DB2, AS400, ISeries, and homemade EMR

[d11aad38d]

Our CIO loves this platform for its stability, security, and scalability. He is now bent on using the "80% of what you need" that he says is on the 400, even tho' nobody can list what that 80% is.
As PM for an Enterprise-wide Web-based eHEALTH Vision Project, I am faced with a CIO who is protecting an ancient fiefdom and a CMO developing a new Vision for our group. Oh, BTW, they don't talk except to chomp on each other in Exec. Comm meetings.
I can find no, nada, none, nothing tieing EMR's to a DB2. These have been Googled in many different combinations.
Can anyone give me a direction to find anything about DB2 and EMR's for the ambulatory setting? :confused:
Does anyone know of anyone using an AS400 for an ambulatory EMR at this time? If used during another time, why was the change made? :rolleyes:
I can use some help!!

[Kursk]

Well as our IT people always say, "Sure, we can do that, what's your budget?" and "We can do anything with unlimited time and resources". Budget is king to these people, and Pareto's Principle (http://www.public.asu.edu/~dmuthua/pareto's_principle.html) has led many an IT project astray - trying to find that magic 20%.

Most implementation guides will tell you that EMRs are not typical IS projects, that they are certainly not IT projects and should be led by physicians with business requirements determined by physicians.

Show the CIO an article or two on the Cedar Sinai CPOE disaster (http://www.optimizemag.com/article/showArticle.jhtml?articleId=19505645&pgno=5):

"No matter what potential benefits a new system may bring to an enterprise—even saving lives—if the intended users aren't sold on it, you may as well pull the plug. That's what Cedars-Sinai Medical Center in Los Angeles had to do last year when doctors rebelled against using a computerized patient-care system designed to eliminate medical errors."

So the cost/benefit ratio can reach infinity (Cedar's won't confirm it but it is widely reported that the system cost over 30 million dollars).

So "growing your own" enterprise system may be a bit risky. Better buy-in some expertise from First Consulting (http://www.fcg.com/healthcare/ps-cpoe-research-and-publications.asp) or some others to help you in your battle with the entrenched CIO. There is a lot at risk!

[mel]

AS400? :confused: Is that the old UNIX based system?

[ozzie]

AS400 is old read VERY old IBM mid range iron .. Ok I started playing with these in my early IT career..

I could write a book on why I WOULD NOT touch the old legacy AS400 system with a barge pole.. great iron and reliable ..
BUT not new world technolgy ..
AS400 and the web READ huge mess and lots of pain..
Since the code for AS400 was written much has changed and its just old technology.
What it would cost to replace just the hardware would make me shudder..
Then some brainiac will come along with idea of web "izing " the as400 read install an application that makes the code look like its web but not really web . Think windows Xp emulation looks if you will .
Then you discover all the down sides like how ugly it looks under a vpn and unusable and half the laptops the screen resolution is out of whack. of course as soon as the security folks get wind of having external connections to the AS400 out goes that idea..
Years back I worked with a major bank and we spat out mortgage rates every day thru an AS400 to a bunch of about 800 odd brokers.. To do that simple task was a major undertaking via the web.. In the end the AS400 was canned and we built a couple of servers and saved a bunch..
snip
Our CIO loves this platform for its stability, security, and scalability.
snip
Well the B52 was great aircraft too but by todays standards meets none of the above ..

You will have a ball with websphere NOT

[mel]

We are talking about the same program then, I honestly didn't think anyone still used it :p

[Kursk]

We are talking about the same program then, I honestly didn't think anyone still used it :p

No it is iron - hardware. But I think it can run on AIX/UNIX/LINUX but I defer to oz on this point.

[ozzie]

Smaller banks still use the systems for number crunching and flat table type stuff they are good. But money stuff is trivial from a code point of view..
I see it like the old office where you saw a wall of great old oak cabinets full of 4 x 3 cards. Great to repurpose the cabinets and yes all the data on the cards is valuable but the "oak card" system has no place trying to be intergrated into todays modern emr systems.
None of which makes the cards and oak drawers bad things however to leverage that concpet for the future is the wrong road to travel..

oz

[alborg]

>>> "Pareto's Principle "

Kursk- that one's a keeper! :cool:

>>> So "growing your own" enterprise system may be a bit risky. Better buy-in some expertise from First Consulting or some others to help you in your battle with the entrenched CIO. There is a lot at risk!

I'm a "grow you own" fan, but here they are really not growing their own, but incorporating someone else's home-built system, which I agree can be trecherous. The bugs, the other issues s.a. when contracts are broken and developmental costs soar make this type of setup an intense undertaking. A good friend of mine tried that sort of setup trying to get together a software that would manage intrabank communications in Panama. Twice he got ripped with escalating programming costs and broken promises.

Instead of a situation like that, I would try to negotiate a good deal, hopefully purchasing an inexpensive good functional software (not the most expense perceived "best"), then hire an IT guy to help you manage it in-house, either full-time or part-time depending on the size of the enterprise.

If the system purchased is partially "open source" or at least allows the ability to change various forms/templates then the IT guy can use the platform to develop some in-house changes to suite the needs of the institution.

>>> DB2

Now that's anncient! I remember the first database that got me hooked into programming was the DB3 Plus platform- a DOS based software that was available back in 1987. I got a stolen privately programmed airport management software written in DB3 Plus and I figured if this complicated system could be written in an off-the-shelf system, then I could program my own EMR. I never looked back... :)

Regards,
Al

[mpike]

>>> "Pareto's Principle "

Kursk- that one's a keeper! :cool:

>>> So "growing your own" enterprise system may be a bit risky. Better buy-in some expertise from First Consulting or some others to help you in your battle with the entrenched CIO. There is a lot at risk!

I'm a "grow you own" fan, but here they are really not growing their own, but incorporating someone else's home-built system, which I agree can be trecherous. The bugs, the other issues s.a. when contracts are broken and developmental costs soar make this type of setup an intense undertaking. A good friend of mine tried that sort of setup trying to get together a software that would manage intrabank communications in Panama. Twice he got ripped with escalating programming costs and broken promises.

Instead of a situation like that, I would try to negotiate a good deal, hopefully purchasing an inexpensive good functional software (not the most expense perceived "best"), then hire an IT guy to help you manage it in-house, either full-time or part-time depending on the size of the enterprise.

If the system purchased is partially "open source" or at least allows the ability to change various forms/templates then the IT guy can use the platform to develop some in-house changes to suite the needs of the institution.

>>> DB2

Now that's anncient! I remember the first database that got me hooked into programming was the DB3 Plus platform- a DOS based software that was available back in 1987. I got a stolen privately programmed airport management software written in DB3 Plus and I figured if this complicated system could be written in an off-the-shelf system, then I could program my own EMR. I never looked back... :)

Regards,
Al

Al:

You're thinking of dBase. dBase II was put out by Ashton Tate, and dBase III came out later, but was bought by a different company I believe.

DB2 is IBM's mainframe platform.

Ahh, the good ol' days. When we had choices.

The days are coming when we have a choice between Walmart and Walmart for shopping. :(

Mike

[d11aad38d]

Sooooo, thanks to all for repeating most of what we already know!! Rather intense, a known emotion here too. Now back to the original question, Does anyone know of anyone using an AS400 (hardware) for an ambulatory EMR?
DB2 is an IBM relational database.

[Kursk]

The answer is obvious.
Sorry for wasting your time with frivilous dialogue, but it is what we do best.
Good luck on your project.
K

[mbenjam]

Why not just use the AS400 and DB2 as a database backend, and design the frontend using newer technology? Every major programming language has APIs to interface with DB2 these days (well, of the ones I've studied, anyway) (ok, so Perl can do it for sure, but so can Java, right? That's Websphere, right?). DB2 in my mind is interchangeable with other DBMSs, as long as your frontend program knows how to interact with it. You'd probably be better off on a cluster install of MySQL using $5,000 of hardware, but that's none of my business. That's the beauty of the DBD/DBI driver system most languages use these days. The client programs can run on $300 desktops.

Is it possible that this is a very old version of DB2 that won't interface with the new programming languages?

So anyway, tell the CIO that you'd be delighted to use his rusting iron as the data server, and tell the CMO that you'd be delighted to spearhead a new frontend project written in VisualBasic or what have you, and everyone's happy.

Using Visual Basic ADO to interact with DB2:
www-106.ibm.com/developerworks/ db2/library/techarticle/0302hasin/0302hasin.html

I'm sure your IT guys can take it from there, though I'd be happy to expand on this as a consultant... :D

[mbenjam]

If you could figure out how to port the database from MSSQL to DB2. That would probably take a guy like Ozzie an afternoon...

[ozzie]

http://www.saigon.com/~anson/ebud/ebidx.htm go here the CIO go here a read a Sutta or two on clinging.. If that fails give everyone leeches and hollow wooden tubes. Its not a great idea in my mind , to connect a very old technology to a very new technology. For many reasons..
Most of all the finger pointing will kill the process, and cause much agro and little result.
Even in IBM world things do die , and death is real painful when there is much attachment..