BLOG on IT/IS in Healthcare & Life Sciences Joerg Schwarz on Healthcare [Health + Care]

Health care is a major challenge for America. Since Senator McCain asked, here is an answer why health care should be a top agenda item for the new President.[Read More]

We did a scalability benchmark with the open source EMR/PHR package from Tolven. Yon can find a detailed report here. If you are a developer or ISV, and would like to perform a similar exercise, feel free to contact me. In order to get started, you can easily get a free Sun SPARC CMT server through our Try-and-Buy program and join our Sun Developer Network (SDN). 

New Sun SPARC CMT single and dual socket servers T5140 and 5240 provide great performance for eHealth infrastructure using web service and relational database technologies. An attractive try-and-buy program allows testing of this technology without risk, and for quick movers there is a special bonus.[Read More]

The Committee for Economic Development (CED), a reputable think tank with significant participation from industry and government, published a report about the benefits of openness for the advancement of health care, both in research and clinical practice.[Read More]

Sun is a founding member of the Open eHealth Foundation, together with Agfa Healthcare of Belgium and Intercomponentware (ICW) of Germany. The open source eHealth stack leverages many of Sun's web services components and it's widely used SOA framework, thus building the foundation for interoperable health care information systems that benefit both health care institutions and health care IT vendors.[Read More]

HIMSS 2008 was a great success for Sun. We showed the Blackbox, a self contained datacenter, and featured SunRay care giver mobility in several different settings.[Read More]

This blog entry reflects on globalization in healthcare and it's impact and dependencies on IT/IS in light of a large health care conference in Dubai.[Read More]

Sun hosted the NorCal HIMSS chapter last month (December 2007) in our Santa Clara campus. I waited with my blog entry for the presentations to be posted, and here they are. Ted Eytan from Group Health did a fantastic job - I really recommend his presentation and his blog entry. When we talk about PHR, it often sounds like science fiction. But Group Health is using personal health records extensively already in their daily practice. It's also impressive (and entertaining when watching the respective TV commercials) to see how Group Health uses secure email for the communication between physicians and consumers. 

Dr. Linares from Anthem gave a strong endorsement for PHR from the point of view of the the largest health plan in the US. Their PHR pilot works with our partner CentriHealth on Sun CMT technology. 

The bottom line: for HMOs like Group Health and Kaiser (they were in the audience and were interactive with Ted), PHRs are already reality. It might be that the economic model for HMOs is a strong driver for this - better, more efficient health care (based on available medication records) create healthier, happier subscribers and more profit - sorry - surplus for research and lower premiums. But also conventional payers like Anthem see the benefits and look for ways to include the different stakeholders to make PHRs useful. 

 

Yesterday was the day of acquisition announcements: Sun intends to buy mysql and Oracle intends to buy BEA. And one way to look at is shows how both moves are related: companies engaged in middle ware are building out their LAMP stacks.

LAMP as Jonathan blogged yesterday w/o further explanation stands for Linux, Apache (as a placeholder for Apache's Tomcat app server), MySQL (as a placeholder for an RDMS) and PHP (as a placeholder for database to web middle ware). Sun now has a complete LAMP stack in open source: Open Solaris, Glassfish, MySQL and  JDBC, while Oracle has a similar stack completely in a conventional proprietary license model.

In both cases developers and users will be able to get integrated products and professional support. The difference of the Sun stack is that we can optimize innovative technology to servers like our SPARC CMT, achieving mind blowing price performance. Do we need this in Healthcare? Absolutely!

Modern EMR applications, PHR's, and HIE all can be designed in a classical web tier architecture, as demonstrated by Tolven.  Put Tolven on the Sun stack, which we are doing, actually, in our benchmark center right now, and you will see unprecedented price performance. We will be able to deliver EMR, practice management systems, e-prescribe solutions and so on in a SaaS delivery to millions of physicians at the fraction of the cost. Today, our Niagara servers are heavily taxed by Oracle because their performance is so good they charge licenses fees that easily quadruple  the system price. With MySQL we can even improve performance and further lower cost per user. Entire Hospitals can run on a small blade center, including all their desktops. Sounds like Science Fiction? We'll show it to you at the upcoming HIMSS show.

So the bottom line for Health Care is that an open source LAMP stack from Sun will combine the best of two worlds: mission critical support (for a fee), which we need for production in health care, combined with the open and flexible data architecture of open source at zero license cost.

Watch this space for upcoming performance results on some open source stacks - you'll be amazed!
 

The key thing with medical images is that you have to keep them for a long, long time. A mammography scan taken on a 40 year old woman today will have to be retained for about 40 years, maybe fifty or sixty-  for the life of the patient. The average life expectancy of data media is five years - disk or tape. Even if you archived on optical media, like a WORM, that can keep data for 50 years, there is no guarantee reading devices will still be around in 50 years (has anyone seen a 5 1/4 inch floppy disk drive lately?). This is just to say that data migration of large amounts of data is not a trivial, yet essential design element for medical image archive solutions.

Read Jonathan Batchelor's article in HealthImaging, and it will become clear 
why our SAM-FS  technology is such a big deal. Dr. Cecil describes it very simple and correct: just add new media, and it'll fill up. The file handle never changes, from media to media, so a data recall after 20 years will access data that has moved four or five times, maybe more often, its physical location.

RSNA is around the corner. Come see us at booth 5155 in McKCormick Place South Building, Hall A. We will demonstrate a few cute things and we'll be happy to discuss our archiving technology with you. Also, we have an exhibit of our new content addressable storage system, the Sun Storagetek 5800 (also known to many under the code name Honeycomb). We'll be happy to discuss with you how content addressable storage will change the way we can drive science and medical practice alike to new efficiencies, while maintaining the cost leadership and reliability you can expect from a Sun StorageTek product.

Musings about open source and open standards in health care, or the need thereof, after attending AHIMA 2007 in Philadelphia.[Read More]

 

So, today I read in both the Mercury news and the Chronicle  about Microsoft's launch of Health Vault. Isn't that interesting?

Don't we all love to update our virus scanners, pop up blockers and so on every day, because Windows is designed so bad that it has all those vulnerabilities. And now the company with the track record in building such faulty systems applies their knowledge to storing sensitive health information. Let's all go sign up right away!

Wait a minute - I was just kidding. First, I was so sick and tiered of my 15-minute-boot-up and virus scanning laptop that I switched to Mac, and never looked back. And apparently a lot of people are doing this now, so not everybody likes to run virus scanners all day. Second, signing up to Microsoft and partners web site might not be without peril. Read for example this review. Smells like passport to me, smells like collecting information and disclosing it to all kinds of affiliate parties. Of course, all for the good. Who would ever assume evil marketing ploys when it comes to Microsoft - exactly!

But maybe we are all jumping to conclusions. Maybe Microsoft will run the site using Solaris with trusted Solaris extensions, so the data is secure (confidential, complete & right, and available when needed). And maybe they will encrypt all the data, all transactions and identities all the time. Yes, and maybe they will open the data standard for everyone to build a connected system, so professional data can be connected?

yeah, right. Maybe.

 

In order to give you another view on the Health 2.0 conference, read Amy Tenderich's summary of Health 2.0 at Diabetes Mine, one of the, if not THE most popular health care blogs. Amy was also a moderator of one of the panels, and she provides a good overview of some of the companies that launched at the conference. She also posted a link to the opening video, which is great and worth a look!

 

What would a new frontiers conference like Health 2.0 be without some brand new launches a the show?

Here are two of them:

Enhanced Medical Decisions launched DoublecheckMD The idea is brilliant: patients type in the data about their prescriptions and the site comes back with information about reported drug interactions. The idea is that the MD doesn't know all potential interactions while making the prescription.

However, wouldn't it be nice if there was a personal health record that already did this kind of check while the doctor makes the prescription? Old economy, brick-and-mortar guys like me would call this CPOE, and Dr. Greenes at HBW gave many talks about CPOE systems that do this (checking interactions and reported allergies) when the doctor makes the order entry to avoid, prevent risky prescriptions. Greenes reports that it's hard to do in the controlled environment of a Hospital. One can only speculate that DoublecheckMD might be better off since patients care about themselfs, but also question where the data will come from, how patients find it and what they will do with the info.

Another new startup was peerclip.com, a social networking site like sermo.com. I enjoyed Sermo's free water, coffee mug and after show party, so I'm maybe biased towards them, but fact is that there was little info about peerclip at the conference. Fact is there is at least two of everything, so why not two social networking sites exclusive to Physicians?

Anyway - my final remark on Health 2.0 - somehow I have to think about that sock puppet from pets.com all the time. Why? Don't know. Maybe because they put so much money into advertising so everyone knew the sock puppet, but somehow there was not enough margin in shipping a 40 lbs bag of dog food across the country (free shipping, of course) for 10 bucks (undercutting the Petco around the corner) to feed that sock puppet past the dog com....ehm dot com bust. I had to think about that sock puppet. No offense to peerclip, or healia or doublecheckMD, but I wonder where the dog food will come from this time around, if you know what I mean.

Reflecting on Health 2.0 - What did and what did not happen, what was there and what was not there

My Mac ran out of power yesterday, so I had to stop live blogging half way through the conference. But that's maybe a good thing, because it provided some time to reflect on what happened and what didn't happen yesterday.

So first, we know what happened: the entire crowd was very pleased. Great panels and panelist, good organization, exciting topics. In the audience were many big companies (I was, for example, by sheer happenstance, on a table with Pfizer and United Healthcare and a Venture Capital representative), something we used to call "old economy" in the .com days, and on stage were the young start-ups, the ".com's". It felt like 97 or 98'ish - I was in Silicon Velley then, and boy, it felt much like it at Health 2.0 yesterday. For the good, and for the bad.

Health 2.0 is about using Web 2.0 technology to personalize health experience. From personalized health search on specific sites like healia.com, to social networking sites for patients, like patientslikeme.com, or care professionals, like sermo.com, or the linkedin for physicians, within3. I could go on and on describing these sites, but that's not the point.

Jay Silverstein from RevolutionHealth and Esther Dyson from EDventure made excellent remarks in the final feedback panel: the conference was great, lots of innovation and enthusiasm, but Health 2.0 is too fragmented, too complex right now. Too many sites to go to. People, Life is not one dimensional.

While google and yahoo are not as good for health searches as healia, they are good to find sites like healia, and people do not search for health related things all the time and exclusively. One might b e a patient at one time and use a site like patientslikeme, but why would that patient have to go to quickenhealth or HealthEquity to keep track of medication expenses, or go to ? Why does that patient already registered in a patient community need to go to Vimo or Careseek to find a doctor rating? Shouldn't all health plans do what BCBS of Minnesota does with Healthcare Facts, anyway?

Marty Tenenbaum had, like me, a deja vu feeling, and said he felt like the first e-commerce meeting before companies like Amazon and eBay figured out that they have to build an entire supply chain, not just offer a point solution and a good idea.

That brings us to the other side of the day and to what did not happen. google, for example, did not announce it would acquire WebMD (or anyone else of the startups present at Health 2.0). No, google will instead look carefully at the evolution of Health 2.0 said Missy Krasner. Maybe I was right, and google discovered that the true value comes from connecting personal health records to the legacy systems.

And that's what was not discussed - how data from existing and future EMR/EHR systems could be used to populate PHRs with quality data, instead of relying on data provided by patients. Besides quality concerns, who has time to update detailed health information all the time? But once one has a PHR, that PHR could become the central hub for managing finances, searching for good providers or joining social networking groups. But none of this joining, aggregating and connecting was discussed yesterday. Maybe I think too much in old economy, but it was a common sentiment in the audience.

Esther Dyson made another good point when she said that is was also not discussed how to keep someone from becoming a patient - the whole lifestyle, wellness and prevention story was absent from Health 2.0. Why? No good ideas?

There will be a next Health 2.0 conference, maybe soon, and those questions need to be addressed, or the people who were there in 1999 will have another deja vu in regards to Health 2.0, but this time one the reminds of 2002....

As promised - live from San Francisco and the Health 2.0 conference.

Half an hour to go and the room is filling up. Looks like a number of VCs are here looking to find investment targets.  They have an interesting audience response system here (wireless dell handhelds). Nice idea to involve the audience and the price to pay for free wireless. Great tool for polls during the panels - let's see how that works out.

First question on the ballot is who will be the next President of the United States. I didn't know Lindsey Lohan was running. Maybe I should pay more attention to the presidential primaries. Anyway, turns out Hillary Clinton won and Lindsey Lohan got a respectable 11%.

Room is packed, intro video about the history of Health care by scribe media is really cool.

Matthew describes what "health 2.0" means. Definition of Web 2.0 from Tim O'Reilly 9/2005.

Holt's best guess: personalized search that finds the right answer for the "long tail", better presentation of integrated data, communities, putting people in charge of their own health) enablement. The key is really personalization when contrasted to conventional tools.

Continuum from user generated health care to users connected to providers to eventually impacting drug development (personalized drugs).

Allegedly 120 mio adult Americans use the web already for health search, 2/3 of the physicians, and the

Doesn't sound like google is going to buy webMD - in fact, they are very careful in regards to monetization. Yahoo, and of course Microsoft, are way more aggressive. Neupert announced that they now have nine customers for their own clinical software. [Must be interesting to be a Microsoft partner, rely on their middleware AND compete on applications. Sun doesn't do applications]

Dr. Brailer: "health care services users do not overlap with the web 2.0 / health 2.0 users". "tech is not the only solution"  ...."by the time web 2.0 users become health service users, we'll be at health 3.0 or so". 2007 and 2008 is the shake out o health verticals - some companies will be there for the long run, others will find out it's a marathon, not a sprint.

Really nice search presentation by healia.com (google type search specialized on health, includes medline) - announced new feature to search clinical trials. The key discussion, however, is the business model of specialized health search engines. Anyone can use google or yahoo every day, but is looking for health info only once in a while - for example after returning from a doctor consultation. There is no question that the specialized engines are helpful, but the business models need work. An audience question covers how web 2.0 content is included. Quality control is of course an issue, and some sites allow tagging and comments (ala wikipedia)

When is the last time to heard about a sold out conference in Silicon Valley? 

Tomorrow I'll have to drive to San Francisco [sighh]- the city that introduced a 200 million dollar universal health care plan this week - and stay overnight. I do this very rarely, since I live closeby in Silicon Valley and hate the SF traffic and parking mess. So what's the big deal? Matthew Holt's Health 2.0 conference on Thursday, that's the big deal.

Matthew posted on his Blog (and sent in an email to registered attendees) details about his troubles: Health 2.0 is sold out, and he encourages everyone to be there early, like at 7am, to avoid standing room for the day. Ok, so I'll get there the night before and bring a sleeping bag (kidding). What is this fuzz about?

Could it be the rumor about google and WebMD (the "google situation")? Sure. First there is a general craziness about web 2.0, and then there is generally craziness about everything google. Mix the two, and you get a sold out conference. Nobody wants to miss the train departing for the next 'youtube' ride. But what would google gain with a WebMD acquisition? I'm not a big fan of disconnected PHRs. A PHR without data feed from EMR and EHR is basically useless for care professionals. And if google was to buy WebMD and leave it at this level, I don't think it would be worth their money. Professional, connected PHRs will outperform disconnected PHRs as they become available, because they are more valuable to consumers, providers AND health plans alike.

Disconneced PHRs with nice flashy portals are easy, and there is actually not much web 2.0 about them. We had user communities for certain disease communities from the early days of the Internet. Connected PHRs on the other hand are a lot of work, because they require interfaces and data, and the data needs to come from physiscians who are today reluctant to adopt EMRs.

However,  google coming to the party could change dynamics. Their gameplan could be to use claims data collected by WebMD's lesser known clearing house to populate the PHR. Throw in some e-prescribe service and you got something interesting going. google could also give a free PC (or better yet, a Sun Ray :-) to every doctor who subscibes to their service (800,000 " $500 = $400,000,000, a mere joke given google's cash) and build up the database with physician interaction in no time. Now this would be exciting!

So, Health 2.0 shall indeed be an interesting conference this Thursday, and since I registered early, I got a ticket to ride. Watch this space on Thursday for a live report from the scene!

One of the most attractive exhibts of our booth at RSNA last year was the GE Medical 3D visualization software demonstrated on a pair of  Sun Rays. People were literally looking for the catch - but there is none.

Dr. Linda Feliingham and her team have developed 3D visualization software that uses an off load renmdering engine (= bunch of graphics cards) to allow true 3D on Sun Rays. Good enough for ultra hi res 3D recontruction models, for example for surgical planning, that would otherwise need full blown Workstations.

And the cool thing is that you can test it yourself. Check out Linda's blog and read instructions how to play with 3D on a ultra thin client.

 

 

My 13 year old son Magnus sometimes reads my BLOG - salve filius meus - and he complained that I'm too US centric for someone with a global scope. Magnus lives in Germany, I live in California, so its understandable that he questions the international scope of my BLOG.

So, with all this said, I will still write about HIPAA (Health Insurance Portability and Accountability Act of 1996). For Magnus and all other international readers - HIPAA is the data privacy standard for healthcare in the United States. As far as I know, every country has some privacy laws, and usually they cover health data. Fact of the matter is, many privacy standards are more strict then the privacy standards here in the US. So please allow me to use HIPAA as a placeholder for the many laws and regulations around the world that protect privacy in health care.

HIPAA's security rule from 2003 with a compliance requirement of April 2005 is, as the name says, about keeping health records secure. In information security, e.g. ISO 17799, we always refer to three different aspects when it comes to information assurance:

• data security, which means only authorized persons can access data, (this is also called confidentiality)
  
• data integrity, which means means the data needs to be correct and complete, and
  
• data availability, which means authorized users can access data they need whenever they need it.  

And what do you know, that's exactly what you can find in the code of federal regulation (CFR), title 45 (public welfare), subtitle A (Department of Health and Human Services), part 164 (Security and privacy), paragraph 306 (Security standards, General rules):

"(a) General requirements. Covered entities must do the following:
  (1)  Ensure the confidentiality, integrity and availability of all electronic protected health information the covered entity creates, maintains or transmits." 

HIPAA regulates furthermore who can access health information and establishes standards that allow auditing who actually accessed and/or altered data. Latter aspect is important to create accountability, a namesake of the act and thus one of its chief goals. You can read a summary for consumers at the website of the Departement of Health and Human Services website.

Now, many times we read articles like this:

---

 

 

 Ivy league universities like Johns Hopkins are not immune:

---

 

---

 
And, for our international readers, it doesn't happen only in the US, but also in Canada (and elsewhere in the world):

---

Common in all these cases is one thing: patient data was stored on a PC or harddisk and got lost, hence exposed to whoever might find the disk or has stolen the PC.  Errare human est, as the romans said, people make mistakes. If you look for it, you can find incidents almost weekly where someone leaves a laptop behind, or a PC gets stolen. As a consequence, hundreds or thousands of records are exposed and CFR 45, 164.306 is violated.

This is so unnecessary. Steve Nelson, our resident security expert and passionate sandal bearer (see his picture in the Singapore blog entry) is always upset to find these reports, and he actually provided the links used in the blog entry.  I actually had to promise a blog entry to him so he would calm down this morning.

Sun in Healthcare defined "mobility with security" as one of our focus areas. We know that health providers need access to data, and they roam in the hospital all day and night long. But why would anyone want to load data on a PC or Laptop and take the risk of exposing records, a clear HIPAA violation?

Our model is to keep the data safe in the data center, where we have physical access control in place, and provide access to it over the network. We call this concept Secure Global Desktop.  And if you need devices all over the hospital, we virtualize the desktop, be it Windows or Mac or Linux or Unix or Mainframe or all of the above, still keep it in the datacenter, and deliver the encrypted desktop bitstream to an ultra thin device called the Sun Ray. And voila, all a care giver has to carry around is a smart card for access control. Maybe a smart card with a picture, as that would establish what security experts call three factor authentication:

• something you must have (the card)
• something you must know (password)
• something you are (the picture on the smart card)

Is there any better way to achieve HIPAA compliance and avoid embarrassing exposure of health records, while saving money and the planet (a SunRay consumes only 6 watts or so, substantially less than a PC or Laptop)?

 We know we can do this, because we have many references. Here in the US for example UAB - see the blog entry and video - in Canada - also on this blog - and (nota bene, filius) elsewhere in the world, like in Korea, Spain and Germany. In all these places, hospitals use Sun Ray ultra thin client devices in day-to-day operation, avoiding any HIPAA or other privacy law violations. It's just a better model.

Steve - I said it. I'll say it again. Next week we will probably still have another case, but don't worry, one by one, we will get this message across!

The Kaiser Family Foundation has just published it's most recent study on health insurance premiums in the United States, providing great empirical data about the state of our health care system.

 

Over the years, those rising health inurance costs take more and more money out of employee’s pay checks. The average cost for health insurance is now over $12,000 dollar for a family of four (12,106 according to Kaiser). Employees pay an increasing part of this ($3,281), but the lion share ($8,825) has to be paid by employers. For a person with minimum wage, a health insurance would consume almost 100% of the gross salary, and for an employer to provide a health plan would almost double the headcount expense. No wonder Medicaid expense increase dramatically year over year. Medicaid and Medicare together represent $590B or 20%  of the federal budget.
This has consequences for competitiveness both from an macroeconomic and microeconomic view. Macroeconomic effects are that labor migrates to cheaper cost countries. Even if wages, which unions and workers can influence, remain moderate, the health care cost becomes a major cost factor.
The US spends 16%  of the GDP on Healthcare, most other countries much less than 10%. A company would save substantially by moving labor to low health care cost countries, all things equal.
The second effect is that more companies do not provide health benefits anymore, reaching a low of 60% in 2007. Consequently, the number of uninsured rose to over 17% of the non-eldery population, more 46 million Americans. And we know that uninsured have no or minimal preventive care and therefore not good medical outcomes, resulting in expensive emergency room visits, which promotes the problem.

Fortunately, the cost trend is not inevitable. Much of the cost is caused by ineffecient or non existing communication, resulting in redundant procedures or bad medical outcomes

 (Copyright: Joerg Schwarz, Sun Microsystems)

I could write pages about this (and I will, because it'll be part of my dissertation), but just in a few short sentences we believe that the collection of individual episodes of care from different providers (electronic medical records, EMR) into an electronic health record (EHR) will build the foundation for a solution to the cost containment problem.

If we can use the efficiency of B2B tools to connect providers, pharmacies, payers and consumers, we could actually prevent all or most of the unnecessary procedures. Secondly, Physicians could use the information to make better decisions, for example based on drug efficacy for target groups prescribing a better medical regiment the first time, which would lead to improved outcomes. And better outcomes reduce cost with faster recovery and avoidance of emergencies.

EHRs are also the base for personal health records, with are the B2C tool that involves patients actively in the management of their own health with life style choices and compliance monitoring.

And we all know, prevention is the best way to cost containment - which is the reason we chose the apple as the logo for Sun Healthcare (Apple a day keeps the doctor away, you know).

 

 

 

 

 

It's been a while since my last Blog entry. We've been at AHIP and MMIS (and at internal sales trainings to bring our sales teams up to speed).

What I took from these meetings is the feeling that PHRs are coming big time. Just before AHIP, the american health insurance plan conference that was in Las Vegas this year, Wellpoint announced their 360 degree patient management program. They had conducted a proof of concept with one of our partners. Exciting stuff.

At MMIS, the Medicaid Management systems conference last week in San Diego, several states expressed their interest in adding health and disease management functionality (alas PHR) to their program. The State of Minnesota will go as far as prescribing EHRs as a mandatory requirement by 2015 for anyone treating state employees. One can only hope other states and private payers follow this bold move.

These developments are exciting because they could provide the missing motivation for physicians to adopt EMRs and participate in RHIOs.  Read the article about the failure of the Santa Barbara RHIO from the California HealthCare Foundation. Many of the failures listed there could be addressed if there was a clear economic incentive to participate in a RHIO, in addition to good technology, of course.

Now - good technology, at least, is available. Check out the material on single patient view based on our JCAPS technology. Single patient view is a critical component for Health Info Exchanges, providers and payers alike. Our marketing folks put together a really nice information session on the technology and use of single patient view in health care.

 

What is the bigger challenge - scaling up or scaling down?

Scaling down! With enough space and money its easy to build reliable infrastructure necessary for a hosptial environment, but that comes at a price that might be prohibitive for smaller institutions. But what if you want to provide the IT infrastructure for an entire hospital, with redundancy and failure recovery, for less than 20k in a single, professionally designed, easy to manage box. Would you have to beg, borrow or steal to accomplish this?

Not anymore! Today we announced a new generation of Blades. Normally I don't talk about specific products in this blog, but the B6000 is different, so have a look at the virtual tour 

Eloy Rodriguez, our business development manager for healthcare in Spain, wanted to share the following comment in response to my earlier posting on open source in the context of health care:

"Don't be surprised if Spain leades the Healthcare Open Source in coming years. A lot of public initiatives, groups of standarization (not only in technology, common semantics is really important), consensus on interoperability, etc are in the agenda of the market. But my feeling is that "open" is passive and the real value is linked to the concept of "community". Community means share and grow. And don't be surprised if Sun plays a nice role..."

Eloy's comment is really important - open is much about the ability to share, for example data architectures. That's why its so important not to confuse the issues. Open source is not only about cheaper software acquisition, its very much about being able to create an architecture that allows everybody to share and access data, and maybe also using some shared tools. Still, not everybody will have to use the same applications. Different communities can build specific extensions. As long as the architectural foundation is solid, this provides the flexibility needed for an infrastructure for an entire region or even nation. I'm specifically glad that Spain also addresses the semantical compatibility issue. Does I.V. mean in vitro or intra venous? Depends on the context, right?  

Sun's decision to open source our software must be seen in this context - if you build an identity infrastructure for the healthcare system of an entire nation, would you want the data to sit in a black box, for which you have to pay money to get a little glimpse in, or would you want to build in on top of industry standards with an open source implementation, so you could continue, modify and extend it as a community whenever you want to?

 

Last week, the first pan Asian health care conference was organized by HIMSS in Singapore. The Sun Healthcare Team was there to meet customers and partners in the APAC region.

In my opinion, Singapore is about as exciting as Indianapolis - and don't get me wrong, I like Indianapolis. It's not as exciting as New York, as much as Singapore is not Hong Kong or Sydney, but it's nice and clean.

 

The conference was in a shopping center - yes, that's right - called the Raffles City. Shopping is the main cultural activity in Singapore, so it makes sense to combine shopping with conferences, shopping with dining, shopping with entertainment and so on.

 Since it was the first conference of this nature in Asia, we dipped our toe in the water with a do-it-yourself booth.

 

Although it looks like Steve Nelson, Mark Handy and myself are just standing around, we really try to figure out how to turn this into something presentable.

After HIMSS in New Orleans I complained that we didn't have a nice demo that shows the popular hotdesk technology (session mobility) of SunRays in a health care context. Steve Nelson and Curtis Cunnigham took this bait and  put considerable effort into building such a demo. In Singapore we showed it for the first time.

 

I think its pretty evident from this picture that we had to improvise quite a bit.

Meanwhile, Mark and Joerg try to figure out how to set up the network.

 But finally it all works and Steve is happy to show the demo to Ambre, our Marketing guru.

The booth was only 10*20 ft, so we had quite a bit of discussion where to put the stuff - furniture, demo stations, hand outs. But we figured it out and had in the end a cosy and inviting little space.

 

Ambre and Amelia from the Sun office in Singapore are waiting for the flood of potential customers and partners in front of our booth.

 

It's amazing to see that everybody wants to get to EHRs, but everybody has different obstacles. In Malaysia, for example, bandwidth is still expensive. China is surprisingly backward - but like everything in China this will change rapidly, I'm sure. Prof. Li's new annual survey will be a good documentation of this effort. Australia is very active, and we were fortunate to have some members of the team down under at the booth. Our next presence in Asia will be at the Japanese conference in Yokohama in July. Our friend Masaaki Sato from the Japanese team joint us. met several Japanese speakers and took notes for the Yokohama show.

So all in all good reason for a nice farewell dinner on top of yet another shopping center.

 

(all pictures courtesy of Ambre Chevalier)

this is not about the Augustana song.....(although it's one of my favorites)

Two weeks ago I visited the MIT Healthcare conference in Boston, which was really great. The topics were quite a variety, from health care policy issues like EMR adoption by primary care physicians, to nano machines injecting single molecules of an active drug right into a target cell.

The advances on the technical side are enormous, and there might be no better place on this planet than MIT to get a glimpse of what the future will bring in advances to drug- and care -delivery. There were also good talks about drug research (Whitehead and Broad Institute), but less spectacular - good discovery life sciences is done in many place.

On the other hand, the lack of progress in EMR adoption is frustrating. Nobody seems to have found the secrect receipe to make the transition to evidence based medicine happen. Micky Tripathi, in his talk about the Mass eHealth initiative,  quoted data from the commonwealth fund which shows slow adoption of EMR with PCP's in clinics of 9 or less physicians. Only low 20% adoption and a 1.5% CAGR. In larger clinincs, adoption is higher. Why is it important? Look at Greenes' talk about computer assisted decision making. Evidence that these systems can improve medical outcomes exists, but organizations not only fail to implement them, in some cases they had to turn computer based decision making off. So is the solution building better systems?

When I heard this, I thought medical informatics should look over the fence to Information Systems more often. In IS, we deal with the problem of implementing information systems scientifically - that's one key differentiator  to computer science. In IS, the most current model for user acceptance is UTAUT - the unified theory of  acceptance and use of technology (Venkatesh, Morris, Davis & Davis, 2003). Bottom line is that the qulity or convenience is only one factor in technology acceptance, which explains why the same system is accepted and used in one organization, but fails in the next.

It just happens that while I was in Boston, I followed an invitation from Blackford Middleton and gave a talk about this very subject, EMR adoption and UTAUT, at the Center for Information Technology Leadership. Here is my actual presentation with a short introduction to the problem, some highlights from acceptance studies based on TAM (technology acceptance model, Davis, 1989) and ideas how UTAUT could be used to predict and moderate user acceptance.

Bottom line is, though, that even if we all agreed on decision support systems, we needed the data from EMR's as their foundation. Back to square one - how can we get PCP's to use EMR in a self sustaining way? This is a good and necessary discussion, and you will hear more about this in this blog.

 So, overall a good trip to Boston. Still glad to be back to California ;-)

On our website you can find for quite some time a video documentary of  the UAB success story.  UAB uses several hundred Sun Ray clients in their day-to-day operations with Eclipsys and Cerner clinical applications.

/> 

Here is a nice video interview on YouTube that documents the use of the same technology in a small clinic. They talk about the use of Sun Ray vs. wireless devices.

 

 

Well, the cool thing is you can have both! check this video out or this one (same Sun Ray notebook, better music)!

 

If you read Future Healthcare, you have seen my two articles about the payor alliance and open source in the current issue. For those who haven't, and those who did, here is some background to the open source article. The article follows after those comments.

After my recent blog and guest column in Healthcare IT-news Europe about open source in health care, I got a few interesting responses. One series of information came from a consultant for VistA, the VA open source EMR. Apparently, I underestimated the use of VistA to a large extend, based on the data I had before posting the BLOG. Now I know OpenVistA has several hundred installations outside the VA and is used outside the US in several countries. This is great. Open source software has really a number of advantages. But here is a word of caution - it's main advantage is not necessarily the free license.

 A 0$ license fee is great for checking out software functionality. It's also attractive compared to a several hundred thousand dollar fee from a competitor. But realistically, the license fee is just one element in a software project. You still have systems design and implementation cost, still maintenance and training are required, and there is still a need for infrastructure, interfaces and so on. Some people think the entire project will be free just because the license fee is free - that's a misconception.

There are other advantages of open source. For example the fact that interfaces and data structures can be examined and exchanged. Anybody who had to deal with proprietary software will appreciate this immediately, especially in the context of regional information exchanges.

Another huge advantage is the ability to customize, which obscures the line between standard sw and custom software. However, more customization means more departure from support - unless the modification is put back into the open source and a group of users evolves.

So my attitude is really not negative about open source, it's very positive. But that doesn't mean we must certainly not throuw out the window years of experience in systems design and forget about support, design and so on. So that's what 'cui bono' really addresses: open source can be very good, but you still need to address the typical systems design parameters and budgets for it. And when open source comes with commerical grade support (as is the case from Sun, Tolven, MySQL, etc.) it's even better and more cost effective, because one doesn't have to build a custom support organization.

 

---

 

Open Source in Healthcare: Cui bono?

Open source software has become
ubiquitous. Spreading from academic developers and users, open source
is now in appliances, IT infrastructures and on millions of desktops.
In the healthcare industry, open source adoption has not been as
quick. Only a few hundred hospitals in the US use open source EMR,
despite the apparent need for digitization of patient records. This
article reflects on the benefits and risks of open source software as
they relate to the healthcare environment.

Information systems designers are
familiar with Royce’s waterfall model for software development,
which consists of five stages: requirements specification, design,
implementation, verification and maintenance. The implementation of
major software projects like an EMR usually follows this model. After
the requirements definition, organizations usually have to a make a
“buy or make” decision, which evaluates cost, benefits and risks
of customer development or commodity-off-the-shelf (COTS) software.
Reality of course is not black and white, and COTS software still
requires careful implementation planning, verification and continuous
maintenance, albeit more focused on the specific adaptations.
Implementation of open source software is, in this regard, no
different than COTS.

Open source software has become popular
for a variety of reasons. For one, software licenses can be acquired
at no cost, which saves the licensing fee. Beyond that, open source
also allows customization and adaptation far beyond COTS products. If
COTS is supported on platform A and B, because the software vendor
has a relationship with platform vendors A and B, users of the
software are locked into those platforms. Open source users can
theoretically chose any platform and make necessary adaptations,
namely porting the code to the desired platform. They could also make
functional changes with self developed modules. In this regard, open
source software creates a bridge between COTS on one side and self
developed packages on the other, as they allow any degree of
customization, while leveraging a vast amount of existing
intellectual property.

The quality of the initial open source
package is certainly a huge decision factor. In health care, not many
comprehensive open source packages are available. The Veteran’s
Administration (VA), however, did release a very comprehensive EMR
product, used in the daily operation of over 130 VA Hospitals and
clinics across the nation, as OpenVistA. OpenVistA was registered
with SourceForge in 2003 and receives about 100 downloads a day,
which is considerable for a specialized EMR system. OpenVistA has
been translated into several languages and implemented in many
Hospitals outside the US, which shows the potential of open source
for global cooperation. Currently, it is based on MUMPs and C#, but
since it is open source, a potential user could re-write the front
end for example in Java to make it platform independent. Another
emerging open source package is with TOLVEN, which is more focused on
the needs of primary care physicians and personal health records.

But open source economics can be
deceiving. Just because OpenVistA or Tolven are free, OpenVistA or
Tolven projects are not necessary low cost. Just like in any other
information systems project, the five stages of the life cycle have
to be planned and funded. While cost of licenses is saved, there
might be some development cost for adaptation in the implementation
stage. Furthermore, ongoing support is an important and considerable
issue. A hospital EMR is mission critical, as care providers rely on
information stored in the EMR for medical decisions. Commercial EMR
vendors usually provide service level agreements (SLA) with short
response times and performance guarantees. Support availability,
quality and cost should be a considerable trade off decision. Not
only should a potential open source project evaluate availability and
cost of support, it should also simulate how any customization
impacts support availability and cost. Therefore, the project cost
for the implementation of open source application packages really is
comparable with software development projects, with the degree of
customization as the independent variable.

In the case of OpenVistA, the VA open
sourced a code base developed with US tax payer’s money. Hospitals
could adopt OpenVistA and leverage the huge investment in the code
base, for example sharing the EMR across multiple organizations in a
regional health information organization (RHIO). But it should be
clear to implementers that, while license cost for the application is
zero, cost for support is not. It is questionable if open source
community support models popular for Linux or PostgreSQL are
sufficient for the mission critical needs of hospital environments.
Tolven, like MySQL or Sun, is an example for a new hybrid, as it
provides open source access with the benefits of commercial support.
Sun Microsystems, the largest commercial contributor to the open
source community, provides access to its software in an open source
style, but provides with the option for commercial support. A similar
model is planned for Tolven. Some consultants provide a fee based
support for OpenVistA, which is a desired and encouraged open source
support business model.

Decisions involving open source
software should include legal considerations. Open source software
based on the GNU license includes a recursive copyright. That means
any software implemented with statically linked, GNU copyrighted
software components becomes subject to the GNU license itself.
Furthermore, implementers could become subject to copyright
infringement suites, as the recent Novell-Microsoft settlement
demonstrated. OpenBSD based licenses allow almost any type of use,
including commercial adaptations, which limits the possibility of
copyright infringement law suites. Commercial open source licenses
usually seek some variations between OpenBSD and GNU. Sun for example
allows for Solaris royalty free re-distribution and even indemnifies
Solaris adopters from potential copyright infringement suites, but it
does not allow the re-distribution of the Java Enterprise System,
which includes the popular JCAPS stack as well as identity management
and web-server products. Use of JES is free, as long as no support is
desired.

Summary

Although open source software is free,
many costs in projects involving open source software are comparable
to projects involving regular licensed software. Potential open
source adopters should carefully consider if support options for
their application of choice meet business requirements, and if any
legal exposure can be mitigated. However, open source packages allow
customization and sharing of both data and applications in regional
or even national communities. In cases were the benefits of open
source are outweighing the risks, sufficient support infrastructure
should be put in place and considered in the “buy or make”
decision.

The Answer to the opening question, who
benefits from open source in health care, is that organizations who
are prepared to create and invest in a support infrastructure can
benefit from the low acquisition cost and leverage the freedom to
share applications with collaboration partners, unencumbered by
licensing and data migration issues. However, legal considerations
like indemnification and nature of the copyright deserve careful
consideration.

 

 

 

 

 

Long time in the making, we announced on April 17th that Sun and the Province of British Columbia entered a contract to build a provincial laboratory information system - read more.

What's the big deal?

Given the fragmentation of healthcare information about any individual in almost any country, there are currently various activities to create a more comprehensive view of patient data by aggregating those fragments into virtual, longitudinal records. Many scientists and physicians believe that comprehensive, longitudinal records can help making better decisions and therefore lead to better medical outcomes.

In many cases, composite records have to be virtual, as the original data has to remain where it was generated for compliance reasons. A key pre-requisite are electronic medical records - paper and film are not good media for sharing information across long distance and among many partners. At the same time, it is challenging to pull highly sensitive data from various systems and maintain full control over access compliance and data privacy and security at the same time.

So when we saw the PLIS RFI first, we though we had to build a rock solid foundation of identity management, and on top of that use our capabilities to collect data from various systems, generate composite information and connect this with the applications used by care providers every day. Our JES stack  has the key ingridients to build such a system - the IdM Suite and the JCAPS suite. We proposed to build the infrastructure, operate and manage it, together with our partners, and convinced the Provincial selection committee that we can do this better and more cost effective than others.

What is really big about this, is the fact that this will create the largest digital health information exchange in North America AFAIK. Add to this the experience we collected in the SPINE project in England and various regional or local RHIOs in the US, and it's not an exaggeration to state that Sun is the leading provider of Health Information Exchange solutions at this time. What's really cool is that we are not just clobbering together someone else's IP - much of this is Sun IP from the ground up - Solaris, Java, IdM, JCAPS. What a great example for open technology (all of this is open source and can be downloaded for free) driving innovation.

Think about the implications of exit cost for a moment. If Sun would decide tomorrow to stop software development and sell only Intel Servers with Microsoft's operating systems (zero chance of that, but just hypothetically), how would that impact the infrastructure for PLIS? Amazingly, the Province could just go on without re-engineering the entire concept. Although we use so much of our own IP, the Province is not locked in - nor should they. They could buy any x86 server to run the software; our software is open source so they could take it over themselfs. Of course, it's cheaper to let us do the maintenance, but it's good to know that there are options for such a critical infrastructure. In contrast, it's almost inconceiveable that a Government would implenent a provincial, state or national system on a proprietarty technology - imagine asking General Motors to build a freeway for GM cars only. Sorry if you have a Toyota - you'll have to drive on surface roads ;-)

 So, the big deal about PLIS is that it's another case where governments take action and build the infrastructure we need for a digital transformation of health care, and it's also another case where they have chosen a stack of open, interoperable IP (from Sun) to build this new infrastructure.

 

Those people who are skeptical about IT in Healthcare should look into this as a beacon of hope - change is in the air, and it will spread!

 

 

 

This article discusses open standards and open source in health IT. The European Commission sponsors open standards and there is considerable activity around interoperability in Europe - more than in other regions. However, even in Europe, open source EMRs have not been successful yet. The article suggests some reasons and potential actions to solve this dilemma.

[Read More]

HIMSS'07 is coming up later this month in New Orleans - as everybody in Health IT knows.

Less known is the fact that Sun will have

• a client summit prior to HIMSS and
• a booth with cool exhibits.

At the eve of HIMSS, on the Sunday before the show starts, we organize a client summit to discuss our solutions for health care and demonstrate some of our application partners. You can review the Agenda and register here.

Later this year we will also participate in the new HIMSS Asia conference, the second HIMSS show in Europe, and of course we'll be back to RSNA. But among all these shows, HIMSS New Orleans will be a prime opportunity to meet with the entire Health Care Industry team and experts from our different product sales organizations. Our booth number so so easy to remember, even I won't have to look it up: 1111. So - let's meet in New Orleans!