I had dinner with a friend of mine whom I hadn’t seen in a few months. Turned out that she had been mugged in the lobby of her apartment in Hell’s Kitchen, so she’s spent the last few months moving to a safer neighborhood and dealing with all the pain that goes along with that. How sobering. Thanks to self-defense classes that she took in college, she was able to ultimately fight off her attacker by pushing him away just long enough for her to scream for help.

You can’t scream when he’s pressed up against you.

The bastard actually impersonated a police officer in order to get access to her building. Her one piece of advice for me (or anyone) - “Learn some self defense,” and blowing away any myths I might have believed, “Men are targets as often as women because they tend to have more money on hand.”

I’m slowly in the process of moving my site from PHP/MySQL to OpenACS. Things are broken in the meantime. Specifically, I need to import all my old photos, steal MarkD’s book module, and implement my family-tree module. If you find something that’s broken, feel free to let me know.

I’m currently using an open-source stylesheet from The NYPL, but I need to edit it at some point. But for now, at least, it looks better than anything I could drum up by myself.

My introduction to computers came when I was about 6 or 7 years old - I can’t quite remember exactly when… Venu Mammen (my uncle) brought home a terminal from work. It looked like a typewriter, but it had 2 suction-cup-thingys where you could attach a phone headset. (Phone headsets were pretty much all the same shape back then). Venu Mammen issued some magical command and then up on the screen came the most amazing phrase. ADVENTURE v1.0. (Actually, I can’t remember for sure if it was a LED screen or if it was a dot matrix printer output - Do you remember Manu?.) Anyway, I stayed up late that night (read: 9PM) moving around an endless maze without really ever getting anywhere. I always sucked at adventure games… Despite this failure, I was hooked on computers.

Sinclair ZX-80

Around 1980, my father brought home a Sinclair ZX-80. This computer had a whole 1K of RAM, and if you filled up the RAM - no need to worry. Just hook it up to any handy portable cassette recorder and you could save all the information you’d ever need (well, about 20K or so). I think my brother wrote a ‘number guessing’ program… I was so impressed. He saved the program on a cassette tape and I still remember trying to listen to the tape on the cassette player, half expecting it to speak, “Guess a number from 1 to 100.” No luck, though - programming apparently was a bit more complicated.

Apple IIe

Some time later, we moved up to the big time (in my estimation). An Apple IIe with the Apple III monitor, 1 MHz processor, 64K of RAM (which we later upgraded to 256K!!!!), 2 floppy disk drives and AppleWorks software. This computer served me throughout high school. I began to dabble with programming in BASIC and assembly. I think that Apple is still sitting in my parents basement. I’ll have to see if I can load up One-on-One again…

My dad bought a Mac Plus for work when it came out and he let me play on it. I couldn’t believe what I was seeing. No command line interface. A little arrow that moved in response to my movements on a “mouse”. Folders that you could move around. I was in heaven! I perused all the manuals that came with it trying to learn as much as I could about this new beast. Around this time, my dad asked me to help him computerize the medical testing that his lab does. I fooled around with HyperCard for about a week and wrote my first Macintosh program.

Mac Classic II

The first couple years of college, I was computerless. My junior year, I decided to splurge and I bought a Mac Classic II - 4 MB RAM, 80 MB HD, 16 MHz 68030 processor. That was 1992, back in the beginning days of the internet. It’s pretty surprising, but this was actually marketed as a business computer. I used that computer faithfully for 5 years. Even after buying a new computer, I kept my old Classic II cuz I just couldn’t bear to sell it (especially since I’d probably only get $20 for it). I finally donated it to Goodwill in June of 1999.

PowerCenter

In February 1997, I decided I needed a new computer. Apple was in the middle of their worst year financially - losing over $1 billion. And unfortunately, their line of computers was a reflection of their financial incompetence. Despite this, I still liked the Mac OS platform, so I investigated the Mac Clones which were much more technically sound. I bought a Power Computing PowerCenter 150 MHz PowerPC 604 with 48 MB RAM, 2 gig HD. So far, this has been an ideal computer and I see it being usable for at least another couple years. (and perhaps even longer, now that I’ve installed Linux on it!)

PowerBook G3 Series

I spent most of 1999 travelling around the country (see my Work page), so I decided I needed a laptop…ok, ok… I wanted a laptop. So, I donated my PowerCenter to my roommate, Gerry, and bought a used PowerBook on eBay. It’s a Wallstreet 2 - 266 MHz PowerPC G3, 192 MB, 4 gig HD, and a DVD. I bought this in July 1999 and it was everything I could have imagined. I’ve extracted all my CD’s into MP3 format and I bought Virtual Game Station, so I could play Playstation games. This computer is now a complete entertainment system - and it only weighs 8 lbs! Now that I have 192 megabytes of RAM, it’s hard to believe that our original ZX80 had only 1 kilobyte of RAM! And even my Classic II had a only a 80 megabyte hard drive! That’s less than half the amount of RAM I have now… The way things are going, this computer should last at least a few weeks!

eMac

Wow, it’s been a while since I updated this page. Around the time I wrote that last paragraph, I was beginning to get interested in Linux, so I bought a cheap AMD-K6 processor and put together a little Linux playground. After trying out a few other distros, I finally installed to run Debian GNU/Linux and it’s provided me a wide-open window into the world of Free Software (I’m associate member #1088). I’m not coming back :-).

Powerbook 12

My Wallstreet finally died in the fall of 2002 and I mercilessly sold it for the value of its parts. I bought an eMac because it appealed to my sense of “bang for the buck”, but I missed the portability of my laptop. So, I sold the eMac to Sung’s parents and bought the 12-inch Powerbook in the Spring of 2003. I miss the beautiful, expansive screen of the eMac, but I really love this new toy. The most lovable part of the Mac is the new operating system, OS X. It combines the stability and geekiness of UNIX with the beauty and don’t-make-me-think usability of the Mac.

Apple pictures and movie obtained from Apple-History website. ZX-80 picture obtained from Albert’s Virtual Computer Museum

I’ve been meaning to put this pic up for a while. That’s me in back, with my super-cool brother and his even-cooler Harley Davidson VRod. It was about 40 degrees F and despite sitting in back and despite having the warm helmet, I was still freezing. I can’t imagine what Manu was feeling. Halfway through the ride, I asked him, but his face was too frozen to talk. :-)

Got Mac OS X? Want a cool screensaver? I stole this from MarkD over at BorkWare and then modified it to allow multiple drips and configurable block sizes.

• Get the ScreenSaver (expand archive and put it in “~/Library/Screen Savers”).
• Get the source code.

My parents and I went to India for 2 weeks. We spent a couple days in Delhi and then the rest of the time in Kerala. I thoroughly enjoyed myself and actually found the vacation relaxing. Kerala is more beautiful than I ever remember (somehow I seem to forget in between trips). It’s lush and green and the food is amazing. I can’t wait to go back. Here are some pics. Unfortunately, my camera battery died while I was in Kottarakara and I was too dense to bring a 220V converter, so there aren’t nearly enough pictures from Kerala.

It’s snowing in NYC. Really hard.

I finally finished Gerry and Kim’s wedding site. The HTTP address is temporary. I had made a promise to Kim back in September that I would finish it, but that I needed a deadline. My brain seems to refuse to work without a deadline. Anyway, she said, “end of the year” and sure enough, it went live on 12/31/2002. It’s running on OpenACS 4 (CVS HEAD 12/25/2002), using an Open Source CSS stylesheet from http://w3.org. All comments welcome.

Developing this site using a comprehensive framework like OpenACS was certainly overkill. I started by using PHPNuke, phpwebsite and then MovableType. All of them would have been fine (as would my old favorite, Frontier). The thing is, none of these worked exactly as I want them out-of-the-box and I’d much rather customize OpenACS for my needs than wade through PHP or Perl. MovableType was the exception in this group as it worked exactly as advertised, but it’s advertised as weblog software, which is not what I was building. I also wanted to test out OpenACS 4 since I’ve spent so much of my free time working on it. Like Jeff said, you actually gotta use the software before you figure out where it needs fixin'.

If you’ll scroll down a bit, back to June 28th, 2001, you’ll see that I promised ‘Video’ of the wedding of Gerry and Kim. Go ahead, scroll… I’ll wait.

Back? OK… I actually said ‘Video at 11’, but I didn’t specify exactly when ‘11’ was. Well… it’s now!!!! OK, so it’s not exactly video, but it’s a slideshow of the festivities (4.6 Mb).

I wrote up a little intro piece on the use of open source software in medical applications (PDF). HTML below…

Improving the communication of patient information is the key to improving healthcare quality and efficiency. In most health care organizations, written records are the main conduit of this information, yet they are notoriously illegible and difficult to browse. Reviewing past problems and past interventions is time-consuming, if not impossible. The record is used mostly as a one-way device for the clinician to record data. It is used much less efficiently to provide important data or reminders back to the clinician. This leads to redundant testing and treatment, as well as to critical data being overlooked. Multiple providers and administrative departments need access to patient information, yet the paper record can only be in one place at a time. As a comprehensive storehouse of patient information, the paper record is obsolete.

The electronical medical record (EMR) is an important weapon in the fight against healthcare inefficiency. EMRs are software programs that allow the clinician to record patient information. They are usually tied in with other ancillary services such as pathology, laboratory, pharmacy and radiology to provide a comprehensive view of a patient’s condition. EMRs have several advantages over written records. Records can easily be searched for pertinent information leading to less duplication of testing and treatment. The record can be programmed to notify the clinician of drug interactions, abnormal lab results as well as routine preventive health recommendations for which the patient is eligible. This decreases medical errors and improves quality. Decreasing errors will, in turn, decrease malpractice costs. EMRs can recognize missed appointments and automatically send letters to the patient reminding them to reschedule. Computers can handle the routine, repetitive aspects of patient care that they do best, allowing the clinician to concentrate on the care of the patient. EMRs can be programmed to translate clinician diagnostic information and documentation into insurer-required diagnosis and billing codes, which makes billing and reporting more accurate and efficient. EMRs can include access to medical references, thus allowing the clinician to apply new knowledge immediately. EMRs have the potential to improve communication between patients, providers and consultants. All three groups can have access to the record at the same time. Properly done, clinicians in different geographic areas and using different computer systems can access and update a patient’s record. Finally, EMRs also promise to improve the security and confidentiality of patient information. Paper records can be inspected by any number of people while in transit. EMRs will only let individuals with proper credentials view patient information and different parts of the record can be guarded with different levels of security.

Most EMR software packages are proprietary. The source code is controlled by the vendor and there are restrictions on the use and distribution of the software. As an example, Microsoft Word is a proprietary program. The source code is not available. Users must purchase a copy of the program and use it on only one computer. They are not allowed to copy or redistribute the program and there is no way for them to make changes to the way the program runs.

Proprietary software is not a good solution for the healthcare industry. The software is expensive, often prohibitively so for small clinician groups or for those in developing countries. The software is complex and doesn’t meet the clinician’s needs. This is due to the fact that clinicians are often not involved in the EMR development process. While the software may be developed in conjunction with advice from a few clinicians, users in the field have little abilty to provide feedback or direction to developers. Proprietary software packages often address only one problem in the healthcare arena, such as recordkeeping, billing, transcription, scheduling or medical reference. Unfortunately, since each application is built by a different vendor, there is no communication between applications, thus reducing their overall benefit. There is no incentive for an EMR vendor to allow their application to communicate with another vendor’s billing application, for example, because it would reduce their profit potential if they were ever to develop a competing billing application. This leads to the problem of vendor lock-in. Once a healthcare organization implements an EMR, it becomes very difficult to make a change. The vendor can essentially hold the organization hostage to their pricing decision because the organization’s patient data is tied up in the current system. This especially becomes a problem when the organization requests that problems be fixed or features be added. They are at the mercy of the vendor who may be slow to fix problems which don’t affect their larger customers, or to introduce features which won’t be widely demanded. Unfortunately, in the healthcare space, one size does not fit all. If the vendor ever goes out of business or changes their focus, then the organization faces an expensive migration. In addition, many current EMRs run on Windows which has poor reliability and security records in comparison to Linux and UNIX. For all these reasons, using proprietary software is a poor solution for healthcare.

The alternative is Open Source Software (OSS). OSS is free. Free, in this definition, means liberty, not necessarily price. While OSS may have a price tag, the user always has complete liberty to do with the code as they wish, including redistributing it for a charge or no charge. Users have complete access to the source code which defines the behavior of the software. They have the liberty to change the code to suit their purposes. There are various open source licenses (GPL, BSD) which each have their own various stipulations.

The most visible example of OSS is Linux, an operating system which competes with Microsoft Windows. Apache, the most common web server in use, and Openoffice, a suite of applications similar to Microsoft Office, are 2 other successful examples. Anyone is free to download these software programs, change them and redistribute them, if desired. The strength of these programs come when users make changes to fix problems or add features and then return those changes back to the original developers.

David Wheeler has done an amazingly comprehensive analysis of OSS and concludes:

OSS/FS (Open Source Software / Free Software) has significant market share in many markets, is often the most reliable software, and in many cases has the best performance. OSS/FS scales, both in problem size and project size. OSS/FS software often has far better security, particularly when compared to Windows. Total cost of ownership for OSS/FS is often far less than proprietary software, particularly as the number of platforms increases. These statements are not merely opinions; these effects can be shown quantitatively, using a wide variety of measures. This doesn’t even consider other issues that are hard to measure, such as freedom from control by a single source, freedom from licensing management (with its accompanying risk of audit and litigation), and increased flexibility.

OSS also lowers the cost for vendors who support it. When proprietary vendors create a product, they start with an empty slate, often rebulding many pieces that have been built by other vendors. OSS vendors start with an established code base, saving time and money. Their final work then gets returned to the community, thus improving the base at which other vendors start. There is much less reinvention and correspondingly more room for innovation. Users are empowered with OSS. Since they have full access to the source code, they are able to fix problems that are important to them. Alternatively, they can hire outside developers to fix problems. There’s no need to rely on the original developers, so if the original developers ever leave the project or “go out of business”, the consumer can simply hire another team. The code is theirs. In 1999, Medicalogic introduced an internet-based medical record called LogicianInternet. Physicians took advantage of this system to keep an EMR which was housed on Medicalogic’s computers. When Medicalogic was sold to GE Medical Systems, GE decided to discontinue the program. All the physicians now have to print out their patient records and switch back to paper records, or re-enter all the patient information into a new EMR which they must purchase and install. If the application had been open source, the users could continue to use it, hiring new developers as needed to keep it updated. Arsdigita was a company which specialized in building community-centered websites and they released their code as open source. As the dot-com economy recessed, they were forced to change focus and were eventually bought out by another company. If they had not released their code, all the users of their system would have been forced to live with the program as it was, without updates or enhancements. Alternatively, they would have had to migrate to another system, which would be expensive. Instead, a community grew around the software, called OpenACS (http://www.openacs.org). This community has taken the software and improved it. Arsdigita’s customers benefit because, unlike Medicalogic’s customers, they can continue to use and customize their software. In addition, since the OpenACS community has flourished, former Arsdigita customers can find developers through OpenACS to fix problems and customize their software.

OSS is especially relevant in the medical arena. The medical community is known to value academic freedom. Researchers regularly publish their findings and if an intervention improves patient care, it makes sense that it should be spread as widely as possible. The same should be true of software which improves patient care. Proprietary software is also more risky in the healthcare arena since medical software companies are, as a rule, small outfits and are more prone to failure as a business. The reasons small companies abound in the current arena is because medical practices vary widely in their sizes, characteristics and needs and their problems can not be solved with one homogenous solution. Yet they still need to be able to communicate with other heterogenous systems and legacy systems need to be interfaced, not thrown away. OSS is an ideal method to develop conduits which translate between proprietary software programs, if not to replace them completely. In addition, OSS is a natural fit for international collaboration due to the lack of intellectual property issues and the need for customization of software for local needs. Healthcare organizations in other countries will feel more secure having control over the source code, rather than relinquishing such control to vendors who may be in other countries and subject to different regulations. Overall, OSS will lead to software that improves the clinician’s ability to function and improves patient care.

There are numerous vibrant OSS healthcare projects in existance and I encourage readers to follow the URLs to their websites for more information: (See attached list)

How can you get involved? It’s easy to make a difference in the Open Source world. All it takes is personal interest, diligence and a willingness to learn. Competence is rewarded. Try out the sotware. See how it works and how it breaks. Report difficulties that you have or features that you’d like to see. Be aware that developers are sometimes working on their own free time, so be courteous and patient. Take the time to read documentation and follow the discussions before jumping in. Offer to help out where you can. Keep notes about your experiences. These notes can help other new users and help developers improve the program. Non-programmers can help by submitting problem reports, writing documentation, and assisting developers with domain advice. Most importantly, promote open standards and open discussions among those around you. OSS software is the best approach to building software that will solve real healthcare problems.

References

• Bray B. Keynote speech at the Inaugaration of the Openkaart Initiative. 9/26/2001

• Bruegger BP, Dal Molin J. Open Source – Open Content – A Perfect Match for Sustainable Development. PowerPoint Slideshow

• Carnall D. Open Source Software in Healthcare

• Carnall D. Vital Signs LinuxUser. October 2001

• Carnall D. Healthy Outlook. LinuxUser. July/August 2000

• Carnall D. Medical software’s free future. BMJ 2000; 321:976.

• Cook B. Wipeout: Lessons on protecting Web-based EMR data. AMNews. 8/19/2002.

• Johnson DL. Open Source Medical Information Management.

• Johnson DL. Medical Enterprises and Open Source.

• O’Reilly T. Ten Myths about Open Source Software.

• Wheeler DA. Why Open Source Software / Free Software (OSS/FS)? Look at the numbers!

Links: General Medical Informatics:

• Handbook of Medical Informatics

• International Medical Informatics Association

• American Medical Informatics Association

Links: General Open Source:

• Open Source Initiative

• Free Software Foundation

• The Cathedral and the Bazaar

• Why Open Source Software?

Links: Open Source Medical Projects:

• An up-to-date list of projects is maintained here

• VISTA – Veteran Administrations comprehensive healthcare software

• GEHR (Good Electronic Health Record)

• OIO (Open Infrastructure for Outcomes)

• OpenEMed

Last Modified: 9/19/02