Re: [openhealth] Re: What do you keep explaining about Health IT
On Sat, Sep 4, 2010 at 11:14 PM, Ian Martin wrote: > Hi, > I'm an Emergency Physician, and program (badly) for the fun of it. > > First of all, there is a paradigm shift between medicine and IT. Medicine > has a > veneer of science, but once you dig a little, there's a truckload of touchy > feely stuff that the science depends on, and that makes it very hard to > tick > boxes on a screen. This ranges from the really complex individual with > multisystem problems who really went to ED because their granddaughter went > away > for the weekend, and they're not mobile enough to manage (unable to cope is > not > an allowed choice) to the elderly person who trips over because their > eyesight > is bad and breaks a hip. The hip is the reason for admitting them; the > underlying problem will almost by definition be secondary (if at all > recorded), > but fixing that is the real long term solution to the patient's problem. > Or how > do you code possible drug seeking? Yes, you can write a database complex > enough; I've used it. There were about 100 options at each of 5 levels > (OK, I > exaggerate, but that's what it felt like); what happens is the most common > generic case becomes a catchall, and IT blame the medical staff because > they > aren't getting good data. > > In part because of that, we're a bit cautious about the black and white > approach > IT has to health care. > > > Without any question, the most common problem we have with IT in healthcare > is > time. The speed to use effect. Several mentions of this... seems like it really is central. > For an IT person, individual logins are a must; for me, it means I'm > logging into physically shared PC's, probably on average about once every > 5-10 > minutes given all the different usernames/ passwords are taken into > account; one > for radiology, one for labs, one for the internet, one for toxicology > database, > one for drug database... and of course autologouts are set so short that in > a > complex case you may have to relogin if you get interrupted (in my job, > interruptions average 30/hour). > > With every new program, it's "only a few seconds to log in". I can fill > out a > paper XR request in 1/4 the time I can do it on the PC (could the people > programming for Kodak please have to use their system for a day? I have 2 > separate usernames entered in 4 locations, two passwords... one Xray > request. > Autofill got invented a few days back). > Great example on how simple issues impact time. > > In my job, sometimes I don't have that time. Doctors don't "get" why we > should > have to go to court and explain why patients die because management > replaced a > paper based system that is efficient at the front end with a computer that > causes up front delays, even if there are significant back-office benefits. > We > are also being held to ransom by lawyers, and cannot afford poorly worked > out > processes- it's our career on the line when bad things happen because the > PC > takes forever to load/ we can't get results because its patch Tuesday/ our > login > died on a weekend and we're useless for an hour or more. We also go home > and > use programs that "just work", so we know we shouldn't have to tolerate bad > design on the job. > > Another major hurdle yet to be overcome is data entry. Doctors are, in > general, > not touch typists. Good point. Typing and data entry. The corellary is that doctors should be suscipcious of any UI that does not get close to keyboard-only control... but do not consider this often enough. > Voice recognition software is still a work in progress. > Until the barrier to entry for getting data on the system rapidly has been > lowered, IT will be seen as a problem, not a solution. Also, most medical > record keeping software also has issues with graphics; a picture may take a > thousand words, but it seems to take at least a million bytes, and I've yet > to > see or hear of a good implementation that allows graphics. And who's going > to > enter it all? I trained for 10 years to... do data entry. Slowly. Most > other > health care workers don't understand the nuances enough to enter accurate > data, > and I don't have the time. > > The data storage issue is bigger than what you've noted. Health care > records > often have to be kept- and be accessible- for the life of the patient, and > more > than once I've found information from the 70's that relates to the problem > I'm > trying to address at the time. In IT it may be acceptable to update your > database/ language occasionally; in my job, accessing old data is more > important > than riding the cutting edge. Don't forget it's not just a court case and > a > truckload of money off to your local litigation leech; it's potentially > lives > lost, and an doctor respecialising in tractor driving. And lawsuits aren't > rare > events: someone once told me the average American doctor spends more time > in > court than the average America
Re: [openhealth] Re: What do you keep explaining about Health IT
Hi, I'm an Emergency Physician, and program (badly) for the fun of it. First of all, there is a paradigm shift between medicine and IT. Medicine has a veneer of science, but once you dig a little, there's a truckload of touchy feely stuff that the science depends on, and that makes it very hard to tick boxes on a screen. This ranges from the really complex individual with multisystem problems who really went to ED because their granddaughter went away for the weekend, and they're not mobile enough to manage (unable to cope is not an allowed choice) to the elderly person who trips over because their eyesight is bad and breaks a hip. The hip is the reason for admitting them; the underlying problem will almost by definition be secondary (if at all recorded), but fixing that is the real long term solution to the patient's problem. Or how do you code possible drug seeking? Yes, you can write a database complex enough; I've used it. There were about 100 options at each of 5 levels (OK, I exaggerate, but that's what it felt like); what happens is the most common generic case becomes a catchall, and IT blame the medical staff because they aren't getting good data. In part because of that, we're a bit cautious about the black and white approach IT has to health care. Without any question, the most common problem we have with IT in healthcare is time. For an IT person, individual logins are a must; for me, it means I'm logging into physically shared PC's, probably on average about once every 5-10 minutes given all the different usernames/ passwords are taken into account; one for radiology, one for labs, one for the internet, one for toxicology database, one for drug database... and of course autologouts are set so short that in a complex case you may have to relogin if you get interrupted (in my job, interruptions average 30/hour). With every new program, it's "only a few seconds to log in". I can fill out a paper XR request in 1/4 the time I can do it on the PC (could the people programming for Kodak please have to use their system for a day? I have 2 separate usernames entered in 4 locations, two passwords... one Xray request. Autofill got invented a few days back). In my job, sometimes I don't have that time. Doctors don't "get" why we should have to go to court and explain why patients die because management replaced a paper based system that is efficient at the front end with a computer that causes up front delays, even if there are significant back-office benefits. We are also being held to ransom by lawyers, and cannot afford poorly worked out processes- it's our career on the line when bad things happen because the PC takes forever to load/ we can't get results because its patch Tuesday/ our login died on a weekend and we're useless for an hour or more. We also go home and use programs that "just work", so we know we shouldn't have to tolerate bad design on the job. Another major hurdle yet to be overcome is data entry. Doctors are, in general, not touch typists. Voice recognition software is still a work in progress. Until the barrier to entry for getting data on the system rapidly has been lowered, IT will be seen as a problem, not a solution. Also, most medical record keeping software also has issues with graphics; a picture may take a thousand words, but it seems to take at least a million bytes, and I've yet to see or hear of a good implementation that allows graphics. And who's going to enter it all? I trained for 10 years to... do data entry. Slowly. Most other health care workers don't understand the nuances enough to enter accurate data, and I don't have the time. The data storage issue is bigger than what you've noted. Health care records often have to be kept- and be accessible- for the life of the patient, and more than once I've found information from the 70's that relates to the problem I'm trying to address at the time. In IT it may be acceptable to update your database/ language occasionally; in my job, accessing old data is more important than riding the cutting edge. Don't forget it's not just a court case and a truckload of money off to your local litigation leech; it's potentially lives lost, and an doctor respecialising in tractor driving. And lawsuits aren't rare events: someone once told me the average American doctor spends more time in court than the average American criminal... Finally, local to you the USA has a huge problem with fragmentation due to the private health care system, and I'm willing to bet not even the President of the US will be able to shout down all the private interests making money out of your ill- health. The billing problems you mention are only the start; given that healthcare is about 18% of GDP, don't bet on being able to inject any common sense. There are too many vested interests making money on it staying just the way it is, and unless greed s
