to use and extend if you belong to the AMA community. You simply have to pay to get into the AMA community. It does unfortunately create a licensing conflict which is similar to the problems with mixing GNU and BSD style licensed software and would not be compatible with either group.

Why would an organization like the AMA need to be involved or why would an open source license need to be modified? The answer is simple, costs. My guess is 5 million dollars to do the first fully functional system, after which it would available for use and development under a modified open source model that I'll refer to as available source. In my last job, I had the opportunity to estimate the costs of three large projects before any design was done or contract signed; in each case I came within 10% of the actual costs. (This accuracy required careful thinking and extraordinary luck.) Here my figure is a guess (not an estimate) with a range between $500,000 and 10 million. I know enough about the complexity, that anything under half a million is has no chance of doing anything useful, and I very much doubt even the AMA could afford 10 million. Even 5 million would really test AMA's commitment and ability to set priorities and divert funds.

An alternative might be for the AMA to buy one of the existing proprietary vendors. I doubt one that had a really comprehensive package and was doing well enough to be desirable would be affordable but only someone who did serious investigation of the market would really know. In addition to having a very flexible feature set, the product would have to be based on high quality code that was very modular and lent itself well to further development.

What I'm suggesting would not draw typical open source developers. The initial project presupposes paid developers and I would assume that modifications would be paid consultants. Since support is one of the revenue models in open source, i.e., developers being paid to make modifications to existing systems, unless a developer was committed to working only on pure open source systems, there is no reason that someone who worked in the open source community could not also work on a project like this.

The reason it makes sense for a professional association to undertake a project like this, is due to the problems with existing proprietary software, especially comprehensive vertical market management software as discussed on the preceding page. The available source model discussed here, eliminates the maintenance lock-in that vertical market vendors typically have over their customers. A project like this could provide funding far beyond what any startup firm in the field could hope to attain, and draw on a much wider client audience who could catalog the problems with their current systems and desires for a new one. As the professional association that covers the entire medical profession, AMA, could provide a benefit more substantial than any they have provided previously and for the foreseeable future, insure membership at growing levels.

Why are pure independent open source projects unlikely come up with viable solutions? This is discussed generically on the preceding page but a look at some current attempts should show the problems faced by traditional open source groups trying to develop comprehensive medical office practice management applications. There appear to be four such projects today. A review of the long list of medical related open source projects at www.openhealth.com shows 5 candidates. sample A www.linuxmednews.com list of projects had the same 5 projects but has been replaced by http://www.linuxmednews.com/LMNProjects, a model of inchoherence. GNUMed used to list the other four under "related links" but no longer seem to. The GNUMed list also included Circare: "Minoru Development Corporation is initiating the Circare project to provide key infrastructure for Regional Health Networks as open source products." This may be an important medical open source project but is not the comprehensive practice management software that I'm discussing.

In addition to GNUMed there are 4 other projects that appear to qualify. These are FreePM, Freemed, Littlefish and TK Family Practice. I'm not including Littlefish because repeated attempts to reach their web site have failed and GNUMed describes them as 'a "close to" open source project with a similar purpose as GNUMed. Unfortunately still based on a proprietary database core.' I don't think this qualifies as open source. Despite multiple references to "FreePM", I can no longer find a site for it. Freepm.com is now a sex site and freepm.org seems to be a "spam" product site. For now, I'm leaving the description below I wrote some time ago regarding FreePM, which may explain why it appears to no longer be available.

Thus I see three open source projects that attempt to provide comprehensive medical office management software: Freemed, GNUMed, and TK Family Practice.

Regarding TK Family Practice, GNUMed states "One of the few open source medical packages already up & running in daily practice." To find out if this is actually used, I had to follow their link to SourceForge where they say "Used in a 4 doctor group for 5 years." TK Family Practice, in the tradition of most successful open source projects, was written by someone, Alexander S. Caldwell, M.D., who needed "help accomplishing certain tasks in my daily routine." It was first "used in 1996 when it sent it's first computer generated fax prescription refill."

TK Family Practice appears to be by far the oldest of the open source medical practice management products and the only one that is clearly in use in a real practice. Despite approximately a three year head start, three or four other open source medical practice management projects have been formed and none of them appears to be using TK Family Practice as a starting point. Why? Is it the use of Tcl/Tk as the programming language? Was TK Family Practice coded fast and dirty? Does it assume too much based on how a family practice rather than other medical specialties work? I have no way of knowing.

What I do know from the SourceForge page is that there are only three developers involved with TK Family Practice. One of the others besides Dr. Caldwell is also an M.D. Besides this we know nothing of either. They may both work for or with Dr. Caldwell. For whatever reason, other persons interested in open source medical practice management have not adopted TK Family Practice. Open source depends on numbers and three is not enough. For unknown reasons, TK Family Practice does not appear to have a significant future as open source medical practice software.

The FreePM.com site had a semi-slick presentation and mentions a few future capabilities: "Patients will be able to personally enter data such as Past Medical history, Address and Phone number changes. Patient review - patients will be able to securely login and view their medical records." From the what I know about doctors and medical practice management software, I suspect this would more likely scare doctors than induce them to try the product. Since almost everything on the .com site is about technology and the related www.freepm.org site was even more technology related, the very LAST place I looked for a product description on their site was under the "technology" heading of the .com site. Here and not under "products" or "about us" I finally found a product description. In its entirity:

Seamlessly integrated "Point and click" Electronic Medical Records AND Practice management.  °  Universal Web Browser (requires, ie. NS 4.x, IE 4.x or other frames capable browser).  °  Comprehensive patient, provider, and resource scheduling module.  °  Electronic prescription writing with drug interaction database.  °  Managed care contracting functions that help prevent underpayment and unauthorized care delivery.  °  Integrated email system.  °  ICD-9 and CPT coding on the fly.  °  Electronic insurance filing.  °  Integrated drug database.  °  Template driven EMR.  °  Template generator included.  °  Integrated employer & insurance plan information.  °  Complete Customization available.

Not very comprehensive and obviously in the wrong place on a the web site but it's at least a start. If they put this in the right place, where a doctor might actually find it almost immediately, they might at least get a phone call to their 800 marketing number. Given their current site organization, no doctor is likely to explore the site thoroughly enough to find it. They have no user guide. When I tried their demo I encountered some error messages and circular links that lead nowhere (home page to empty page with only a link back to the home page). Some would dismiss this as just marketing but we exist in a market economy. Marketing is often the difference between success and failure in a market economy.

FreePM has a SourceForge link. They don't appear to be doing anything with this link. Some comments were posted between late October and early December 2000 but that's it. Seven developers are listed; none chose to reveal anything about themselves. This is not encouraging for an open source project.

The remaining two projects that are entirely open source medical practice Managment projects, Freemed, and GUNMed share some common characteristics. The key point is that they tell us almost nothing about what they do as applications. I've included them in this grouping because of their names and brief descriptions found on other web sites. We don't know if they are aimed as small practices or large, whether they are for one medical specialty, several or attempt to be neutral regarding medical practice areas.

Actually someplace on the GNUMed site I stumbled across a mention that it is aimed at a General Practice. I couldn't find this again when I tried to. They admit that the project is in an "early development / planning stage . . . not in a state yet where anything could be used yet in daily practice." Still there are 17 developers on this project, 13 of whom are Doctors. You'd think that somewhere they would be emphasizing what the application did or what they hope it will do rather than talking about languages, databases, three tier system, 3NF normalization, and modular architecture. About the best that you can do is infer from four screen shots what it actually does.

The GNUMed project is not totally unaware the problem. They are asking for people to help with "Documentation: it would make sense to start writing the documentation along with the development of the software, otherwise we would never get proper end user documentation." Still, it makes me wonder if open source turns everything about software development on it's head. I know from personal experience that lots of software development never goes through a formal development process. Things like utilities and even operating systems sometimes do better just evolving. You put in what you think it needs and work with. Then you realize something else would be really useful and add it.

I just can't see this working with a comprehensive vertical market application. It seems obvious to me, that if you don't go through a formal requirements analysis and design phase, that you'll end up with some gapping holes at the end. I don't see how letting everyone do what interests them can possibly make a complete project. A requirements analysis would look a lot like a detailed description of what GNUMed was supposed to do. A few wording changes should turn it into the first draft of the marketing oriented web pages. That there are no such pages and that they are asking for help with documentation at this late stage, strongly suggests they haven't done traditional requirements and design, don't have anyone interested in documentation and most likely have a product missing some very important functionality.

Freemed suffers from problems similar to GNUMed but appears smaller with only 4 or 5 doctors. Their idea of a features page lists: new download site, bug tracking, cost effectiveness, upgradability, etc. It also discusses three actual product features: Electronic Data Interface, High Security, Language Agility which have descriptions of a few paragraphs. Their web site categories are as confusing as any I've ever seen. Not counting the under construction pages, several categories are pure guesswork as to what you'll find.

It would be great if one of these projects lead to a successful medical practice management suite. Widespread use of low cost, high quality software in the medical field would have a small but beneficial impact on the high cost of medical care that we all contribute to one way or another. Still, if I were a betting person, I'd not put any money on any of these projects succeeding.

To have a chance of being selected, an open source product will need to be compared to multiple proprietary products. If the buyers are on the ball they will go through at least a two step selection process. I haven't seen any open source comprehensive vertical market management product that deserves to be included in any first cut let alone a final selection.

If buyers are really doing their job they will have a formal Request For Proposal (RFP). A serious response to a sophisticated RFP can easily cost tens of thousands of dollars. For a small doctors office the response is likely to be around a thousand dollars. How can open source models based on service revenues even participate in this game?

The answer is simply that they can't, at least not on their own. Only a major professional association has the stature and organizational wherewithal to put together a project like this, where the profession has hundreds of thousands or more members.

If a doctor's selection of practice management software is not driven by a formal selection process, the slick marketing materials of a proprietary company are even more likely to sway their judgment. The one thing that would influence most professionals more than marketing material is the active involvement and endorsement of the professional association to which they belong.

Professionals repeatedly turn to their professional associations for advice on software products to use. By working with their members to develop a new practice management application, a professional association has the ability to develop a product that significant numbers of the association's members would actually try. Especially those actively involved in the design of the product would likely be early adopters. Others switching products or entering the market for the first time would be other prime candidates for adoption.

Often members of a professional association are desperate for a recommendation from the association but the associations need to be very careful about doing this because typically they have not been involved in the development of the product. They also fear loss of revenue from vendors not endorsed. Where they have been involved and the product is being adopted by active members, there is no reason not to actively endorse the product they've built.

Unless the project has been miserably managed, the result will be good and get better the more widely it's adopted. There is no reason over time, that such a product should not become a true industry standard, being extended as far as necessary. Leading the development of an industry wide software system gives a professional association an opportunity to really live up to what is implied in the word "association".

Large professional associations can look at the successes of the large open source projects, study the state of vertical market software serving their industry and potentially, find a better solution for their members. What better way to assure membership than to get for the price of dues, which most professionals should probably be paying anyway, a software system that might otherwise sell for $5,000 to $500,000, with annual support costs / maintenance fees ranging from 10 - 30% of the original purchase price.

Done right, a project like this could guarantee an association's survival. The value to the professional, of a project like this done well, will be so much higher than association dues traditionally are, that the association could allow dues to float upwards by large amounts without coming close to what the members would otherwise have had to pay for management software.

The association should continue to shepherd the project. Instead of each practitioner who needed an enhancement doing it separately and merging back into the common source, the association could maintain lists of desired features and what the features are worth to those requesting them and coordinate developments so that costs were spread among those requesting the features. They should also set standards for those who were in a hurry, defining standards for patches that would allow changes made by individual members, if approved for inclusion, to be merged into the standard source tree. Since the value to users of the software, should be much greater than a typical proprietary counterpart, they should be willing to pay more than they would if they were already paying traditional license and maintenance fees. This would assure the continued growth and vitality of the project.

To date the successful open source projects that I am aware of have been in computer related fields where the users possessed the skills to write software that satisfied their own needs. The large cooperative projects are almost purely computer infrastructure. As you get to more specialized projects the number of authors decreases. In some cases, it may be just one and we can only guess at their motives for going open source rather than proprietary. A project as ambitious as GIMP has only two primary authors; are they software developers who love graphics or graphic artists who love computers?

When you get to professions like medicine and law where the computer is simply a tool, only an occasional hobbyist practitioner is seriously interested in computers or software. Many still see it as something for the office help to use. It's possible there are enough practitioners really interested in computers to build an open source product that can manage professional offices. Will they also have the skills to get it used by many of their colleagues? I suggest available source as a mechanism by which a few influential professionals, might use their professional association as a way to get out of the clutches of proprietary vertical market vendors.

To be honest, I would not bet on the suggested AMA project either. Though the AMA or other large professional association (in its own field) has the size, organization and financial resources to succeed at a project like this, I expect the lack of understanding of technology, lack of vision, politics, and excessive concern with short term revenues would doom most such projects.

Though few areas would have more to gain from either open or available source software than what I describe as comprehensive vertical market management software, I don't see available source happening in any field or open source being successful in any but the most technical industries in the foreseeable future.

Top of Page - Site Map

Copyright © 2000 - 2014 by George Shaffer. This material may be distributed only subject to the terms and conditions set forth in http://GeodSoft.com/terms.htm (or http://GeodSoft.com/cgi-bin/terms.pl). These terms are subject to change. Distribution is subject to the current terms, or at the choice of the distributor, those in an earlier, digitally signed electronic copy of http://GeodSoft.com/terms.htm (or cgi-bin/terms.pl) from the time of the distribution. Distribution of substantively modified versions of GeodSoft content is prohibited without the explicit written permission of George Shaffer. Distribution of the work or derivatives of the work, in whole or in part, for commercial purposes is prohibited unless prior written permission is obtained from George Shaffer. Distribution in accordance with these terms, for unrestricted and uncompensated public access, non profit, or internal company use is allowed.