Develop Your Own EMR – Are You Crazy?

UPDATE: Check out the update to this post where I say you’re still crazy to develop your own EHR.

I had someone email me asking me what I thought about a small group of doctors developing their own EMR. Then, I was helping a doctor implement an EMR in a new office and his friend asked me why I didn’t just create my own EMR and make a ton of cash. All I have to say is….Are you CRAZY????

While developing your own EMR is a very nice thing because you are able to customize the process exactly the way you like I honestly think this is a big mistake. You will honestly be spending just as much money developing your own EMR as you will spend purchasing from an EMR vendor. Doing this for a small group of doctors is even worse. Gaining a consensus of these doctors on what is “best” for an EMR is like asking which type of ice cream is best. There are certainly some that are better than others, but it really just depends on a person’s preference. EMR choice is pretty much the same.

One of the biggest problems of creating your own EMR from scratch is what happens once your EMR is “created”. You have to continue paying development to continually create enhancements. If you stop enhancing your EMR program then you get farther and farther behind and lose some of the advantages of an EMR. If you go with an EMR vendor then you continue to benefit from the enhancements that they continue to create. Often there is a small update fee, but much less than you developing all these features yourself over time. An EMR vendor is able to generate a lot more revenue which can be funneled into enhancing your EMR. Do you really think that one programmer can compete with a whole team of programmers? Besides the time factor it is really hard to find a programmer that can do all of the specialized programming for an EMR. They would have to know how to do an HL7 interface, learn CCR, document management, reporting, just to name a few.

Even more important is it is just not realistic for one programmer to be able to build an EMR that has all the feature set of a good EMR system. Sure, I have the technical skill to program an entire EMR. It would just take time. Not a little bit of time, but hordes of time. Individual components of an EMR system really aren’t that complicated. When you pile them all together it would just require a lot of work to develop an EMR from square one. I really think developing your own EMR is a poor strategic decision.

Another point is that there are some great EMR companies that are reasonably priced and will certainly pay for themselves over time. There are even some different pricing models that don’t require an enormous up front fee which allows a doctor to minimize the risk of EMR implementation.

Here’s a few examples you might look at:
Doctors Partner No up front, just a monthly fee
Medtuity Charged per visit(does not include a PMS)
I know both of these companies personally and they are quality organizations and people with a unique cost structure.

There are so many EMR vendors to choose from. In fact, look at my enormous list of EMR systems.

Every doctor should be able to find an existing EMR system that meets their needs(pricing and features).

One other quick tip is to choose an established, but smaller EMR. These smaller EMR’s are able to listen and implement your specific requests much more quickly than a larger EMR that receives tons of requests.

About the author

John Lynn

John Lynn

John Lynn is the Founder of the, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference,, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • Taking on the task of developing such a system is a HUGE task but as more and more technologies emerge to enable more agile development methods such as Flex, Rails, etc… the actual “programming” involved become much less daunting and time consuming. I guess the key would be to make a system which would enable doctors to concentrate on what they need and not all the tech know how which comes along with such an undertaking. Even then you still have a large amount of work. What do you think about maybe open source communities building and supporting such systems?

    Interesting post.


    Nick Selvaggio

  • I was going to say the same. Instead of going the Micro$oft way and trying to engage in the ‘profits are everything’ world, why not use an existing CMS like Plone or phpNuke and spend time customizing portlets or modules? The only trade-off would obviously be that the physicians would have to rely on the community of programmers for updates and bug fixes instead being able to piont the finger at sue-able company.

  • And who would support it? You need the capabilities of a support organisation which is only likely to be cost effective over many customers’ support fees. Do not have the developers doing support!

    Also, you need a vibrant community of users to do post-Beta debugging and achieve anything like tolerable stability.

    Configure what is off the shelf, but bespoke is dangerous territory.

  • Nick,
    My take on open source communities is that the open source EMR communities haven’t quite made it yet. They are making a lot of progress and there are a number of people committing a lot of time to the process, but they just aren’t where they need to be in my opinion. I expect that within the next 2 years they can be considered a viable option.

    The other major problem with open source EMR projects is that there isn’t enough of a community to support the project at this point. If you needed help I don’t think there are enough open source EMR developers to help you out. I guess you could dip into the open source development community as a whole, but if you do that your essentially developing your own EMR. A good open source community shares a lot of the development. There’s just not enough people to share with yet. I hope this changes.

    I don’t know too much about flex and rails, but from the programming languages I’ve seen we are still a long way from separating the technical from the development process.

  • Fredric,
    Are you serious when you suggest that someone use Plone or phpNuke for an EMR system? They are good products that have their place, but being used as an EMR system would be a huge mistake. Talk about opening yourself up to vulnerabilities. There’s no way I’d trust the security of those 2 systems to keep my patient information private. There’s just too much risk. They weren’t designed with privacy in mind.

    Privacy aside I still think it’s a bad option since it would take so much programming to make it usable as an EMR.

    The idea of having Microsft as a sue-able company is a very fine point. Sure is a nice security blanket. I’m just not sure I know many lawyers that think they have a chance against Microsoft’s lawyer department.

  • Amen to beta testing! Developing your own EMR means you’re the beta test site and will forever be the beta test site. At least you have the developer’s attention if there is a problem, but just plan on problems. It’s a part of the EMR development process.

  • In addition to the functionality of an EMR for a particular site there is the issue of interoperability. Eventually, to reap the real benefits of these solutions we need full communication and accessibility between systems, across the nation. That would be my greatest concern about implementing anything less than the best quality application (within obvious budgetary restraints).

  • Hi Chris!

    A long time ago I came to the conclusion that writing my own was the way to go, and I’ve never looked back. These are some of the reasons why:

    1) There are inexpensive software programming platforms that are made to help you in your endeavors using programming languages that are easy to work with. In the past I started out with Paradox for DOS then ported everything to MS Office using Visual Basic for Applications.
    2) Building your own EMR is amazing and entertaining. If I want a particular feature, I’ll add it in within 24 hours. Sometimes I bring up an expensive EMR and reverse engineer the feature into my system.
    3) Over the lifetime of practicing medicine I will have the advantage of not having to go through the incessant upgrade process which over time will cost hundreds of thousands of dollars.

    You guys are just too negative. It’s not that difficult…


  • I agree with Dr.Borges. It is really not that hard to have a functional EMR. I am a neonatologist and am forced to use an extremely user unfriendly system at the University hospital. However, I designed a whole EMR for my wife’s Allergy practice using microsoft accesss without writing a single line of VBA code. As Dr. Borges said, any update, new report or form can be created within a few hours. With a little help from a programmer, I am planning to add workflow management and integrated document management. Again, I don’t think it is going to take more than a couple of days to accomplish. The advantage of such a program is that it is clean, simple and without the clutter of tons of screens and buttons.

  • The interesting thing to note from the physician comments above is that they think different than us techies. Number of features is a big deal for us, but simplicity and ease of use is bigger for them. They want the features but not to the point where usability is compromised. Also, today’s EMRs are just giant ERP systems. It is all about business (scheduling, billing) and very little about clinical information. Even the clinical information that is collected is all analog (free text) and not discreet which makes analysis impossible. Ask any doctor anywhere today to give you the number of diabetes patients with left foot neuropathy that they saw in the last 6 months and they cannot tell you.

  • Well, there’s no doubt that almost all techies think different than doctors. That probably explains a lot of the reason why so many of the EHR are so unusable.

    I agree with you on the comparison of an EMR to the giant ERP systems. They both suffer from many of the same problems. Billing definitely has driven the development of EMR’s to date.

    Discreet elements is an interesting topic. The EMR that I think does discreet data like this the best is Medtuity. They even save each piece of discreet data individually so that multiple people can access the same chart at the same time. Ask most existing EMR vendors to do that and you can just start to laugh as you see them try.

  • While I agree that EMR is going to be important for lowering costs and improving patient outcomes the costs are largely do to the fact that we do not have a standard for data storage. Standards lower costs and improve quality. Without them only big players can innovate. When we look open source programs like Firefox or Thunderbird they work because of standards for the Internet.

    I hope to create a group that will present a standard to the ISO. Having one will allow for software vendors to create interfaces that can send information to different systems.

  • Kino,
    Have you thought much about how you might structure the standard? Would you build it on top of other standards? How would you get adoption of the standard you create?

  • You should check out

    This is a free full service EMR that my clinic- 12 providers, two sites- have been using for 3 years. It was created by one programmer. He is now giving it away for free because the development costs are very negligible.

  • Rakesh,
    Looks like is following the ad based model that Practice Fusion has as well. Basically it looks like MD Bug shows ads in the EMR. Is that correct?

  • Thanks Rakesh. Although, don’t you think it’s a bit disingenuous to say that they are offering it for free since development costs are negligible? Sounds like MD Bug is trying to make a buck like everyone else. I’m not faulting them for this. I do the same on this blog also, but you made it sound like the programmer just does this for the good of all.

  • Everyone wants to make money. It’s all about how much the doctors are going to pay for it. I think that doctors are skeptical of free EMRs- you get what you pay for mentality.

    My point is that one programmer CAN do an all inclusive EMR. It’s not that complicated as all the other EMRs make it out to believe. So, doctors should not spend thousands of dollars on cost and maintenance.

    Let them all make money from advertising- who cares- but let’s not be stupid to throw our own money away when we don’t need to- just because an EMR company has high costs. ( I paid about 100K in software and hardware costs 6 years ago and was paying 15K a year for maintanence)

    I think MDbug’s main objective is to change that mentality and force other EMR companies to become more efficient- and make tons of money along the way.


  • Rakesh,
    You’re right about many doctors being skeptical about free EMRs that pay for it with ads.

    In development, 1 great programmer is definitely better than 10 average programmers.

    I agree with you completely that people are paying hundreds of thousands of dollars for EHR software that is more or less unusable. They should keep looking. There are a LOT of options now with every pricing plan under the sun.

    Thanks for joining in on the conversation. This is an important message for doctors.

  • I wonder where all these “usable” EMR’s are. I’ve looked at many, and they are terrible at capturing the narrative of medicine–for us, the narrative is everything. Patients just don’t fit templates or check boxes. In addition, currently available EMRs are cluttered, ugly, and terribly disruptive to work flow. Come on, guys–don’t you work in hospitals? Why can’t IT folks wrap their heads around our work flow and needs?

  • House Whisperer,

    I challenge you to tell me what you an EMR to do. You say cluttered, ugly, disruptive and would like to work for you in the hospital and clinic. The problem is not technology, it’s the doctors.

    Being a doctor myself, the problem is: we don’t really know what we want. Seriously, If you can tell me exactly what you want and how you want it, I WILL show you an EMR that does that.

    The other issue is that doctors are only concerned about one thing: themselves (me included). The EMRs are only concered about one thing: the doctor (who buys the EMR). I think that all of that will change. Doctors and EMR companies will have to be concerned about one thing only: Patient Care.


  • House Whisperer,
    I’ll admit that I’m not much of a hospital EMR guy. Maybe there aren’t any good ones in a hospital. Although, developing an EMR for a hospital from the ground up is even worse than one for a small clinic.

    Certainly many of the EMR software out there are because IT people don’t have a clue about clinical environments. However, as Rakesh says, many IT people have tried to reach across the clinical lines and can’t get the clinical people to define what they want. Meeting that balance is incredibly difficult.

    Back to the original part of this post, start to list the detailed features that you want in an EMR and you’ll get a better idea of the challenge of developing one from scratch.

  • Folks,
    Historically, in development, and in technology as a whole, the formation of a standard that everyone agreed upon has helped the advancement of a particular technology. This is what ANSI does. I think if we doctors are going to develop EMR, we need to develop something that conforms to an ANSI standard or is good enough that it will merit an ANSI standard. I believe that interoperability is a major concern. We could make the emr as simple and uncluttered as we want but, at basic level its nuts and bolts (messaging, document architecture etc.) needs to conform to an ANSI standard. Learning something like HL-7 has a steep learning curve. I am sure we can do it, but it will take some time. Then there is the issue of certifying the emr you create with claims clearing houses, pharmacy clearing houses (so they will communicate with your emr). A lot of time involved here. However, there is something to be said for a co-op model of social networks. A pool of committed physicians with programming knowledge could split the task into smaller assailable chunks. What do you think of that?

  • Physicians can do all they want, but until the EMR vendors choose to follow a standard, we’re going to get no where. Plus, there’s as many political and legal issues as the lack of an established standard.

  • I think this article is written in so much negative sense. Building your own EMR with bare minimum features will satisfy small physicians more than purchasing a giant costly EMR from vendors. Programming help is way cheaper and easily available on internet and doctors can get it done from their friend circle if they have programmer friend.
    Building according to their needs is what is needed, rather than software full of crappy screens and buttons all over the screens ! Not all features are needed by most doctors.

    I did some search on open source projects, but couldn’t find any good looking EMR, apart from OpenEMR and OpenMRS and few others. I am a software developer with experience in EMR software development and willing to contribute to any such projects.

    Looking forward to few other nice projects coming up in this field and hoping I also will be able to contribute in some way to this cause.

  • Jayesh,
    Your comments illustrate why this post is important. No doubt this post is a little old (from 2006). So, I’ll see about doing an update as I’ve learned alot since I first posted this. However, you seem to illustrate the lack of understanding of the volume of functions that are needed in an EHR. It’s not that any one component is that challenging, but all together it’s overwhelming. Plus, if you make it commercial and want it to work for anyone but yourself, then you’re looking at another layer of complexity.

    I applaud the brave souls that do it, but it’s an amazing challenge to do it well.

    OpenEMR and Vista are the top open source EMR projects. Definitely the top 2 to look at if you want to look at open source EMR software. I’m sure they’d love your contribution. I’d love it as well since I’d love for their to be a killer open source EHR option.

Click here to post a comment