Because I can tell you, from 2 decades of experience, you are dead wrong. EPIC has been at this since 1979, MEDITECH since 1969. This industry changes dramatically on the whim of Congress and the data stored and processed is extremely complex and inter-related. Most HIT vendors have been through at least 2 if not 3 or more full iterations of their software, and lets not forget that many if not most facilities/health systems pay much less than $270 million for their HIT systems. Epic tends to be on the high side of the price spectrum, and will not even respond to quotes for hospitals or health systems that are under 500 beds (I can’t speak for practices).

Let’s take a look back into the DIY HIS annuls, and pull up this gem: http://www.modernphysician.com/article/20030201/MODERNPHYSICIAN/302010715 which is a Modern Physician article describing the development of Patient Care eXpert (PCX)by Perot Systems for Cedars-Sinai hospital in Los Angeles.

Also note this article from the LA Times about Cedars halting that implementation after a revolt from their medical staff. http://articles.latimes.com/2003/jan/22/local/me-cedars22

That is not to say that all in-house developed programs fail. Octo Barnett, father of the oft-maligned in this industry MUMPS (or M), has been doing it for close to 45 years now, quite successfully. But he has a large staff (60+, including 14 MD’s/PhD’s), large facilities backing him, and the benefit of being a Harvard professor, who has 45 years of experience doing this sort of thing. Most critical access and community hospitals in this country have none of these things, and quite honestly cannot and should not attempt to develop a HIS system from scratch.

The other thing I think you fail to grasp is the scope of such an endeavor: It’s really not like developing one application, more like developing 30 large applications. There are so many aspects to this business and the information systems that serve it that I think it’s naiive to think that this can all be done in one monolithic application. Even in the article you reference, it is not clear what it is he’s actually developing; it sounds more like he will be creating a system to connect the disparate systems they already have in place.

Lesson learned: Just because you can develop it, does not mean you should. Hospitals do not deal in widgets, and errors in medical systems can mean serious injury or death for patients. Even a well-planned out system, with professional developers and physician input can catastrophically fail.

I beseech you SameOldStory, if you do not understand this business, please do not go around the web bitching about the amount a hospital or health system spends on the licensing and implementation (usually much more $$ than the license fee, and includes things such as highly skilled staff’s time for training and getting their input in the design/setup of the system) of a mature, integrated and proven product in an industry that really cannot tolerate anything less.

Tell me, where would you rather your mother (or wife/child, etc.) be treated? In the place that over-spent for a HIS that is proven, tested, and provides accurate clinical decision support, or the place that is still “working out the bugs” from its home-grown system?

It was mostly internal costs and over 10 Years.

That’s 27 million budget per year for an organization the size of Sentara.

If SOS has a 10 year financial model, I’d bet Mr HISTalk would be happy to share it and I’d bet Bert R. from Sentara would be happy to comment on it.

You could get any number of pretty darn good companies to comment on their total development costs and R&D investments.

- SOS – if you’ve got a plan, just detail it out and let others comment on it.

The cost is astronomical because hospitals have no leverage when negotiating with large vendors. In any other industry, once the 270 million quote arrives the vendor is sent packing and the software is developed in house. There is also an overwhelming assumption in healthcare that healthcare software is more complicated than other industries and development is rocket science. It is not.

Interesting reading from a previous HIStalk post regarding a Hong Kong CIO. Large or small hospital, this would never happen in the U.S.

“This Hong Kong CIO who has worked in US healthcare IT takes a dim view of bloated hospital IT departments that buy big packages that take lots of people to install and manage. “Are they really doing that much? I would say here in Hong Kong we do much more per dollar spent. One of the Adventist hospitals in the US has more than 100 employees in IT and more than 30 of them are implementing and maintaining Cerner.” He also says that HIT systems are relatively easy to develop, vendors don’t know how to turn one implementation into an off-the-shelf package, and systems aren’t designed with user customization in mind.”

http://www.futuregov.net/articles/2009/may/14/diy-hospital-cio/

You’re twisting the numbers and you probably know it, so I write tihs for those who may not realize what you’re doing.