10 ms·

I'm currently working on my thesis, I'm trying to design a radically different take on EMR/EHR systems - http://barnett.surge.sh/ http://barnett.surge.sh/ I've

I'm currently working on my thesis, I'm trying to design a radically different take on EMR/EHR systems - http://barnett.surge.sh/ http://barnett.surge.sh/ I've come to realize healthcare software is a significantly trickier problem than most realize. Not in terms of technical possiblity, but other factors. Building healthcare software is really hard - which seems like such a paradox, because surely ensuring highly trained individuals have half decent tools would be a huge goal for society - but almost all healthcare software is still frankly, terrible. After heaps of reading I've narrowed it down to 3 reasons why healthcare software is so hard to build. 1 - Modern healthcare is complex, and varied. Building any healthcare softawre is no easy task, even simple systems have many other systems to integrate with. But what makes this worse, is that no two places do healthcare quite the same. Between regions, and institutions, healthcare can vary a shit ton. 2 - Healthcare is risky. Sounds obvious, but if your software fails, people die - but this risk creates an attitude of "dont touch it". Changing systems has such a risk, that institutions will use the same software until it's so out of date its not funny. This impeeds improvement and innovation. 3 - Healthcare software is hard to sell. Not in the sense that its unprofitable, but that it takes ages to get users. Say if I made video editing software - I could get professionals trying it out tommorrow. However, if I make a PMS, I have to sell it to the entire practice - not just a few keen doctors. From making contact to actually getting a small practice using your software could literally be years - making it super hard to enter the market. I've been working on this topic fulltime for a year now - and I must admit, I don't have the golden answer. However, I'm designing a system as a response to these factors, an alternative to interoperability, and would love to talk anybody in the industry about it. http://barnett.surge.sh/ http://barnett.surge.sh/ eliot.slevin@gmail.com

dharma1 10y ago
looks good, how far are you with it? is it open source?
- JusticeJuice 10y ago
  It's essentially a giant design proposal - yes it's all open source, but theres nothing you can download and start using - yet :)
doughj3 10y ago
Your homepage there has a heavy focus on health records. Have you considered building upon the FHIR standard?
- JusticeJuice 10y ago
  Yes, thats the main focus of the project. One of my critiques of healthcare systems is they do wayyy to much. Eg a practice management system (pms), handles appointments, records, invoicing, staff management, inventory, decision support - so much stuff. I'm instead focusing on just health records - and the interfaces medical professionals use to work with them. I've had a good look at FHIR and many other standards, but FHIR is built for an interoperability model - eg, your record is sent back and forth between your gp, hospital, specialist, labs, etc. I'm instead focusing on a central access model - so all of the parties interested in your record are all accessing one central point, so I'm doing something a bit different, it almost looks like NPM. That being said, I'm all for any standard becoming the industry standard - I don't care what it is, anything is better than nothing. So I wish FHIR the best of luck :) http://standardhealthrecord.org/ http://standardhealthrecord.org/ is also a standard I really like
  - numbsafari 10y ago
    Have you considered, in your model, how your system would work if a customer wanted to integrate with, as an example, a machine learning platform that attempts to predict the likelihood of a patient's readmission given their current status?
    - JusticeJuice 10y ago
      Yeah, I have. A health record is a collaboration - unlike a medical record, which just shows a slice of your health from the perspective of a single institution, a health record aims to show as many facets as possible. An EHR system should be agnostic about who access the record, and what they use it for - obviously gps, specialists, hosptitals, labs, the patient should all have access - but what about, as your said, software? The ideal system would be,patient visits your system, goes through an OAuth flow, and your system has api access. The tricky part is format. Its really hard to define a format which fits the entire world's use case. If you consider, what is healthcare information - you realize it's any data relevant to the patients health. So how can you even start to define that? My compromise is a system where formats are promoted to become popular, regionally, and condition specific. So if you're building a machine learning readmission software, you can target a specific market - say British Primary care, or a specific condition, say diabetes. You can expand out to other markets by writing transformers to translate from one regional format to another.
fny 10y ago
As far as I can tell, there's no economic incentive for health care systems to undergo these transformations willingly. Remember, the digital transformation didn't occur until the HITECH act in 2009, which mandated the use of electronic health records. Why digitize your operations and worry about maintaining tech infrastructure when you can do everything on paper with a handful of secretaries? You could argue providing better data-driven healthcare to patients, but guess what--most health systems don't care because they make money every time you get sick.[0] With HITECH, the government forced them to transition to EHRs, and here's what hospital systems optimized for... 1. Reliability 2. Optimizing billing 3. Guaranteeing compliance with different protocols, particularly those related to billing Openess? Interoperability? They don't care. You're piling on more technical and administrative costs without the promise of future returns. > ensuring highly trained individuals have half decent tools would be a huge goal for society - but almost all healthcare software is still frankly, terrible. You know who hates EHRs the most? Doctors. Do you know who loves them? Hospital administrators. Guess who pays for the EHR? ;) [0]: An exception to this is Kaiser Permanente. They insure the patients for which they care, so they're incetivized to improve health to reduce costs over time.
- JusticeJuice 10y ago
  > As far as I can tell, there's no economic incentive for health care systems to undergo these transformations willingly Yes. Thing is, say you have shitty software and see 20 patients in one day. Say you have fantastic software - you'll probably still only see 20 patients a day. The cost of upgrading is usually not justified if it doesn't hit the bottom line. I'm trying to tackle this by making a system where improving it isn't a huge risk, or cost. For example, it has an extensible user interface - imagine something like atom text editor, or even browser extensions. Say you look at the immunization screen all day, and hate it. Find / build a small extension to arange it how you want, without changing any of the back-end system, and then share it to a registry for other people in your situation. Easy win. I'm actually designing this system within the context of primary care in New Zealand. We have a socialized healthcare system, which is honestly very effective. Everybody pays for it via tax, and resources are evenly distributed. We even bulk purchase generic drugs for the entire country. There isn't so much of an evil incentive to ensure all their 'customers' stay sick, but instead funding is a bit hard to come by. Funding happens regionally, and building national healthcare software systems is a huge task nobody but the government is willing to try. The market is so small (population of 4mill) that vendors can usually make more money elsewhere - so getting NZ tailored solutions is rare. This is why I think the opportunity is in making a crowdsourced system, as opposed to a 'product' to sell.
zupreme 10y ago
So my startup hosts and supports a decent number of EHR/EMR solutions and I even seriously entered talks at one point for us to partner up with a large medical conglomerate to write a new platform. That idea died for the same two reasons most similar initiatives in healthcare die: Government and Insurance (called "Payers" in the healthcare industry). You can write the absolute best EMR & PM (Practice Management) platform on the planet but if you can't get the payers to integrate with it you're dead in the water. The only practices who would even consider you without those integrations are ones not currently using an EMR, which means that your first hurdle with them has to be convincing them to go electronic - which means you own all the logistical hurdles associated with a large-scale chart-scanning operation, user training, mindset shifting, and much more. And don't even get me started on the barriers to entry the Department of Health and Human Services have put in place....
- JusticeJuice 10y ago
  Interesting. I think there's a real argument that an integration platform for healthcare should be a government run entity - ensuring that all parties have a chance. I'm curious about these barriers you speak of - I'm based in New Zealand, and I must admit I haven't run into many government based barriers to entry - just barriers created by privatized companies.
  - shaftoe 10y ago
    I work in healthcare. A lot of the complexity is FROM government attempts to help. Look at HIPAA. Originally about medical record portability, now it largely deals with security, privacy, and availability. These sound good, but it is frustratingly vague, subject to massive reactive fines, and assumes a lot of practices that always feel out of date. It's way more of a pain than private compliance frameworks like PCI. The problems in US healthcare are complex, entrenched, and Brook no simple fix. The government is incredibly buearacratic, designed by lawyers who value an openness to interpretation by courts that creates ambiguity, and creates solutions in their own mold.
- nradov 10y ago
  All of the major payers have standard interfaces for submitting electronic claims now. The payers aren't a barrier to launching a new EHR.
uzoodoo 10y ago
Minor thing — your page title is "Page Title"
yourapostasy 10y ago
I strongly suspect that for the US healthcare system setting you are talking about, the incentives are misaligned for building a better EMR/EHR system. What we endure today satisfies those who profit mightily from it just fine, and improvements in health outcomes represent profit erosion, hence no true incentives to truly improve. Could there be more room for improvements building from the ground up in alternative healthcare environments, like co-ops, hospitals where billing is in cash only and fees for standard procedures posted up-front, concierge, etc.? I don't think the mainstream US healthcare industry will change for the better unless it is presented with an immediate (<6 months), concrete, existential threat to its existence as we know it today. Considering how unfathomably unlikely that is, then substantive improvements in health outcomes could lay only "off the grid" of the current system. Going down that path is absolutely more risky and less remunerative. Your post gives me the impression that while compensation ranks for you, I'm-On-A-Boat FU money at the expense of poor health outcomes for those you serve might not rank, so I thought I might throw out some outlier ideas and see what discussion falls out of that, if any. To give an idea of how misaligned the US system is, this blog post [1] from my reading list today came up --- but to pick one at random out of a 3-second Google would be pretty easy --- describes how in the US a $5K MRI tab is not unusual. You can fly to NRT, pay in cash in Japan for an MRI and consult, fly back, about four times for that much. Cray cray. [1] https://market-ticker.org/akcs-www?post=231917 https://market-ticker.org/akcs-www?post=231917
- JusticeJuice 10y ago
  Just to be clear, I'm actually designing this system with New Zealand in mind, which has a very socialised healthcare system. I've got no idea how to help the american healthcare system
jldugger 10y ago
There's also some structural reasons. Why is it that larger practices don't write their own software? Well, maybe a few do, but: a) nearly all states require that ownership in medical practice be restricted to those licensed to practice. This reduces access to capital, and more or less ensures that practices remain small businesses. b) nearly all care in the US is fragmented across specialists in different practices. c) healthcare in the US is largely paid for by parties fairly disinterested in actually transacting. While most folks on HN might A/B test UI improvements looking at click through rates, the implicit idea in insurance billing software is that submitting claims is someone else's expense to deal with. When you look at places where there is a scale to medical practice, opportunity for software opens up. VA and DoD both operate and staff their own health care facilities, and have software systems they've implemented. But even DoD has problems receiving and retaining records from outside care providers (if Wikipedia is to be believed). If there were fewer, larger medical practices in the US, it would both be easier to agree on standardized data sharing protocols, and more likely for patients to see multiple doctors working for the same employer, thus never necessitating a data transfer. Admittedly, it'd be harder entrepreneurs like yourself to break in. Perhaps you'd be an employee or Engineering manager working on the problem instead. Or perhaps you'd be selling a plugin to a larger EMR.
- yellowapple 10y ago
  In my observation, I've found that there are quite a few larger healthcare providers, and that even the "smaller" hospitals are pretty big in terms of IT needs. In my hometown alone, there's Kaiser Permanente, Sutter Health, and UC Davis. Sure, none of them are quite nationwide, but they're by no means "small", either.
  - jldugger 10y ago
    Part of the challenge is that hospitals represent only one portion of the care chain. My mom's GP, for instance, has never used a hospital as their office. There are dozens of hospitals within 30 minutes driving distance, of which few are owned by the same company / non-profit. KP is in the vein of what I'm suggesting though, just wish it was more widely available.
fredsmith42 10y ago
The EMR market is consolidating on two big vendors: Epic and Cerner. Depending on how you measure it (Hospitals on a system or Medical Records under management), either Cerner or Epic is #1 and the other is #2. At this point, in the U.S., it is almost impossible for a newcomer to break into the market. There are many companies out there slowly losing market share to Epic or Cerner. Allscripts, Nextgen, McKesson/HBOC, GE, and others. Some are big, well-funded companies, others are more like startups with OK funding. Most of them are falling by the wayside, casualties to the bitter Epic/Cerner market war. The HITECH act really exacerbated this point by accelerating spending on EMR's since 2008. Meaningful Use meant that a lot of homegrown systems had to be replaced, simply because hospitals couldn't keep up with the regulation and complaince. Likewise, smaller companies had to spend more money on compliance than adding features and also faded. Sure, there are plenty of Allscripts and HBOC customers out there, but their number is shrinking, not growing. In a way it's a shame, because the two biggest vendors each have critical flaws. Cerner has a decent store in Oracle, but is hobbled by reliance on its own homegrown language for integration and middleware. The language is CCL and is like a cross between PL/SQL and TCL with HL7 Domain knowledge thrown in. Epic has an ancient store with Caché, which is really MUMPS with some OO and relational hooks tossed in. Cerner excels at customization, while Epic excels at disciplined project implementation. You can browbeat Cerner into giving you what you want, but you can implement Epic pretty much close to on time and close to budget with an impressive reliability. Neither system really wants to interoperate with the other. It can be done, but it takes a months-long project with specialists writing HL7 interfaces to pass demographic, lab, and document data between each system. And this interoperability, once established, doesn't really help the patient. Docs can see orders entered in Cerner Powerchart displayed in Epic Hyperspace, but the patient doesn't necessarily get access to either the order or the result unless it goes to either system's portal. Finally, you have to realize that you are dealing with a very powerful and conservative management culture. Many Silicon Valley startups with whizbang solutions have foundered when they run up against the entrenched incumbents. Well-funded companies with highly competent engineers have failed because they don't have the patience to spend a couple of decades building up a user base. Investors realize that they can get more bang for their buck elsewhere and leave the game. If you even want to make a dent, you have to be willing to spend at least 15 years and untold amounts of money just trying to take on the big guys. If you're going against a market with relatively short turnarounds, like a calendaring app or even a game console, you have a shot. If you're going against a market that is very conservative and waits at least a decade before reevaluating installed systems, you have a very tough row to hoe. Ultimately, at least in the U.S., I think that implementing a new EMR is a sucker's game. You can spend millions and lose. In New Zealand, the market is different at this point, but how long can they resist the network effects of the big vendors? Cerner and Epic are currently battling it out in Australia, and will exert pressure in NZ once that market is settled. I know for a fact that Cerner went after NZ in the late 90's but pulled back after it became clear that entrenched corruption in the NZ procurement system would determine the winner. I don't have names and places for you, but I was in Sydney at the time and remember the disappointment of the executive team when they realized that NZ at that point was not a genuinely open market. You have mentioned in another post that you are concentrating just on records, and avoiding the extra complexity of the other systems in terms of patient management, practice management, supply management, etc… With all due respect, I think you are being naïve. A modern EMR system has to take into account the entire needs of the practice and how to manage it. When a health care practice is spending a buck, they want to get as much out of that dollar as possible. If vendor A is offering "just records", but vendor B is offering records plus practice management, they will go with vendor B. Every time. In any case, I wish you the best of luck. I have a lot of experience in this sector and would be happy to share my knowledge with you.
Steeeve 10y ago
EHR is not difficult. There's a lot to the required data model, but building a working system doesn't require mental gymnastics, just time and effort. Selling isn't particularly challenging either. There is a tried and true model for selling to medical practices and hospitals that works. Customer acquisition is expensive and resource intensive, but it's doable with the right team and right investment for any product. There is risk, but HIPAA is a lot more of a risk than a logical failure. With every HIPAA violation, there are fingers pointing in every direction, and that means that resources are required to respond to everything - even when it's painfully obvious the software isn't at fault. I would say that the biggest challenge to creating a market for a new EHR product is that there are generations of people resistant to any real user interface changes. I'm not even talking about adding tabs or checkboxes. If an EHR product logically maps to one of the legacy powerhouses like EPIC, it makes sense to people. If it logically maps to health care instead - No dice. People are entrenched in a system that was designed by engineers for health care because that's all that was available for a generation. They've gone to classes to learn how to work in that environment, they've worked in it for years, and it works for them. A major shift of any sort means retraining doctors, nurses, call center operators... anyone and everyone up and down the food chain. Even in a small practice that's a lot of time and energy to invest. There is resistance every step of the way.
brady_m 10y ago
The combination of OpenEMR being open source, open community, volunteer driven, physician driven and getting significant use in the real world makes it a perfect vehicle to glean really useful insights into EMR use and I highly recommend looking into it in your research endeavors. For example, in subjects such as EMR system development and the effects of regulation. A really nice and innovative example on EMR system development is the development of the Eye module released in OpenEMR's most recent version 5.0. A practicing ophthalmologist, Dr. Ray Magauran, decided to develop a ophthalmology module in OpenEMR with the goal of actually improving efficiency over paper (as a physician, I would confidently say I have never used an EMR that is faster than paper). What this physician/developer did was simply amazing and the details of this eye module(including the makings of it) can be found here: http://www.open-emr.org/wiki/index.php/Eye_Exam http://www.open-emr.org/wiki/index.php/Eye_Exam What I think is innovative here is that this was designed, built, tested, and used by a practicing physician on the front lines. Then on the flip side can also learn about the dichotomy of fulfilling regulations versus developing an innovative and efficient EMR system, which OpenEMR provides a ring side seat too since the road to complete meaningful use certification was done the open source way(ie. in full public view).
ThomPete 10y ago
The primary reason why healthcare is hard is legal and politics more than anything else. I think a lot could be solved if someone decided to build the healthcare version of Zapier.

dharma1 10y ago

looks good, how far are you with it? is it open source?

JusticeJuice 10y ago
It's essentially a giant design proposal - yes it's all open source, but theres nothing you can download and start using - yet :)

doughj3 10y ago

Your homepage there has a heavy focus on health records. Have you considered building upon the FHIR standard?

JusticeJuice 10y ago
Yes, thats the main focus of the project. One of my critiques of healthcare systems is they do wayyy to much. Eg a practice management system (pms), handles appointments, records, invoicing, staff management, inventory, decision support - so much stuff. I'm instead focusing on just health records - and the interfaces medical professionals use to work with them. I've had a good look at FHIR and many other standards, but FHIR is built for an interoperability model - eg, your record is sent back and forth between your gp, hospital, specialist, labs, etc. I'm instead focusing on a central access model - so all of the parties interested in your record are all accessing one central point, so I'm doing something a bit different, it almost looks like NPM. That being said, I'm all for any standard becoming the industry standard - I don't care what it is, anything is better than nothing. So I wish FHIR the best of luck :) http://standardhealthrecord.org/ http://standardhealthrecord.org/ is also a standard I really like
- numbsafari 10y ago
  Have you considered, in your model, how your system would work if a customer wanted to integrate with, as an example, a machine learning platform that attempts to predict the likelihood of a patient's readmission given their current status?
  - JusticeJuice 10y ago
    Yeah, I have. A health record is a collaboration - unlike a medical record, which just shows a slice of your health from the perspective of a single institution, a health record aims to show as many facets as possible. An EHR system should be agnostic about who access the record, and what they use it for - obviously gps, specialists, hosptitals, labs, the patient should all have access - but what about, as your said, software? The ideal system would be,patient visits your system, goes through an OAuth flow, and your system has api access. The tricky part is format. Its really hard to define a format which fits the entire world's use case. If you consider, what is healthcare information - you realize it's any data relevant to the patients health. So how can you even start to define that? My compromise is a system where formats are promoted to become popular, regionally, and condition specific. So if you're building a machine learning readmission software, you can target a specific market - say British Primary care, or a specific condition, say diabetes. You can expand out to other markets by writing transformers to translate from one regional format to another.

fny 10y ago

As far as I can tell, there's no economic incentive for health care systems to undergo these transformations willingly. Remember, the digital transformation didn't occur until the HITECH act in 2009, which mandated the use of electronic health records. Why digitize your operations and worry about maintaining tech infrastructure when you can do everything on paper with a handful of secretaries? You could argue providing better data-driven healthcare to patients, but guess what--most health systems don't care because they make money every time you get sick.[0] With HITECH, the government forced them to transition to EHRs, and here's what hospital systems optimized for... 1. Reliability 2. Optimizing billing 3. Guaranteeing compliance with different protocols, particularly those related to billing Openess? Interoperability? They don't care. You're piling on more technical and administrative costs without the promise of future returns. > ensuring highly trained individuals have half decent tools would be a huge goal for society - but almost all healthcare software is still frankly, terrible. You know who hates EHRs the most? Doctors. Do you know who loves them? Hospital administrators. Guess who pays for the EHR? ;) [0]: An exception to this is Kaiser Permanente. They insure the patients for which they care, so they're incetivized to improve health to reduce costs over time.

JusticeJuice 10y ago
> As far as I can tell, there's no economic incentive for health care systems to undergo these transformations willingly Yes. Thing is, say you have shitty software and see 20 patients in one day. Say you have fantastic software - you'll probably still only see 20 patients a day. The cost of upgrading is usually not justified if it doesn't hit the bottom line. I'm trying to tackle this by making a system where improving it isn't a huge risk, or cost. For example, it has an extensible user interface - imagine something like atom text editor, or even browser extensions. Say you look at the immunization screen all day, and hate it. Find / build a small extension to arange it how you want, without changing any of the back-end system, and then share it to a registry for other people in your situation. Easy win. I'm actually designing this system within the context of primary care in New Zealand. We have a socialized healthcare system, which is honestly very effective. Everybody pays for it via tax, and resources are evenly distributed. We even bulk purchase generic drugs for the entire country. There isn't so much of an evil incentive to ensure all their 'customers' stay sick, but instead funding is a bit hard to come by. Funding happens regionally, and building national healthcare software systems is a huge task nobody but the government is willing to try. The market is so small (population of 4mill) that vendors can usually make more money elsewhere - so getting NZ tailored solutions is rare. This is why I think the opportunity is in making a crowdsourced system, as opposed to a 'product' to sell.

zupreme 10y ago

So my startup hosts and supports a decent number of EHR/EMR solutions and I even seriously entered talks at one point for us to partner up with a large medical conglomerate to write a new platform. That idea died for the same two reasons most similar initiatives in healthcare die: Government and Insurance (called "Payers" in the healthcare industry). You can write the absolute best EMR & PM (Practice Management) platform on the planet but if you can't get the payers to integrate with it you're dead in the water. The only practices who would even consider you without those integrations are ones not currently using an EMR, which means that your first hurdle with them has to be convincing them to go electronic - which means you own all the logistical hurdles associated with a large-scale chart-scanning operation, user training, mindset shifting, and much more. And don't even get me started on the barriers to entry the Department of Health and Human Services have put in place....

JusticeJuice 10y ago
Interesting. I think there's a real argument that an integration platform for healthcare should be a government run entity - ensuring that all parties have a chance. I'm curious about these barriers you speak of - I'm based in New Zealand, and I must admit I haven't run into many government based barriers to entry - just barriers created by privatized companies.
- shaftoe 10y ago
  I work in healthcare. A lot of the complexity is FROM government attempts to help. Look at HIPAA. Originally about medical record portability, now it largely deals with security, privacy, and availability. These sound good, but it is frustratingly vague, subject to massive reactive fines, and assumes a lot of practices that always feel out of date. It's way more of a pain than private compliance frameworks like PCI. The problems in US healthcare are complex, entrenched, and Brook no simple fix. The government is incredibly buearacratic, designed by lawyers who value an openness to interpretation by courts that creates ambiguity, and creates solutions in their own mold.
nradov 10y ago
All of the major payers have standard interfaces for submitting electronic claims now. The payers aren't a barrier to launching a new EHR.

uzoodoo 10y ago

Minor thing — your page title is "Page Title"

yourapostasy 10y ago

I strongly suspect that for the US healthcare system setting you are talking about, the incentives are misaligned for building a better EMR/EHR system. What we endure today satisfies those who profit mightily from it just fine, and improvements in health outcomes represent profit erosion, hence no true incentives to truly improve. Could there be more room for improvements building from the ground up in alternative healthcare environments, like co-ops, hospitals where billing is in cash only and fees for standard procedures posted up-front, concierge, etc.? I don't think the mainstream US healthcare industry will change for the better unless it is presented with an immediate (<6 months), concrete, existential threat to its existence as we know it today. Considering how unfathomably unlikely that is, then substantive improvements in health outcomes could lay only "off the grid" of the current system. Going down that path is absolutely more risky and less remunerative. Your post gives me the impression that while compensation ranks for you, I'm-On-A-Boat FU money at the expense of poor health outcomes for those you serve might not rank, so I thought I might throw out some outlier ideas and see what discussion falls out of that, if any. To give an idea of how misaligned the US system is, this blog post [1] from my reading list today came up --- but to pick one at random out of a 3-second Google would be pretty easy --- describes how in the US a $5K MRI tab is not unusual. You can fly to NRT, pay in cash in Japan for an MRI and consult, fly back, about four times for that much. Cray cray. [1] https://market-ticker.org/akcs-www?post=231917 https://market-ticker.org/akcs-www?post=231917

JusticeJuice 10y ago
Just to be clear, I'm actually designing this system with New Zealand in mind, which has a very socialised healthcare system. I've got no idea how to help the american healthcare system

jldugger 10y ago

There's also some structural reasons. Why is it that larger practices don't write their own software? Well, maybe a few do, but: a) nearly all states require that ownership in medical practice be restricted to those licensed to practice. This reduces access to capital, and more or less ensures that practices remain small businesses. b) nearly all care in the US is fragmented across specialists in different practices. c) healthcare in the US is largely paid for by parties fairly disinterested in actually transacting. While most folks on HN might A/B test UI improvements looking at click through rates, the implicit idea in insurance billing software is that submitting claims is someone else's expense to deal with. When you look at places where there is a scale to medical practice, opportunity for software opens up. VA and DoD both operate and staff their own health care facilities, and have software systems they've implemented. But even DoD has problems receiving and retaining records from outside care providers (if Wikipedia is to be believed). If there were fewer, larger medical practices in the US, it would both be easier to agree on standardized data sharing protocols, and more likely for patients to see multiple doctors working for the same employer, thus never necessitating a data transfer. Admittedly, it'd be harder entrepreneurs like yourself to break in. Perhaps you'd be an employee or Engineering manager working on the problem instead. Or perhaps you'd be selling a plugin to a larger EMR.

yellowapple 10y ago
In my observation, I've found that there are quite a few larger healthcare providers, and that even the "smaller" hospitals are pretty big in terms of IT needs. In my hometown alone, there's Kaiser Permanente, Sutter Health, and UC Davis. Sure, none of them are quite nationwide, but they're by no means "small", either.
- jldugger 10y ago
  Part of the challenge is that hospitals represent only one portion of the care chain. My mom's GP, for instance, has never used a hospital as their office. There are dozens of hospitals within 30 minutes driving distance, of which few are owned by the same company / non-profit. KP is in the vein of what I'm suggesting though, just wish it was more widely available.

fredsmith42 10y ago

The EMR market is consolidating on two big vendors: Epic and Cerner. Depending on how you measure it (Hospitals on a system or Medical Records under management), either Cerner or Epic is #1 and the other is #2. At this point, in the U.S., it is almost impossible for a newcomer to break into the market. There are many companies out there slowly losing market share to Epic or Cerner. Allscripts, Nextgen, McKesson/HBOC, GE, and others. Some are big, well-funded companies, others are more like startups with OK funding. Most of them are falling by the wayside, casualties to the bitter Epic/Cerner market war. The HITECH act really exacerbated this point by accelerating spending on EMR's since 2008. Meaningful Use meant that a lot of homegrown systems had to be replaced, simply because hospitals couldn't keep up with the regulation and complaince. Likewise, smaller companies had to spend more money on compliance than adding features and also faded. Sure, there are plenty of Allscripts and HBOC customers out there, but their number is shrinking, not growing. In a way it's a shame, because the two biggest vendors each have critical flaws. Cerner has a decent store in Oracle, but is hobbled by reliance on its own homegrown language for integration and middleware. The language is CCL and is like a cross between PL/SQL and TCL with HL7 Domain knowledge thrown in. Epic has an ancient store with Caché, which is really MUMPS with some OO and relational hooks tossed in. Cerner excels at customization, while Epic excels at disciplined project implementation. You can browbeat Cerner into giving you what you want, but you can implement Epic pretty much close to on time and close to budget with an impressive reliability. Neither system really wants to interoperate with the other. It can be done, but it takes a months-long project with specialists writing HL7 interfaces to pass demographic, lab, and document data between each system. And this interoperability, once established, doesn't really help the patient. Docs can see orders entered in Cerner Powerchart displayed in Epic Hyperspace, but the patient doesn't necessarily get access to either the order or the result unless it goes to either system's portal. Finally, you have to realize that you are dealing with a very powerful and conservative management culture. Many Silicon Valley startups with whizbang solutions have foundered when they run up against the entrenched incumbents. Well-funded companies with highly competent engineers have failed because they don't have the patience to spend a couple of decades building up a user base. Investors realize that they can get more bang for their buck elsewhere and leave the game. If you even want to make a dent, you have to be willing to spend at least 15 years and untold amounts of money just trying to take on the big guys. If you're going against a market with relatively short turnarounds, like a calendaring app or even a game console, you have a shot. If you're going against a market that is very conservative and waits at least a decade before reevaluating installed systems, you have a very tough row to hoe. Ultimately, at least in the U.S., I think that implementing a new EMR is a sucker's game. You can spend millions and lose. In New Zealand, the market is different at this point, but how long can they resist the network effects of the big vendors? Cerner and Epic are currently battling it out in Australia, and will exert pressure in NZ once that market is settled. I know for a fact that Cerner went after NZ in the late 90's but pulled back after it became clear that entrenched corruption in the NZ procurement system would determine the winner. I don't have names and places for you, but I was in Sydney at the time and remember the disappointment of the executive team when they realized that NZ at that point was not a genuinely open market. You have mentioned in another post that you are concentrating just on records, and avoiding the extra complexity of the other systems in terms of patient management, practice management, supply management, etc… With all due respect, I think you are being naïve. A modern EMR system has to take into account the entire needs of the practice and how to manage it. When a health care practice is spending a buck, they want to get as much out of that dollar as possible. If vendor A is offering "just records", but vendor B is offering records plus practice management, they will go with vendor B. Every time. In any case, I wish you the best of luck. I have a lot of experience in this sector and would be happy to share my knowledge with you.

Steeeve 10y ago

EHR is not difficult. There's a lot to the required data model, but building a working system doesn't require mental gymnastics, just time and effort. Selling isn't particularly challenging either. There is a tried and true model for selling to medical practices and hospitals that works. Customer acquisition is expensive and resource intensive, but it's doable with the right team and right investment for any product. There is risk, but HIPAA is a lot more of a risk than a logical failure. With every HIPAA violation, there are fingers pointing in every direction, and that means that resources are required to respond to everything - even when it's painfully obvious the software isn't at fault. I would say that the biggest challenge to creating a market for a new EHR product is that there are generations of people resistant to any real user interface changes. I'm not even talking about adding tabs or checkboxes. If an EHR product logically maps to one of the legacy powerhouses like EPIC, it makes sense to people. If it logically maps to health care instead - No dice. People are entrenched in a system that was designed by engineers for health care because that's all that was available for a generation. They've gone to classes to learn how to work in that environment, they've worked in it for years, and it works for them. A major shift of any sort means retraining doctors, nurses, call center operators... anyone and everyone up and down the food chain. Even in a small practice that's a lot of time and energy to invest. There is resistance every step of the way.

brady_m 10y ago

The combination of OpenEMR being open source, open community, volunteer driven, physician driven and getting significant use in the real world makes it a perfect vehicle to glean really useful insights into EMR use and I highly recommend looking into it in your research endeavors. For example, in subjects such as EMR system development and the effects of regulation. A really nice and innovative example on EMR system development is the development of the Eye module released in OpenEMR's most recent version 5.0. A practicing ophthalmologist, Dr. Ray Magauran, decided to develop a ophthalmology module in OpenEMR with the goal of actually improving efficiency over paper (as a physician, I would confidently say I have never used an EMR that is faster than paper). What this physician/developer did was simply amazing and the details of this eye module(including the makings of it) can be found here: http://www.open-emr.org/wiki/index.php/Eye_Exam http://www.open-emr.org/wiki/index.php/Eye_Exam What I think is innovative here is that this was designed, built, tested, and used by a practicing physician on the front lines. Then on the flip side can also learn about the dichotomy of fulfilling regulations versus developing an innovative and efficient EMR system, which OpenEMR provides a ring side seat too since the road to complete meaningful use certification was done the open source way(ie. in full public view).

ThomPete 10y ago

The primary reason why healthcare is hard is legal and politics more than anything else. I think a lot could be solved if someone decided to build the healthcare version of Zapier.