There was a hole in the wall of our bathroom that was a painful reminder of a bad encounter with a plumber.  Yes, that hole has been there about a year, and it has been on my to-d0 list for the duration, daring me to show if I inherited any of the fix-it genes I got from my father.  Why not hire someone to come fix it?  I also got (as I mentioned in my last post) dutch genes, which scream at me whenever I reach for my wallet.  So this hole was giving me shame in surround-sound.

I attempted to fix it the hole last year, even going to the degree of asking for  a router table for my birthday.  Since there was previously no way to get to this all-important access to the shower fixture without cutting through the sheetrock, I decided I would take a board, cut it larger than the hole, then use the router to make a rabbet cut so the panel would fit snuggly.  Up until then, I thought a rabbet cut was a surgery to keep the family pet population under control, but my vocabulary was suddenly expanded to include words like rabbet, roundover, chamfer, dado and round nose.  Unfortunately, my success only came in the realm of vocabulary, as I was not able to successfully master the rabbet cut without making the wood become a classic example of the early american gouge woodworking style.


I am not sure why, but something inside me told me today was the day to give this another shot, and to my shock (and that of my family), I was successful!


Yes, there still is a minor wallpaper issue, but note the total absence of gouges!


Also note this fine example of the roundover cut.

This home project is actually a late comer to the DIY party I\’ve been holding for the past few months.

  • Don\’t like your practice?  Build your own from scratch!
  • Don\’t like the health care system, build a new one!

My latest DIY venture is in an area I swore I\’d not go: I\’m building my own record system.

There are several reasons I\’ve avoided doing this DIY project:

  1. If I fail, I\’ve wasted a bunch of time I should have been building my practice.
  2. If I succeed, I don\’t just have a practice to manage, but a piece of software.
  3. I tend to get obsessed with details, losing hours coming up with elegant solutions to problems for which simple solutions are available.
  4. It requires that I spend far too much time thinking about HIPAA and security issues.  I hate that kind of thing.  It bores the socks off of me.  I fell asleep three times while writing this bullet point (and I have no socks).

Business is good; we are up to 250 patients and are managing the volume pretty well.  But I\’ve had to keep a cap on growth while I figured out what system I would use to run the practice. Obviously, EMR systems designed to produce enough E/M vomit to scare away Medicare auditors don\’t fit with my business plan. Other systems seem to have become so obsessed with \”meaningful use\” that they don\’t do basic business functions.  Expecting a system designed to work with the Economics Through the Looking Glass of American healthcare to function in the real world is folly, and so I had to choose: do I stay with my current non-system and let the quality of my care suffer, do I keep growth of the practice to a minimum, ignoring the reality of 3 kids in college next fall, or do I give in to the belief that I know what I need and can build a computer system that will work with my type of practice?

I decided on what\’s behind curtain number 3.  Unfortunately, this all happened just as I agreed to an interview with a local TV station – an interview that went viral and now has people as far away as Idaho and San Francisco wondering if they can be my patients.  Now the pressure is really on to make this thing work.  I can no longer be indecisive; I will either live by the database or die by it.

So far, it\’s been going well.  Despite a few \”unfortunate\” moments where I deleted all records of everything (thank goodness for paranoia about backups), I have broken the code of working with a relational database, and my nature as an internal medicine problem-solving nerd has served me well.  In truth, this is not much different from what I did with the EMR system at the old practice.

  • I think about where the greatest pain is for me and my nurse, and fix those problems.  Where can time be saved, and jobs be made simpler?
  • I think about where the greatest risks for patients are, and fix those problems.  What things are easily forgotten or missed?  How can I set the system up so it assures the safety of my patients?
  • I think about where I want to go with the practice in the long run, and set up a system that will set us up to go in those directions when we are ready.
  • I think about the questions I ask myself when dealing with a patient, the information I want to know the most, and put that information in a place where it\’s easily accessed.

In reality, the software borrows heavily from software real businesses use:

  • Contact Relations Management to keep track of interactions with customers (patients)
  • Business financial management to keep track of costs and of who has paid (and who hasn\’t)
  • Document management to handle the reams of information flung at me on a daily basis.
  • Task management to keep important tasks in front of me and my nurse (and eventually patients)
  • Spreadsheets to organize numbers
  • Reminders to tell when important things are due
  • Communications systems both between office staff and with patients

It\’s really a hybrid of all of these, with the additional plan to securely share much of the data with my patients online.  My hope is to build something good enough to get the interest of someone who actually knows what they are doing in writing software.  I know what problems need to be solved, and am learning much about how a good database program can do that (I am using Filemaker Pro because it\’s cheap, it\’s easy, and it works on both Macs and Windows), but I know my limitations.

I still have no desire to become a software tycoon.  I am doing this only because it\’s the only way I could see to make this practice work.  The practice is still at the center of my motivation.  If it doesn\’t help me serve my patients better, I won\’t do it.  The amazing thing is that we used it all of last week and my nurse didn\’t quit.  That\’s a good start, but the real test comes this week, as we take on the barrage of new patients brought on by our recent publicity.

I\’ll keep you posted.

8 thoughts on “DIY”

  1. Wow!

    You never stop impressing! You understand the concepts of relational databases, and you’re savvy
    enough to pull it together. I don’t think as much coding is needed as in
    the 90s but still, you have to roll your sleeves up and dip into some kind of
    code. #impressed 😉

  2. I’m impressed too – especially having worked with Filemaker. I’m also excited because you’re designing the type of EMR I need for my practice. Hope this is successful and you decide to offer it to others <–gentle hint

  3. If you aren’t already using it, you can vastly expand the power of what Filemaker can do by using the scripting functionality. It’s a strange little scripting language. But once you crack the code, it’s pretty powerful.
    The downside to Filemaker is it handles relational database design very differently from ‘normal’ databases where everything is done by SQL. Coming from a more traditional database world, it took me forever to get comfortable with how Filemaker does things. But, between the Mac, Windows and iPad support, it’s a pretty attractive platform.

    I’m a semi-retired software programmer/entreprenuer & I’d be very interested to see what you have when you get a little bit down the road. I sit on a couple of hospital boards and am convinced your model of practice represents the direction we should be moving with our primary care.

    BTW, here’s what you can look forward to a few yeas down the road via Google Glass:

  4. Yep, I’ve been having fun with the scripting language. It’s really quite fun. I actually cut my database teeth using Crystal Reports to query both Oracle and then SQL. I’m glad, however, that I’ve not had to re-learn things having not really understood relationships until now. It makes intuitive sense to me now, and as an internist problem-solving nerd, this is right in my wheelhouse.
    I just sat down with a software guy who has developed a lot of projects like this and I am happy to say he was impressed with my progress, given my lack of formal training.

    I thought the Google glass thing was cool, although seeing things like that in pre-alpha makes you have to use a good bit of imagination to see what’s possible in the future. Pretty cool, though.

  5. I am definitely planning on sharing the platform I am using, not trying to capitalize on it. I have no desire to own a product I have to sell. If I can get it to work in real life (which I now believe I will), I will be hopefully not too ashamed of any flaws that are still there. Right now it is REALLY early, but it is already far, far better than anything I used from regular EMR systems.

  6. Michael Chen, MD

    Have you looked into my system…NOSH ChartingSystem. It’s not based on Filemaker but uses industry standards like MySQL, PHP, Apache, and JavaScript It is platform agnostic. It was a DIY system for my own practice very much like yours. It’s open source so if you want to tinker with it, you can. It does all the functions you listed at the end of the post for all the same reasons I needed an EMR that worked for me as a solo doc. Why re-invent the wheel?

Leave a Reply