TECH: Quick KP HealthConnect update

Of course the other big non-HIMSS tech news is that KP dumped interim CIO Phil Bruce Tukstra and brought in another outsider. While the LA Times says that Tukstra ran Health Connect, I’m not so sure. Andy Weisenthal told me that he (on the TPMG side) and Louise Lang (on the plan side) ran it, and expressly said that Dodd (the CIO who went a few months back) had not much to do with it. And I’m a little cynical about the LA Times reporting on this last week too.

However, I have extracted a promise from a senior TMPG regional CMIO to go on the record about what he’s seeing—so hopefully I’ll have more details for you all soon but suffice it to say that not everyone was too happy about the Citrix approach

Meanwhile there was an interesting session I sat in on at HIMSS Monday from the CIO of Kaiser Hawaii about how to deal with down-time when you have an EMR. What he said was that every hour a mirror of the latest data is made and that it’s on a separate system that can be accessed if the main HealthConnect goes down. And then he said that most doctors would dream of having that level of data (i.e. what’s in the backup) and he’s probably right! But I guess the real question is was the low uptime (the Times says 80% only in some cases) a function of the power failures last year, or is it still going on? My understanding is that it’s the former.

Finally, I’m going to be on a panel in Vegas on April 30th with two one leading Kaiser basher”* (Justen Deal and Gadly) and apparently Kaiser has declined an invite <sigh>. Which I guess means that I’m going to be their tame blogger and be castigated for defending HealthConnect….

* term adopted from “Japan basher

UPDATE: Justen emailed me to point out that not only had I called Bruce, Phil (hey it was 6 am!) and that he was not going to be on the Vegas panel (should have checked, his name was on an earlier agenda I saw but he never agreed) but also this:

Bruce was, without question, the "architect" of HealthConnect.  Before he became
interim CIO, he was the VP specifically for HealthConnect.  Dr. Liang has always
handled making sure HealthConnect improves the "quality" of our hospitals, while
Andy Wiesenthal’s job is to work with the physicians to make sure the system
meets their needs.  Turkstra was in charge of designing and implementing the
system, while Dodd was in charge of making sure the infrastructure was in place
to support it. Finally, I hope you’ll consider not calling me a "basher"
("Japanese" or American) of Kaiser Permanente. Trying to paint someone as a
"basher," frankly, only makes you look like a "cheerleader." Since neither, as
far as I know, is true, I hope you’d consider not using it in the future.

I don’t know about who did within HealthConnect, so hopefully someone who does will tell me if Justen’s right or not.

As for the "basher" term, this is a little inside baseball, but the "Japan bashers" of the late 1980s actually felt that they were very pro-Japanese but that Japanese politics and society needed to become more open. So I think the reference to Justen as a Kaiser-basher is extremely correct in that context.

Categories: Uncategorized

Tagged as: ,

5 replies »

  1. The american people are blinded, caught up in a whirlwind. we need to step back take a deep breath. Smell the roses, do whatever you have to do. Everyone is fighting.The democrat against the republican and the republican against the democrat. To far right to far left. To this to that. Get ahold of yourself. No matter who the Predident of the United States will be some will like him some won’t. United we stand, Divided we fall. I sometimes wonder what our neighboring countries are thinking of us. They must be laughing at us and our confusion. Our fighting and our bickering. We are feeding into there hands. You don’t dare mention the word God anymore. No wonder our country is the way it is . Let the chips fall as they may. I have now had my say. We all have our own opinion. May God bless AMERICA the land of the free and the home of the brave…

  2. Let me answer this one ..
    “Matthew, I realize this may be a very basic question, but what are the disadvantages of the Citrix approach?”
    Citrix is a system that proxies the client interface to an application. It is comprised of two parts:
    1. A client portion. (I think) this is a java applet that is invoked from a Web browser.
    2. A server portion. That runs several copies of the EpicSystems, Windows thick client on behalf of several corresponding users.
    The Citrix client talks to the Citrix server using a proprietary protocol, transferring a graphical image of the screen. This is a bit like how VNC, or Timbuktu or X-Windows works. Presumably the Citrix software does something to optimize the potentially huge amounts of image data that are sent over the network between the client and server.
    This potentially massive amount of data traffic is the first disadvantage – it imposes a significant load on the network by transferring large amounts of graphical data.
    If there are any advantages to this “architecture” they are:
    1. A user can run the application on any workstation which has a browser, java interpreter and network connection.
    2. The client workstation can be low powered – it’s only really a terminal, acting a bit like a vt-100 or a 3270.
    3. A legacy client-server application, like Epic, can be made to appear to its users as a desktop application running on a browser.
    4. Client workstation configuration and provisioning is confined to the Citrix servers, obviating having to configure personal workstations. This might be a mute point since most large companies use desktop administration software to manage their desktop configurations.
    The other disadvantages are:
    1. The machine(s) running the Citrix server(s) and client Epic application have to be very large Windows machines to be able to handle the work of several users. Large scale deployments require significant numbers of these severs.
    2. Most companies buy decent sized desktop machines anyway which can handle the client processing so they are incurring additional cost by having to buy two machines to do the job that one could do.
    3. If a server goes down it takes several users with it.
    4. There is huge complexity in managing the configuration, which must affect availability.
    5. A load balancer has to sit in front of the citrix servers to route client connections to servers – more complexity.
    6. The servers have to be Intel windows machines.
    7. It’s costly since every software and hardware component is vendor supplied and proprietary.
    8. There is an additional logon step to authenticate with the Citrix server.

  3. Matthew, I realize this may be a very basic question, but what are the disadvatages of the Citrix approach?

  4. So you were actually calling me a KP cheerleader? I can’t keep up, Matthew! In any event, I’m glad we’ve decoupled your Fasano-Turkstra 6 a.m. combination and gotten the job responsibilities straight. Looking forward to your panel in Vegas…