Living and Dying with Fragmented Medical Records
September 26, 2018 9:37 AM Subscribe
Michael Champion’s journey through healthcare in the U.S.
Every year, an untold number of patients undergo duplicate procedures — or fail to get them at all — because key pieces of their medical history are missing. Why? We know it’s not a perfect system. We know there will be gaps. But what choice do we have?
Every year, an untold number of patients undergo duplicate procedures — or fail to get them at all — because key pieces of their medical history are missing. Why? We know it’s not a perfect system. We know there will be gaps. But what choice do we have?
What choice do we have? How about a single payer system with unified medical records?
Regarding scruss's comment, that is borne out right in the article:
Sometimes, the barriers are intentional, DeSalvo says, as vendors don’t necessarily want to make it easy to share data with hospitals using competitors’ systems and may charge them a fee to do so.
posted by elmay at 10:55 AM on September 26, 2018 [6 favorites]
Regarding scruss's comment, that is borne out right in the article:
Sometimes, the barriers are intentional, DeSalvo says, as vendors don’t necessarily want to make it easy to share data with hospitals using competitors’ systems and may charge them a fee to do so.
posted by elmay at 10:55 AM on September 26, 2018 [6 favorites]
I used to help develop Medicaid programs. Many states have attempted to create these unified systems in an effort to consolidate care and therefore save money, and many programs have set out to achieve them. They always fail.
One huge problem is that every provider (and every state-funded program) has its own system, its own codes, its own datafields. It's incredibly difficult to standardize the data on such a scale and with so many contributors. One provider's system might use a 34 information fields and 27 different proprietary codes. Another provider's system might have 100 information fields and 12 proprietary codes. Another's has 200 fields and 30 different codes. And take something like the overarching ICD system that codes different medical conditions -- the ICD updates every several years with more and more codes that branch out to add more specificity. Not everyone updates to the new ICD schema at once.
How do you begin to reconcile that? How deep would the record go to capture data that would be comprehensively important and useful for orthopedists and urologists and psychologists and gynocologists and social workers and your family doctor? And once you set a standard, you'd also have to go back historically for every patient as well. And then each state might have its own requirements, so if you move between states, how would that get shared? What new gaps would arise? And that's assuming the data is correct and correctly entered to begin with. Which it often is not.
The best solution, really, is a national medical system with a national EHR database.
posted by mochapickle at 11:16 AM on September 26, 2018 [18 favorites]
One huge problem is that every provider (and every state-funded program) has its own system, its own codes, its own datafields. It's incredibly difficult to standardize the data on such a scale and with so many contributors. One provider's system might use a 34 information fields and 27 different proprietary codes. Another provider's system might have 100 information fields and 12 proprietary codes. Another's has 200 fields and 30 different codes. And take something like the overarching ICD system that codes different medical conditions -- the ICD updates every several years with more and more codes that branch out to add more specificity. Not everyone updates to the new ICD schema at once.
How do you begin to reconcile that? How deep would the record go to capture data that would be comprehensively important and useful for orthopedists and urologists and psychologists and gynocologists and social workers and your family doctor? And once you set a standard, you'd also have to go back historically for every patient as well. And then each state might have its own requirements, so if you move between states, how would that get shared? What new gaps would arise? And that's assuming the data is correct and correctly entered to begin with. Which it often is not.
The best solution, really, is a national medical system with a national EHR database.
posted by mochapickle at 11:16 AM on September 26, 2018 [18 favorites]
I work on a team developing a unified medical record repository but it's only for one hospital system. We're a big damn health system that covers most of western PA, but even so, the number of different internal record systems that we have is dizzying. Heck, the main hospitals in the system don't even same EMR for in-patient and out-patient records and then since we expanded through acquisition, each new hospital comes in with a different EMR.
We're just one hospital system and we struggle to unify our own records. Doing that across systems is going to be another level of complexity.
posted by octothorpe at 11:33 AM on September 26, 2018 [3 favorites]
We're just one hospital system and we struggle to unify our own records. Doing that across systems is going to be another level of complexity.
posted by octothorpe at 11:33 AM on September 26, 2018 [3 favorites]
These programs always seem to get held up building out a rigorously-specified data model, as mochapickle describes. There are just too many different fields, codes, datatypes, rules about mandatory fields, etc. Every time someone starts out to create a new "One Standard To Rule Them All", what ends up happening is just one more in a proliferation of 'standards'.
One solution to this is to just throw enough resources at the problem to finally crack the nut once and for all. Which means working from the nation-state level on down, because you'd literally have to have a monopoly on violence to get people to use the thing. And it would probably also require a wholesale reform of the US healthcare system—which might be great for other reasons, but that's a hell of a hurdle if you're trying to get EHR.
The second solution is to punt on the whole data model stuff. Stop trying to encode each particular field as a value in a database. Move up a level, and try to come at the problem in a different way.
I'll make an analogy to character encodings. Encoding human writing into computers is hard; it's on the order of the EMR problem, because there are just so many different letters in so many different languages and they can be represented as glyphs in so many ways. It's 2018, we've been working on this problem since the very early teletype era (Baudot, patented 1874), and Unicode, the brute-force one-standard-to-rule-them-all, has really only taken firm hold in the last decade or so—and it's still being worked on. I'd expect EMR, done the hard/correct way, to take about as long.
But it turns out that back in 1929, when various national encodings were proliferating like crazy, and a truly universal encoding was as lofty a goal as universal EMR is today, some people decided to punt on the whole thing. The solution they came up with was called Hellschreiber, and it's (IMO) pretty crafty. Instead of trying to come up with a 1:1 relationship between every written glyph someone might want to transmit and a numeric value, and translate back and forth on each end using an explicit or implicit lookup table that has to be the same (and deal with the inevitable oversights and extensions to the set somehow), they just punted on the whole question of glyphs. They used a 7 pixel by 7 pixel grid for each letter, and transmitted what we'd now call a raster down the wire. If you wanted a new character, fine, whatever, as long as you can represent it in a 7x7 grid. You want umlauts? Great. Currency symbols? Fine. English? German? French? Latin vs. Cyrillic? The encoding doesn't care. (You can even do Simplified Chinese in 7x7 pixels, although it gets ugly.) It also has some nice properties in terms of dealing with noise well, and this is mostly what it's remembered for today, but I think it's a nice example of a solution that solves a problem without going down the same well-trod brute force path as most others.
The analogous solution for EMR might be to stop trying to encode everything down to the field level, and instead just do documents. Most medical records are still reducible to "pages" and "documents" or some similar level of categorization. We have established paradigms that allow for nested hierarchies (folders) while still allowing data to be searched across the hierarchy. Optional metadata, with the basics populated during ingestion (what source did it come from? when? who submitted it?). But it needs to not make the perfect be the enemy of the good, by demanding that the system have a priori knowledge of every type of data that might be stuck into it. This isn't rocket science, it's basically Sharepoint on steroids. And even if that makes a lot of people retch, it'd be a hell of a lot better than what we have now, which is nothing, because everyone keeps throwing themselves at the Hard Problem and busting their skulls trying.
You could probably build a not-totally-useless system on top of any number of document management systems glued together with WebDAV today, if you wanted to. Would they be perfect? No. Would they let you slice and dice medical information perfectly? No — but again, it's better than what we have now, which doesn't allow for that either.
And to be honest, if you want this worse-is-better approach in your own life, today, you can do it by taking a cheap fax-to-email gateway service with a dedicated number, a Gmail rule that moves around or resends those incoming documents, and a document management system that performs OCR like Evernote. Have doctors fax all your records to the number (which they'll almost always do, particularly if you don't make it obvious that it's your personal number), then have Gmail label them and optionally stick them into Evernote where the PDF attachments will get OCRed and made searchable. Fax is effectively the API glue, because it's universal, and the basic datatype is the faxed page, or maybe the PDF document that comes out of the inbound fax link. When you want to share data, you just forward the email to the fax gateway service with the number (depending on how your service works) and it's on its way. This is all very stupid, very kludgy, and likely to make Real Enterprise Programmers vomit a little, but it's better than any "real" EMR system I've used as a consumer, in terms of letting you own your own data, and not being tied into a provider or insurance company system, and easily getting things into and back out of it. (It doesn't do medical imaging very well, is about the only big limitation.)
Anyway, sometimes the best solution is to not keep trying to solve the problem that everyone else has tried and failed to solve.
posted by Kadin2048 at 12:35 PM on September 26, 2018 [19 favorites]
One solution to this is to just throw enough resources at the problem to finally crack the nut once and for all. Which means working from the nation-state level on down, because you'd literally have to have a monopoly on violence to get people to use the thing. And it would probably also require a wholesale reform of the US healthcare system—which might be great for other reasons, but that's a hell of a hurdle if you're trying to get EHR.
The second solution is to punt on the whole data model stuff. Stop trying to encode each particular field as a value in a database. Move up a level, and try to come at the problem in a different way.
I'll make an analogy to character encodings. Encoding human writing into computers is hard; it's on the order of the EMR problem, because there are just so many different letters in so many different languages and they can be represented as glyphs in so many ways. It's 2018, we've been working on this problem since the very early teletype era (Baudot, patented 1874), and Unicode, the brute-force one-standard-to-rule-them-all, has really only taken firm hold in the last decade or so—and it's still being worked on. I'd expect EMR, done the hard/correct way, to take about as long.
But it turns out that back in 1929, when various national encodings were proliferating like crazy, and a truly universal encoding was as lofty a goal as universal EMR is today, some people decided to punt on the whole thing. The solution they came up with was called Hellschreiber, and it's (IMO) pretty crafty. Instead of trying to come up with a 1:1 relationship between every written glyph someone might want to transmit and a numeric value, and translate back and forth on each end using an explicit or implicit lookup table that has to be the same (and deal with the inevitable oversights and extensions to the set somehow), they just punted on the whole question of glyphs. They used a 7 pixel by 7 pixel grid for each letter, and transmitted what we'd now call a raster down the wire. If you wanted a new character, fine, whatever, as long as you can represent it in a 7x7 grid. You want umlauts? Great. Currency symbols? Fine. English? German? French? Latin vs. Cyrillic? The encoding doesn't care. (You can even do Simplified Chinese in 7x7 pixels, although it gets ugly.) It also has some nice properties in terms of dealing with noise well, and this is mostly what it's remembered for today, but I think it's a nice example of a solution that solves a problem without going down the same well-trod brute force path as most others.
The analogous solution for EMR might be to stop trying to encode everything down to the field level, and instead just do documents. Most medical records are still reducible to "pages" and "documents" or some similar level of categorization. We have established paradigms that allow for nested hierarchies (folders) while still allowing data to be searched across the hierarchy. Optional metadata, with the basics populated during ingestion (what source did it come from? when? who submitted it?). But it needs to not make the perfect be the enemy of the good, by demanding that the system have a priori knowledge of every type of data that might be stuck into it. This isn't rocket science, it's basically Sharepoint on steroids. And even if that makes a lot of people retch, it'd be a hell of a lot better than what we have now, which is nothing, because everyone keeps throwing themselves at the Hard Problem and busting their skulls trying.
You could probably build a not-totally-useless system on top of any number of document management systems glued together with WebDAV today, if you wanted to. Would they be perfect? No. Would they let you slice and dice medical information perfectly? No — but again, it's better than what we have now, which doesn't allow for that either.
And to be honest, if you want this worse-is-better approach in your own life, today, you can do it by taking a cheap fax-to-email gateway service with a dedicated number, a Gmail rule that moves around or resends those incoming documents, and a document management system that performs OCR like Evernote. Have doctors fax all your records to the number (which they'll almost always do, particularly if you don't make it obvious that it's your personal number), then have Gmail label them and optionally stick them into Evernote where the PDF attachments will get OCRed and made searchable. Fax is effectively the API glue, because it's universal, and the basic datatype is the faxed page, or maybe the PDF document that comes out of the inbound fax link. When you want to share data, you just forward the email to the fax gateway service with the number (depending on how your service works) and it's on its way. This is all very stupid, very kludgy, and likely to make Real Enterprise Programmers vomit a little, but it's better than any "real" EMR system I've used as a consumer, in terms of letting you own your own data, and not being tied into a provider or insurance company system, and easily getting things into and back out of it. (It doesn't do medical imaging very well, is about the only big limitation.)
Anyway, sometimes the best solution is to not keep trying to solve the problem that everyone else has tried and failed to solve.
posted by Kadin2048 at 12:35 PM on September 26, 2018 [19 favorites]
What sort of systems do the various “socialist” nations use in their national systems? Couldn’t the US just swallow its collective pride and adopt one of those, rather than roll our own?
posted by Thorzdad at 1:36 PM on September 26, 2018 [1 favorite]
posted by Thorzdad at 1:36 PM on September 26, 2018 [1 favorite]
British person here - one thing that puzzled me about the story of the patient was the part where he sits the patient's partner down and says "Because of that, I said, I am going to tell you the details of his medical care." Why doesn't he just give her a copy of his discharge letter with all that information in it? When I've been discharged from hospital I've been given a copy of my discharge letter every time, and someone has checked that I understood it, and understood the medication instructions.
How about a single payer system with unified medical records?
This is also a problem in the NHS. The reason is that from the 1970s onwards, enthusiastic people in various places saw the potential of computers in healthcare, and started working on ways to use them in practice. So multiple systems emerged in the same way they did in the USA, and once an organisation the size of the average large hospital starts using a system, it's much easier to update that system than to change to a new one.
However, it's something that is being worked on, and is improving all the time. Certainly in Scotland, if you're admitted to hospital in the health board area you live in, the doctors will be able to access medication information from your GP, all your previous blood and scan results done locally, as well as all letters from other specialities.
posted by Vortisaur at 1:36 PM on September 26, 2018 [1 favorite]
How about a single payer system with unified medical records?
This is also a problem in the NHS. The reason is that from the 1970s onwards, enthusiastic people in various places saw the potential of computers in healthcare, and started working on ways to use them in practice. So multiple systems emerged in the same way they did in the USA, and once an organisation the size of the average large hospital starts using a system, it's much easier to update that system than to change to a new one.
However, it's something that is being worked on, and is improving all the time. Certainly in Scotland, if you're admitted to hospital in the health board area you live in, the doctors will be able to access medication information from your GP, all your previous blood and scan results done locally, as well as all letters from other specialities.
posted by Vortisaur at 1:36 PM on September 26, 2018 [1 favorite]
What sort of systems do the various “socialist” nations use in their national systems? Couldn’t the US just swallow its collective pride and adopt one of those, rather than roll our own?
It's not a matter of swallowing pride, it's the fact that there are already multiple competing systems out there and agreeing on some common standard and then converting the data over would require lots of money, lots of time, and none of the players have any interest in doing it. The various software suppliers don't want it, because it makes it easier for their customers to move to a competitor. The various hospitals and health organizations don't want to do it because it doesn't actually benefit them. They understand their system. There is nothing that the new system would offer for day-to-day operations that is better than what they have now. Why should they be excited about switching?
posted by It's Never Lurgi at 2:10 PM on September 26, 2018 [1 favorite]
It's not a matter of swallowing pride, it's the fact that there are already multiple competing systems out there and agreeing on some common standard and then converting the data over would require lots of money, lots of time, and none of the players have any interest in doing it. The various software suppliers don't want it, because it makes it easier for their customers to move to a competitor. The various hospitals and health organizations don't want to do it because it doesn't actually benefit them. They understand their system. There is nothing that the new system would offer for day-to-day operations that is better than what they have now. Why should they be excited about switching?
posted by It's Never Lurgi at 2:10 PM on September 26, 2018 [1 favorite]
EHR interoperability is exactly the thing that I work on. It is clearly an ugly problem but there are some reasons for optimism.
The FHIR standard has built up quite a bit of traction in the healthcare tech world. Probably more than any previous effort. US healthcare providers and EMR vendors are having their arms gradually twisted by the "meaningful use" requirements and the incentives behind them.
Hospitals and health systems actually have a surprising amount of reason to embrace standardization - they can't even exchange data inside their own walls. A new generation of patients raised on the internet is not going to take "sorry, our fax machine is broken" as an excuse.
posted by allegedly at 2:44 PM on September 26, 2018 [2 favorites]
The FHIR standard has built up quite a bit of traction in the healthcare tech world. Probably more than any previous effort. US healthcare providers and EMR vendors are having their arms gradually twisted by the "meaningful use" requirements and the incentives behind them.
Hospitals and health systems actually have a surprising amount of reason to embrace standardization - they can't even exchange data inside their own walls. A new generation of patients raised on the internet is not going to take "sorry, our fax machine is broken" as an excuse.
posted by allegedly at 2:44 PM on September 26, 2018 [2 favorites]
We have a friend who's an expert on US electronic medical record systems and gets invited to address conferences on interoperability. She thinks the US systems are already pretty good (they certainly sound better than Australia's) but her description of the underlying technology makes me think it's held together with tape and binding wire and will fall over as soon as Bobby Tables gets admitted.
posted by Joe in Australia at 4:21 PM on September 26, 2018
posted by Joe in Australia at 4:21 PM on September 26, 2018
What sort of systems do the various “socialist” nations use in their national systems? Couldn’t the US just swallow its collective pride and adopt one of those, rather than roll our own?
Some of the universal health care nations don't have systems at the national level. In Canada it is provincial. In England it was regional as well (authorities I believe). So they have some of the same complexities of different jurisdictions and different coverages.
What they don't have is the complexity of insurance companies' and hospitals' obfuscatory pricing schemes.
The UK's NHS universal health care record project has been a huge multibillion pound boondoggle.
posted by srboisvert at 4:52 PM on September 26, 2018
Some of the universal health care nations don't have systems at the national level. In Canada it is provincial. In England it was regional as well (authorities I believe). So they have some of the same complexities of different jurisdictions and different coverages.
What they don't have is the complexity of insurance companies' and hospitals' obfuscatory pricing schemes.
The UK's NHS universal health care record project has been a huge multibillion pound boondoggle.
posted by srboisvert at 4:52 PM on September 26, 2018
I have worked at the same hospital for my entire career. My department was one of the first to adopt electronic record keeping, back in the early 1990s. (This system, to be exact.) Before it was implemented, a number of us from the department spent an entire week at a nearby hotel with representatives from the vendor learning the system in great detail. It was met with grudging acceptance by most of us, but the big advantage that got people to actually like it was it's ability to automatically record vital signs from our monitors, a function that is especially useful when starting a complicated case where you may be putting in various vascular lines, nerve blocks, etc. and not having to try and reconstruct vital signs every 5 minutes on the anesthesia record was a great benefit. (As an aside from an old physician, not having to manually write down vital signs every 5 minutes can also make it easier to be less vigilant and not notice untoward trends as early as possible). But the biggest problem in implementing that record was getting all of our monitors from various manufacturers to output data to the record. There is no standard format for medical data, so each station had to be configured for the exact monitors it was hooked up to, sometimes with a great deal of trial and error; and not always a lot of cooperation from the monitor manufacturers regarding their proprietary communication protocols. Fast forward to 2016. We are upgrading to our 4th iteration of an electronic anesthesia record, this time one supplied by the same vendor as the EHR adopted by the whole hospital. This time however, training was limited to a few hours at most, much of which was not tailored to the unique needs of our department. I could go on about how the user interface is incredibly complex (over 100 clickable elements on any given screen, most of which I have no idea what they do), how the program has a habit of hanging up or slowing down at random times, or how every "improvement" to the EHR involves additional dialog boxes you must go through before actually doing anything useful (AKA "death by a thousand clicks"). But the real test came when we went live with the new system one Monday in October. There was the usual grumbling about having to do something new and different, but as soon as the operating rooms start, more panicked call started coming in. Flashback to 1992; none of the vital signs were being recorded on the record; this is a basic requirement for an anesthetic and if it isn't done we can't bill for our services. Not enough field testing of the new system had been done, and there were no vendor reps on site to troubleshoot (apparently that was an extra cost option, which makes about as much sense as having tires be an extra cost option on your new car.) Things were eventually worked out, but with great disruption to the operating room, which is normally one of the profitable parts of the hospital. And despite the vendor's assurance (Cerner, Epics biggest competitor) that we would be fully integrated into the hospital's EHR. that is not close to happening.
But that is not even my biggest frustration with our EHR and how it is implemented. One of the times I was most angry and frustrated at our medical informatics department was when I found out, completely by accident, how scanned in records from outside institutions were handled. For reasons that were never made clear to me, all those scanned in records were for years filed under the date January 1, 2005. And this was not widely publicized, so for years I wondered why I couldn't find outside records that had definitely been sent to us. What this meant in practice, is that if you wanted outside cardiology records from November 2012 for a patient born in August 2011, you had to look in the medical record several years before the patient was born to find what you needed. This has since been fixed (although not retroactively as far as I have seen), but who thought this was even remotely a good idea in the first place? It seems to me that without a strong outside influence (government) forcing all of the parties involved to make their systems communicate, then EHRs will remain just another source of frustration instead of the mechanism for expediting health care that they are being sold as.
posted by TedW at 5:58 PM on September 26, 2018 [7 favorites]
But that is not even my biggest frustration with our EHR and how it is implemented. One of the times I was most angry and frustrated at our medical informatics department was when I found out, completely by accident, how scanned in records from outside institutions were handled. For reasons that were never made clear to me, all those scanned in records were for years filed under the date January 1, 2005. And this was not widely publicized, so for years I wondered why I couldn't find outside records that had definitely been sent to us. What this meant in practice, is that if you wanted outside cardiology records from November 2012 for a patient born in August 2011, you had to look in the medical record several years before the patient was born to find what you needed. This has since been fixed (although not retroactively as far as I have seen), but who thought this was even remotely a good idea in the first place? It seems to me that without a strong outside influence (government) forcing all of the parties involved to make their systems communicate, then EHRs will remain just another source of frustration instead of the mechanism for expediting health care that they are being sold as.
posted by TedW at 5:58 PM on September 26, 2018 [7 favorites]
A good friend of mine used to work on a team that did data conversions for a company that sold hospital management software. If a hospital purchased his company's product, a member of his team would go work with that hospital, and write custom software to convert all the existing records to the new system.
The conversion was always a custom project, even if the hospital had previously used some other popular product. This is because nearly every hospital (or at least the larger ones) had managed to get some customizations added to their local installation. So even if they were previously using the well-known FizzBuzz Hospital Management System (FBHMS), what they were actually using was FBHMS with fifty local tweaks that affected the format of their medical records.
Based on my friend's stories, nearly every records conversion was a multi-month project. It was often complicated by the fact that the destination format was a moving target... because the hospital was trying to get some similar-but-different set of tweaks added to the new system, too.
posted by fencerjimmy at 7:43 PM on September 26, 2018 [5 favorites]
The conversion was always a custom project, even if the hospital had previously used some other popular product. This is because nearly every hospital (or at least the larger ones) had managed to get some customizations added to their local installation. So even if they were previously using the well-known FizzBuzz Hospital Management System (FBHMS), what they were actually using was FBHMS with fifty local tweaks that affected the format of their medical records.
Based on my friend's stories, nearly every records conversion was a multi-month project. It was often complicated by the fact that the destination format was a moving target... because the hospital was trying to get some similar-but-different set of tweaks added to the new system, too.
posted by fencerjimmy at 7:43 PM on September 26, 2018 [5 favorites]
"And to be honest, if you want this worse-is-better approach in your own life, today, you can do it by taking a cheap fax-to-email gateway service with a dedicated number, a Gmail rule that moves around or resends those incoming documents, and a document management system that performs OCR like Evernote. Have doctors fax all your records to the number (which they'll almost always do, particularly if you don't make it obvious that it's your personal number), then have Gmail label them and optionally stick them into Evernote where the PDF attachments will get OCRed and made searchable. "
I do this with my special needs child's medical records, and I don't actually have to resort to faxes; usually I just ask and they either give me a printout (that I then scan) or will e-mail it to me, and if it's in my hospital system, they have a web interface where I can go pull the records myself and download them. If I can't find what I need, I can e-mail the doctor's office staff from inside the web interface and ask them to send it by mail or e-mail. Sometimes I have to sign a form saying I agree to have a medical record transmitted insecurely over e-mail, but they send me a PDF, I put it into a folder on my desktop and then pop it into Evernote so I can search on the go from my phone and immediately e-mail relevant documents to other doctors/teachers/therapists/whatever.
I think it must be a much more common request these days because I used to get some guff about it, but now everyone acts like it's normal and usually sends me what I need in the format I want within 24 hours.
posted by Eyebrows McGee at 8:03 PM on September 26, 2018 [3 favorites]
I do this with my special needs child's medical records, and I don't actually have to resort to faxes; usually I just ask and they either give me a printout (that I then scan) or will e-mail it to me, and if it's in my hospital system, they have a web interface where I can go pull the records myself and download them. If I can't find what I need, I can e-mail the doctor's office staff from inside the web interface and ask them to send it by mail or e-mail. Sometimes I have to sign a form saying I agree to have a medical record transmitted insecurely over e-mail, but they send me a PDF, I put it into a folder on my desktop and then pop it into Evernote so I can search on the go from my phone and immediately e-mail relevant documents to other doctors/teachers/therapists/whatever.
I think it must be a much more common request these days because I used to get some guff about it, but now everyone acts like it's normal and usually sends me what I need in the format I want within 24 hours.
posted by Eyebrows McGee at 8:03 PM on September 26, 2018 [3 favorites]
Going to a new eye doctor this AM and they emailed a pdf of a form that I had to print out and then fill out in pen because they apparently don't have a way for me to fill out my demographics online or at a station in the office. So someone is going to have to sit there and re-type in all my information.
posted by octothorpe at 5:17 AM on September 27, 2018 [1 favorite]
posted by octothorpe at 5:17 AM on September 27, 2018 [1 favorite]
I did medicaid case management for a little while during a transition of MCO's (basically one state contract ended and another took ove) r the care of thousands medicaid patients.
It was a clusterfuck of epic perportions. I worked with highly disabled adults that recieved home based services and the amount of data we recieved was almost none. We barely knew who we were paying for services for, much less their medical conditions, needs, and previously authorized services. We basically just started over and blanket authed until we could review.
It took almost a year.
posted by AlexiaSky at 6:02 AM on September 27, 2018 [1 favorite]
It was a clusterfuck of epic perportions. I worked with highly disabled adults that recieved home based services and the amount of data we recieved was almost none. We barely knew who we were paying for services for, much less their medical conditions, needs, and previously authorized services. We basically just started over and blanket authed until we could review.
It took almost a year.
posted by AlexiaSky at 6:02 AM on September 27, 2018 [1 favorite]
And despite the vendor's assurance (Cerner, Epics biggest competitor) that we would be fully integrated into the hospital's EHR. that is not close to happening.
Most of my days are punctuated by someone grumbling about Cerner, it's an absolutely rubbish system. And their Wikipedia page is...interesting. https://en.wikipedia.org/wiki/Cerner#Controversy
posted by elsietheeel at 7:29 AM on September 27, 2018 [2 favorites]
Most of my days are punctuated by someone grumbling about Cerner, it's an absolutely rubbish system. And their Wikipedia page is...interesting. https://en.wikipedia.org/wiki/Cerner#Controversy
posted by elsietheeel at 7:29 AM on September 27, 2018 [2 favorites]
Yeah it's probably possible to forgo the fax-to-email stuff today, but honestly I prefer using fax as the "universal medial API" than deal with a whole mess of two-bit physicians' office proprietary web portals. If it's a choice between sending a really officious REQUEST FOR RECORDS TRANSMITTAL fax, or creating yet another goddamn login to a shitty EMR portal, I'm just gonna send that fax. It takes less time to just go into Word and change the name of the office and the inside address and slam it into the outbound fax queue than it does to deal with most of those crap web interfaces. YMMV, of course, and if you visit the same doctors multiple times, I could see that being preferable. But I've never managed to go to the same doctor more than a handful of times before changing insurance and inevitably going somewhere else.
(The last time I tried to play along and use one, after the usual account-creation fuss, I had to download some ancient version of Java, enable the Java browser plugin, and then turn all of my browser's security settings down to the computer-to-computer equivalent of bareback sex with a stranger in a dark alley. All to view TIFFs, but you could only get them through this grotty plugin, or a similarly hideous IE-only ActiveX thing. In 2017, for a major regional hospital chain.)
posted by Kadin2048 at 11:17 AM on September 27, 2018 [1 favorite]
(The last time I tried to play along and use one, after the usual account-creation fuss, I had to download some ancient version of Java, enable the Java browser plugin, and then turn all of my browser's security settings down to the computer-to-computer equivalent of bareback sex with a stranger in a dark alley. All to view TIFFs, but you could only get them through this grotty plugin, or a similarly hideous IE-only ActiveX thing. In 2017, for a major regional hospital chain.)
posted by Kadin2048 at 11:17 AM on September 27, 2018 [1 favorite]
Previously, regarding EPIC (and its culture, as well as MUMPS): Blocking health records for fun and profit.
...
they just punted on the whole question of glyphs. They used a 7 pixel by 7 pixel grid for each letter, and transmitted what we'd now call a raster down the wire.
The analogous solution for EMR might be to stop trying to encode everything down to the field level, and instead just do documents. Most medical records are still reducible to "pages" and "documents" or some similar level of categorization.
We have this, it's called PDF and that would be fine if the only point of going to EMRs were the various benefits of paperlessness. But it's not. Hell, this is ultimately what fax transmission is: a bitmap of a physical document's image.
posted by snuffleupagus at 6:20 AM on September 29, 2018 [1 favorite]
...
they just punted on the whole question of glyphs. They used a 7 pixel by 7 pixel grid for each letter, and transmitted what we'd now call a raster down the wire.
The analogous solution for EMR might be to stop trying to encode everything down to the field level, and instead just do documents. Most medical records are still reducible to "pages" and "documents" or some similar level of categorization.
We have this, it's called PDF and that would be fine if the only point of going to EMRs were the various benefits of paperlessness. But it's not. Hell, this is ultimately what fax transmission is: a bitmap of a physical document's image.
posted by snuffleupagus at 6:20 AM on September 29, 2018 [1 favorite]
« Older The Newcomer | “I want to rewrite history.” Newer »
This thread has been archived and is closed to new comments
posted by scruss at 10:24 AM on September 26, 2018 [3 favorites]