Technology in the Medical Industry

I started an interesting conversation over on my blog and thought it would be good to discuss here.

My two posts are here: http://radio.weblogs.com/0001011/2005/04/07.html#a9829

and here:

http://radio.weblogs.com/0001011/2005/04/07.html#a9830

And the comments I've received so far are:

I work at a hospital, and have worked in the medical field for most of my working life, both as a health care provider and in IT. Patient medical records are done electronically in many hospitals. In fact, the hospitals I've been to in the past few years provide wireless devices for nurses and doctors to take to the patient's bedside for charting and monitoring. Many of the monitoring devices are networked, at an increased cost. However, and this is the thing that gets lost in translation for non-medical people, there are many things that cannot be monitored by machine and require hands-on interaction. And while there is always room for improvement, some things will never be able to use a software solution. Candy • 4/7/05; 4:45:29 PM

This gives a whole new meaning to the phrase "Blue Screen of Death"... Steven J. Ackerman • 4/7/05; 4:51:34 PM

You hit the nail on the head. The opportunity is getting systems to talk to one another. We can see this today in a single vendor's proprietary/closed "stack" of devices and protocols that work well together. This is easiest to implement, plus it gets consumers hooked on a single brand. Interoperability across products from different, competing, vendors will come eventually but is a much harder problem to solve. In many respects, hardware "hackers" are blazing new ground here. Look at all of the cool tricks people are doing to their iPods, TiVos, Media Centers, whatever. These people want their devices to communicate, and they are making it happen, despite obstacles put in place by the original vendors. In the end, software and hardware is more valuable when you assemble the pieces together in unexpected ways. Vendors that "get this" and produce hackable/pluggable/interoperable products will win in the end. The current state of proprietary devices that do not talk to each other cannot and will not last. Eric Burke • 4/7/05; 4:56:30 PM

Opportunities there are, but also barriers a-plenty. I happened to give a presentation a few months ago alongside a chief architect for a major supplier of medical systems. He said that while the company was indeed making systems that would work together, one of the issues is that hospitals deinfe their own ways of using these systems. It would seem that we do not have a lingua franca that spans even hospitals in the same city, let alone nationally or globally.. The end result is that it becomes extremely expensive to retroactively impose universal schemes - even within the same hospital organisation. And then it's down to "what is the chance that we would be hit by a major malpractice suit versus the cost of putting things right". Bean counters rule, because shareholders want returns on their investment. Business investments get made following one of three drivers: Greed, Fear, and Rightness. That last driver - do it because it's the right thing to do - is, alas, the weakest and far behind the other two. Such is human life. Geoff Coupe • 4/7/05; 4:59:04 PM

I see this.. http://meddb.plocp.com/"> http://meddb.plocp.com. My mini immune db I created for a doctors office. They use to crank out this immune forms for school/camp/sports leagues. It tracks stuff. Now only to get some salesforce working on spreading it thoughout the land... http://meddb.plocp.com/" rel="nofollow">aram comjean • 4/7/05; 4:59:17 PM

Sorry mate, saying "the opportunities are there" without having any idea of what the challenges that IT organizations in healthcare face is useless. Budgets and politics are the two largest barriers. I know you said convincing people is key, but it's deeper than that. In a typical healthcare IT shop, the people who make the decisions aren't the people who know what the technology can do. So you get this barrier where IT has to first convince itself that it's a good idea, then convince itself it's a priority (most projects die in either of these two stages) and then finally work on convincing the outside world (this is where real politicing comes into play). It's rare for a healthcare IT group to have any authority to change anything. A perfect example is the last hospital I worked at. Our 3 year, "revolutionary" (for our province) plan? Standardize server OS and hardware. Move to a common directory services platform. Move to a common mail platform. Move to a common desktop management platform. None of that changes patient care stuff. But without that type of foundational stuff you can't even get to the patient care opportunities. Then the problem becomes who pays for patient care IT upgrades. Not IT. And patient care budgets would rather spend money on more beds, more nurses, better analysis, etc than spend it on IT. It's a far bigger conundrum than anyone who's never been there can understand. Jeremy C. Wright • 4/7/05; 5:10:38 PM

Amen! I'm amazed at the stories I hear about the lack of technology in hospitals. Why not do some amount of monitoring of vitals using something like Statistical process control. Charting the results of vitals would allow looking for trending. While this type of automation wouldn't eliminate all of the measures, it would reduce the amount of time it takes to do a patient review. If anyone is starting a company to do this, sign me up! I would love to work on this type of thing. Mike Mangino • 4/7/05; 5:16:42 PM

A few thoughts - simple, glib answer "what happens if a blue screen of death becomes literally that..." (i.e. the technology is good, but has to be many orders of magnitude more reliable when it is dealing with issues of life or death, not the latest sports score, blog post (Blog or Die not withstanding), or hot single) Less simple, less glib answer - that while video and technology holds a lot of value, there is a lot more going on with an in person visit and interaction than can be captured even by the best HDTV cameras. Smell, sounds, tones, subtleties of how someone reacts quickly, what their body is doing when they answer etc. I'm not a doctor but think about just within our own families - can't you tell just by looking at your loved ones when they are feeling pain or discomfort, even if they put on a good face, something subtle about their posture, face, tone of voice, little microgestures etc sends out a "not doing so well" vibe. Plus, much like how a waitperson's tip has been shown to increase if they even briefly physically touch their guests (a brief pat on the back, brush of the hands etc), the physical interaction with a doctor goes a long way towards how the doctor is viewed. I know there have been studies which show that more than the actuality of whether or not a doctor made a mistake, the best predictor of who will in fact get sued has to do with how they interact with their patients - i.e. a doctor who makes a mistake but who seems nice and responsize gets sued less frequently than perhaps a "better" doctor technicaly but who is rude, abbrasize and seems not to care. From a technical perspective I agree with the other commentors about the unlikelyhood of a "universal" solution getting traction. Rather, I would suspect that systems that "play nice" and work well with legacy systems while gradually moving data to a more consolidated and universially useful form will gain traction - but at the same time they face serious HIPAA (and other) regulatory requirements and restrictions. (imagine what happens if a doctor's PDA holds 1000's of patients medical records and is stolen or lost. But at the same time while security is needed, what happens if a doctor needs to access a patient's chart in the middle of major crisis and forgets a password, miskeys something, is wearing gloves and other protective gear so biometric means are not available etc. Paper charts for all their failures have the advantage of not needing power, boot cycles, login times, etx. All they need is some form of illumination (generally available when a doctor has to work) Shannon Shannon Clark • 4/7/05; 5:19:20 PM

Hospitals going high-tech wouldn't cut down on malpractice suits - people would just sue Microsoft instead. And what a juicy target MS would make, with all its billions in the bank. P.S. How many Microsoft EULAs include the clause about 'do not use this product where failure could result in property damage or loss of life'? No one of consequence • 4/7/05; 5:22:59 PM

Robert, You want a video of a doctor putting their finger in your butt? There are a lot, and I mean a LOT, of privacy issues regarding some of the things you are suggesting. How can you tell from a video tape if the clear liquid that a nurse injected into that line really is the correct drug in the correct dosage? All that a video camera is going to capture is the gross, and by gross I mean large, stuff. Like nurses smacking patients that they don't like and such. Here's the thing, I agree that the technology has to get better and at a lot of hospitals it is. Who's going to pay for it? It's not just a matter of cost though. Hospitals are pretty rich. Look at the average patients chart, well you can't really. You can look at your own chart. HIPPA probably won't let you look at your wife or another family members, maybe your sons. Take a good, long look at all of the informtion contained in it. Now, count up how many people you see in the waiting room and think about how long it will take to enter in all of their data. Again, that's just the people you can see in that waiting room, on one floor, in one hospital. Imagine doing that for just everyone that works at Microsoft. And those people probably don't have anything all that serious wrong with them. We have entire file drawers dedicated to a single patients chart for some of our bone marrow transplant patients. A lot of that information is kept in an EMR, electronic medical record, but what happens if the patient is in a car wreck on vacation and they are seen at another hospital? How do you keep the information synchronized statewide, let alone nationwide or worldwide. If you say BizTalk I'm gonna smack you with a cluebat. How do you convince machine manufacturers that it's in their best interest to make their data available in a standard format? How long did it take for Microsoft to open up it's Word format? Doesn't it make sense to have a standard format for information? So I can open up my Excel worksheet in Lotus 1-2-3 right? Every manufacturer is going to say, "our format should be the standard because we did the most research and determined that this is the best way to represent this data". What happens if their machine supports a very useful diagnostic feature that doesn't fit into the standard? Do you think there are standard data interchange formats for medical data? There's standard terminology, but what data points mean can vary from institution to institution and , most often, from physician to physician. How do you represent that in a standard format? There have been some attempts at medical information standards, but the problem is that there isn't one large guiding force in the medical industry. I'm right there with you that things have to change. Heck, I'm one of the people working to make them change. It's a LOOOOOOOOOOOOOOONNNNNG uphill battle, but we are winning it somewhat. Look at the new ER in Issaquah, check out the system that Swedish hospital has in place. Tablet PC's and everything. I know a few hospials that my wife worked for in New Mexico were completely electronic. The Group Health hospital I was at just this morning (HIPPA VIOLATION!) has Dell terminals in every room and their EMR is synchronized between all GHC facilities. Is the system failproof, hell no. A nurse this morning couldn't enter some data because someone was "fiddling with the system.". Lets look at your IV pump example. In either case, a blockage or out of fluid, a nurse is still going to have to check it and either clear the blockage or replace the bag. How does having a different beep really help? Isn't the idea behind the beep just to call attention to the machine? Which case is more important. How much does an MRI machine cost? What does a blood pressure cuff have to say to an IV pump? A malpractice lawsuit is much less expensive than replacing every piece of equipment in a hospital with a new piece that can talk to the others. You don't really solve any problems by making the machines smarter, in fact you create new ones. You still rebooting your Tablet PC every night? Imagine doing that for every piece of machinery in a hospital every night? If you don't, what happens when you are working a trauma case and you have to look up a patients chart only to have your Tablet lock up because it's been on for 5 days straight. Worse yet, what happens in a disaster? Think a lot of doctors were booting up their PC's during the Tsunami? A paper chart is never "down for maintainence". Can you tell you hit a hot button for me? I deal with this EVERY DAY. We have to deal with data whose meaning changes over time even if the actual data doesn't due to advances in technology and knowledge. The amount of data and the complexity of the data is MASSIVE. Scott • 4/7/05; 5:25:03 PM

P.P.S. On the other hand, doctors at the clinic I go to all have Tablet PCs.... No one of consequence • 4/7/05; 5:25:31 PM

What are you talking about Scoble. I know people who work in software giant whose main business is medical software. They have really good systems, from pda software for nurses (so that they don't use paper pen so much), to live updates to the software, to powerful servers.. were you in myanmar or something??.. Anyways,there is always room for improvement and technology has a long long way to go. dev • 4/7/05; 5:26:22 PM

on a side note. Is SQL server FDA certified yet? I know we've got some 21-11CFR issues with using it. Scott • 4/7/05; 5:31:18 PM

I think liability fears drive a lot of that resistance to remote doctoring, i.e. "if you'd only been in the room, you would've seen these other symptoms..." But really now, every time that surgeon pops his head in, he's charging your insurance $70 (or more). I wish I could do that. Matthew B • 4/7/05; 5:05:47 PM

It may not be as hard as you first envision. I worked at Children's, here in Seattle, for about 3 years as they Lead Web Developer and I know, all too well, the resistance to change you're talking about. But, if it's kept up and taken slowly breakthroughs can and will happen. I had the opportunity to do on-site usability testing with nurses and it was a serious eye-opener to see how they used technology, or maybe more accurately, didn't use it. Or tried to use it? LOL. As well, I met with many docs, in their own environment and I was amazed at how far behind me tech-wise they were. It'll take time, for sure, but there is so much opportunity to help out there, it's well worth it is my guess. Keith • 4/7/05; 5:24:13 PM

The Healthcare IT sector is one of the hardest to move. main reason: They pay big bucks for complicated systems, and then want to fully depreciate the systems before looking to new ones. Average cycle about 7-10 years. FWIW, Ross Perot had been extoling the virtuals on reducing healthcare costs by reducing the manual way of doing things. He was laughed off as being crazy. I believe he wasn't (I can't vouch for other times ). I agree that the HC IT industry needs to move faster, but we also have to balance the need of having personable people, vs robots (cold people) BTW-- Blogging won't fix it. The solution starts at the CIO level. JeffMc • 4/7/05; 5:28:13 PM

Here is a link to a post at over at engadget entitled "Detroit hospital rolls out "robotic" doctors": http://www.engadget.com/entry/1234000910039231/

Personally, I am of the school that beleives that the hold-out of the medical profession has run its course and that we are on the cusp of an era of hyper-adoption of medical technologies at every level of care.

   

technology in the ER.  the NHS is about 10 years behind america, it's only because america has private funding are we behind, we have to spend our money on MSRA (not microsoft research academics).

recently i had a mumps booster, huge panic hitting various unis thus needing a big immunisation program, how was it done?  paper based no, medical records at all, tell your GP after.  Well its very rare that i ever see a doctor as i'm in excellent health generally.  but wouldn't it have been better if all the nurses there had tablets with all our medical records on? 

electronic records are coming in the UK, so my mum tells me (who works in databases for the NHS) but it's taking it's time and the fact that many surgeries use many differnt systems, not just window systems.

why doesn't MS get more involved, i know Bill wants electronic records and more IT in the NHS.  Why didn't MS UK get involved it would be a great marketing ploy, give the immunisers some tablet pcs and synch up medical records and do a tablet pc roadshow around the UK universitys.  would be excellent marketing.

I don't care about all these CSS 2.x stuff i don't care about linux being > * and MS suxoring (which it doesn't).  but i'd like to see MS work with every other platform supplier and develop a system where every GP's surgery, every NHS hospital, every bit of IT in NHS run hospitals runs with every platform allowing access wherever you are whether dr patel is running solaris or dr evans is running Windows XP Home, it should all work.  thats what i want to see MS do in the next 5 years.

For my UK Imagine Cup SDC last year my team designed an application that allowed paramedics to send data direct from the ambulance to the hospital ready for the arrival of the patient. The icing on the cake was that we could also send a photo (taken using an XDA2) direct to a db and then onto the assigned dr, who could then draw on it using their tablet and send the image back to the ambulance (it impressed the judges).

Technology can help alot in the medical industry, I think there will be some interesting and innovating software making its way into hospitals, especially with tablets which are a perfect tool for doctors.

Microsoft are working with the UK NHS

http://www.microsoft.com/uk/government/health/default.mspx

we have to remember alot of dr's are still quite old, i know mine is in his 50s.  so that means support and lots (if judging my dad is anything to go on) of it.

yes you're right MS do have a huge contract with the NHS, but i wonder how closely MS work with the various departments judging by what my mum has said.

My mother-dear is in the NHS as well... As some random "Senior Physiotherapist"

And to give a general idea of the state of IT in the NHS...

She has to change her NTLogon password every 30 days... she doesn't even know where "a" is on the keyboard.

All her documents and memos are sent over email, she can't figure it out so she gets her collegues to do it for her.

Back when it was printed paper, she was used to read through them in the car or at home, she doesn't have the technological know-how on how to transfer her email and documents to read and go over at home or on the road. Besides, I doubt the NHS is going to subsidize a tablet PC with GPRS or Wifi connection so she can do things like she used to.

This is but one shortcoming of increased deployment of IT.

I'm all for it in hospitals and such... but it's really just being counter-productive for those on the road all the time, her mobile phone doesn't even support WAP for Exchange Mobile.

we need more UK MS posters here, it feels to me that MS Reading is justa  building possibly with a few janitors in it,  but they never do the cool things that redmond do. 

anyhow, i only ask for UK MS Posters here is because they might, hopefully would, understand the UK NHS political IT scene than robert or charles does.



Dear Gretchen

I am very passionate about technology and have great communication skills (please don't take any blog or forum posts at current to show this).  Please employ me at MS reading to be the UK equivalent of Scoble.

Cheers very much

Jaz

Jaz wrote:

Please employ me at MS reading to be the UK equivalent of Scoble.

That's Johnathan's job

Unfortunately the technology, especially protocols that are so arcane they would make internet  users choke (like DICOM and HL7), has not kept up with changes in the rest of the world. I managed a company doing medical speech recognition products, and most of our time was spent trying to work with protocols that could have easily been adapted to XML, but too many consultants make too much money working with the existing stuff so there is no motivation to change.

And change takes so long in the medical business anyway.

I'm not saying there aren't opportunities to improve healthcare using information technology--I worked in healthcare IT for seven years and still consult in the field from time to time--but I think that this approach (a typical techie approach, incidentally) of "what are the opportunities" is part of the problem.  I think that approach is commonly a problem, in that it ignores the special circumstances of any industry, and also that it tends to activate the very attitudes that Robert identifies as being problematic.  He's correct--they are--but they can be minimized by adopting an entirely different approach to integrating information technology than his conventional approach.

Many people are just completely put off when someone, anyone, comes in with little or no knowledge of their job or industry and starts spouting ideas for improvement.  We in the technical world recognize this as open-ended brainstorming, and are less defensive about it than many.  But when you try it on someone who may be apprehensive about technology in the first place, say a medical professional, and worried about their job in the second, you've got a recipe for rejection.

In my experience, it's better to start the other way around--learn the industry and listen to what the people who know it are saying.  Don't be aggressive and far-reaching, and for the love of all that is holy, don't try anything on the edges of functional technology.  This is especially true in the medical industry.  The sorts of teething pains and integration issues that we in technology accept as the price of doing business are completely unacceptable in medicine, due to a number of factors:  One, HIPAA regulation is stringent and far-reaching, and is going to have some teeth when enforcement really kicks in; Two, the pucker factor of messing up information storage and retrieval goes up considerably when you have lives on the line; and three, staffing shortages and revenue demands stretch nerves to the breaking point where a single hiccup in an electronic chart can throw the entire clinic/floor/whathaveyou into chaos and condemnation.

That's why prototyping 50 beds is a lousy idea... it's not so much the small roll out as a demonstration of technology feasability (which is a good idea) but the thought of charging into it without dotting every I and crossing every T.  The bugs that we considerable inevitable will condemn the entire project from the point of view of the medical staff who have to work with it; and word travels quickly.

Then, too, there is the rapid rate of change in medicine that makes it difficult to implement complex, regimented IT systems to handle.  Other than IT itself, there may be no other general field which reinvents itself so quickly and so often.  New procedures, methods, and practices cascade down regularly, and computers are frankly horrible at accomodating them.  There are places where the flexibility of a human mind and hand will always outpace a computerized system in efficiency.

When you start mucking with billing systems, then the patients get the good news as well....

At any rate; don't assume that the failure rates which we have come to accept in other industries are tolerable here, and don't underestimate the repercussions of even ONE goof on a Tablet PC that a pen and paper chart would have handled with elan.  If technology in the healthcare industry seems Dark Age to us, it's exactly because previous crusaders have failed to recognize or adhere to these principles, and the backlash is with us all today.

Get to know the systems you propose to replace, and alter them only in moderation and with thorough bullet-proofing.  Don't try to replace manual systems that cannot be documented in sufficiently regimented detail, and accept that computers are not always the appropriate answer to certain muddled situations.  Sometimes, the muddle really is the most appropriate system.

ScuzzMonkey wrote:

That's why prototyping 50 beds is a lousy idea... it's not so much the small roll out as a demonstration of technology feasability (which is a good idea) but the thought of charging into it without dotting every I and crossing every T.  The bugs that we considerable inevitable will condemn the entire project from the point of view of the medical staff who have to work with it; and word travels quickly.

Exactly. Would you want your loved ones vital signs to be monitored by beta code? Generally, people that enter a clinical research protocol, and experimental protocol, don't have much of an alternative.

At any rate; don't assume that the failure rates which we have come to accept in other industries are tolerable here

And don't assume that every failure of the paper system would have been caught by an electronic gizmo (to use an industry term). If a doctor wrote down the wrong blood type for a patient on the paper chart, what makes you think that the doctor would have entered it correctly into an EMR? What kind of assurance do you have that the EMR that you get from another hospital is accurate? Are you willing to bet your patients life on it? Or are you going to run a test to check the patients blood type anyway? What happens if your database is corrupted and then propagates the corrupted data into the system? If it's financial data, you can probably go to a backup and restore it. What if it's lab data for a patient in the hospital? That kind of data might need to be pulled up if the patient goes south and it would need to be pulled up real quick.

I posted my thoughts here- http://www.clrsql.com/archive/2005_04_07.html#000068.

Basically, if you take even 5 minutes to come up with a short list of ideas on how software can improve your day to day life, you can come up with some pretty interesting scenarios very quickly. I'd love to hear other people's thoughts.

And don't assume that every failure of the paper system would have been caught by an electronic gizmo (to use an industry term).

Good point.  And another point in the same vein is that failures in the pen and paper system are typically seen (appropriately) as a failure of the user.  Failures in electronic systems are predominantly seen as a failure of the technology, even though as a practical matter they are often failures of the end user.  But I don't think that it's entirely correct to blame the user in those cases, either, even if they are technically just as incorrect as in the manual case.  Few technology systems are designed to be as simple to use as the manual system, and errors, even if they are user errors, can reasonably still be attributed to the technology--if it's not as simple or more so than the manual system, it's causing the problems.

We are doing it:
https://www.saintlukeshealthsystem.org/slhs/com/system/eicu.htm

Pretty cool stuff.

Dedalus wrote:

Here is a link to a post at over at engadget entitled "Detroit hospital rolls out "robotic" doctors

MED I BOT!

lukes wrote:

Microsoft are working with the UK NHS

http://www.microsoft.com/uk/government/health/default.mspx

The new NHS system project is huge. Massive. I believe its one of the biggest MS technology products in the world.

However, I have to say that the UK government (under all flavours of administration) has a very poor record when it comes to procuring and implementing IT systems - and the NHS project has its share of problems. Somewhere, somone is taking Billions of pounds worth of tax payers money is going to IT systems that never get implemented, fail in development, or turn out to be next to useless.

As an aside, my small sector of the industry is kind of health related. I am not a billionaire. I think Im doing somthing wrong.

I wonder how many backdoors microsoft leaves open on government software.. shhh;)