As we continue to see EHR hailed as the mechanism by which American health care will be transformed from an inefficient, anachronistic cottage industry into a sophisticated, statistically relevant expression of cutting edge, evidence based perfection we must resolve the cognitive dissonance that follows from comparing the Brent James disciples’ argument that the EHRs should facilitate and amplify the innate desire of clinicians to do good to the actual EHR adoption and failure rate.
If EHRs offer a new paradigm of improved clinical outcomes via more efficient case finding for focused efforts, reminders for complete care and evidenced based care at the point of care then why is the failure rate so high? Chronic clinician EHR critics frequently site the lack of usability, the dearth of clinician input into design, bizarre business rules that drive inefficient EHR functionality and the tension between IT leaders and clinical informaticists that leads to marginalization of the clinicians due to lack of respect and/or lack of project management experience. It is hard to argue with the 2006 Annals HIT review that warned us that the benefits of EHR have been extrapolated from only a few long term self-developed clinical information systems from a remarkably short list of organizations that developed those systems over many years and thousands of Plan-Do-Check-Act iterations and that only with appropriate cynicism should we assume that commercial EHRs are capable of delivering similar results. However it is entirely appropriate to recognize that at least a few of these market offerings are quite robust – certainly good enough to allow population management, case finding and reasonable decision support at the point of decision making. Why then the continuous criticism, the ‘down sizing’ of some aspects of Meaningful Use to something more akin to ‘not quite significant use’? I believe the principle reason is that we have yet to transform most of the health care system from paying for quantity of care to quality of care. If we accept that transformation of the health care system means we will more reasonably utilize resources based on best evidence and best practice then most would agree that primary care has a very important role to play in terms of appropriately matching the consumption of resources to the needs of the patient both individually for acute care and on a population basis for primary and secondary prevention. Yet most primary care physicians continue to practice in small groups organized around maximizing revenue (visits) with a focus on managing acute problems (or acute manifestations of chronic problems). We are all aware of the low utilization of preventative and evidenced-based care in both the clinic and in the hospital at discharge despite the tremendous benefits that would accrue to the country from preventing complications that lead to ER visits, admissions to the hospital, lengthened LOS, not to mention conversion to long term care, disability and death. The current payment system is perfectly designed to deliver the care we find so inadequate today – care that is fragmented, incomplete and at times inappropriate. Is it so surprising then, that physicians view EHRs as impediments to efficiency, when that efficiency is based on providing the minimum required care in the shortest time possible? Is it so surprising that the current care environment which encourages and rewards egregiously irrelevant documentation (re the failed E/M pay for bullets scheme) is equally inefficient when interpreted into EHR functionality? I believe that in 10 years we will look back upon this very fact as one of the great tragedies of the HITECH stimulus dollars; we will have successfully stimulated the purchase and implementation of EHRs but those EHRs will be crippled by designs that forced the worst of the existing incentives into their fundamental architecture. In conclusion:- Currently available EHR systems are available and have been deployed that offer a level of sophistication that can help motivated clinicians achieve remarkably improved results. Kaiser, Cleveland Clinic, Mayo, Geisinger, Group Health and others have demonstrated the value of integrated information systems using existing technology.
- The current failure rate of EHRs cannot be attributed to poor or inadequate design alone
- The failure rate is directly related to the current set of incentives, tradition and lack of transformation change management that prevents clinicians, particularly those in low valued roles like primary care, from being able to afford the costs of redesigning the way they deliver care to focus on patient outcomes.
If we ‘design with the end in mind’ what would the American health care system look like? Would it change the way we pay for care? Have we had that debate? What successful Accountable Care Organizations exist that provide for significant participation of independent small groups? Should everyone be in a Mayo clinic?
I’d be very interested in your thoughts on the matter. 1.Systematic Review: Impact of Health Information Technology on Quality, Efficiency, and Costs of Medical Care. http://www.annals.org/content/144/10/742.full
The Orthopedic Adventure 
Howard, the way I see it, the current purpose of Meaningful Use 1.0 is just to get people off of paper and on to the computer. That’s about all it does. I agree with you that true transformation won’t occur until the next wave, which would be MU2.0, I guess. Of course, the 1.0 wave is further along (maybe about 15-20% of max) than the 2.0 wave, which is in the low single digits. We need a set of MU2.0 criteria which is *concurrent* with the 1.0 set, otherwise we will have to wait for the 1.0 wave to pass before we get to the transformation era. I’m not convinced that financial incentives tied to payments are the best way to get there.Just like like user-generated content spurred Web2.0, it may take health care users (often called ‘patients’ or ‘consumers’) to drive us towards the Meaningful Use 2.0 era. (see my post on licensing of one’s own PHI: http://hitshrink.blogspot.com/2009/11/licensing-ones-own-health-care-data.html )
Well, hm. Okay, a post-length comment.At my very first EMR conference (TEPR+, Feb 2009) the very first impression I got was the stunning realization that the customers (docs) HATE the EMR systems, because they’re hard to use. That’s the short version; there are details about rigid workflow, poor UI, lost data, etc, all frosted with absurdly high prices. (One MD on Ted Eytan’s blog in 2008 cited spending $7 million on a system for their 41 doctor practice. That’s RIDICULOUS.)So the very first thing that was obvious to me, as a non-brilliant person in the ordinarily competitive high tech industry, is that this must be a closed market that’s very hard for innovators to invade. Because it’s not at all difficult to make software that doesn’t suck. Which almost all EMRs apparently do.So I decided to support the innovation game, starting with moving my data into an open PHR platform (Google Health). Imagine how stunned I was to discover that it’s not just the big establishment EMR systems – even my own hospital’s modest in-house system emitted GARBAGE data. Several times this year I’ve heard that the average physician’s office receives 600 pages of fax every day. Is that true?? (Also in Feb I saw Jonathan Bush of Athena Health say they cost-justify their entire service by eliminating the fax machine.) Even if the number’s off, incoming pages have to be manually filed somewhere: that’s cost (archaic inefficiency) and unreliability (pages getting lost). And outgoing pages have to be handled manually – more time-consuming, archaic inefficiency.The industry largely has a 1980 level of technical sophistication. Okay, maybe 1990. Another example: my understanding is that even two EPIC systems can’t exchange data easily – is that right? Correct me if I’m wrong.I understand the economic incentives issue and I agree it’s important. (I’m on the verge of declaring that Fee For Service is officially Unacceptable.) I also don’t know whether the amount of incentive provided by ARRA will make any difference compared to the human (non-IT) cost of changing a business to be computerized. Doesn’t it boil down to something like $5,000 per year per physician? Not something that would excite most businesses *I* know, if receiving it requires doing something that sucks.But more than anything else, I sense that the most fundamental challenge is, the systems largely don’t carry their own weight: the people who use them, from patient to practitioner, don’t get a lot of value from using the systems that exist today.So I’m waiting for speed demon innovators to show up and do one (just one) thing that’s worth doing, and then add features to it. Bottom line, as much as I understand the obstacles to adoption, I wouldn’t for a moment expect adoption until the systems do something that people want.And here’s the surprise ending: one REALLY USEFUL application would be to let patients see (and take) their data, so they can participate in their care. (!) That sure would be a worthy reason to adopt frickin’ electronic medical information systems.
Dave your surprise ending (having the patients see & take their data) is very problematic. In an industry where quality control has always been done by peers giving the customer the tools that will finally give her/him the ability to check & judge the quality of work done by the hired professionals will be so disruptive that you’ll get a revolt before it happens on a large scale. As we already witness, we will also see all kinds of excuses to minimize the access granted to the only ones who have the final and most important stake in the data. This is going to be a tough job. Succeeding in effecting mindset change. But it is one we should pursue vigorously.
Great discussion… and great insight. @epatientDave, ur assessment is nearly dead on… from this physician’s perspective. Yes we are a fax dependent culture. I am assisting a start up that wishes to deal with the typical fax load a physician’s office experiences. I kept tabs for two weeks. I personally rec’d 65-120 pgs/day, and more important, 1-2% were misfiled into another physicians fax bin. Imagine what will happen when that same person (who misfiled paper faxes) has to deal with a fax server and redirect electronically???? Imagine physicians who do not use email trying to find their faxes in their EMR mailbox. Yes… EMRs, as a whole, are very poorly designed. It seems that the back-end designers/engineers didn’t consult with the front end user— nor did they determine what a physician’s typical work flow was— and design a platform to emulate/improve on that. But that is not the only significant issue with EMRs, as currently designed. As pointed out in the initial post “…The current payment system is perfectly designed to deliver the care we find so inadequate today – care that is fragmented, incomplete and at times inappropriate. Is it so surprising then, that physicians view EHRs as impediments to efficiency, when that efficiency is based on providing the minimum required care in the shortest time possible? Is it so surprising that the current care environment which encourages and rewards egregiously irrelevant documentation (re the failed E/M pay for bullets scheme) is equally inefficient when interpreted into EHR functionality? I believe that in 10 years we will look back upon this very fact as one of the great tragedies of the HITECH stimulus dollars; we will have successfully stimulated the purchase and implementation of EHRs but those EHRs will be crippled by designs that forced the worst of the existing incentives into their fundamental architecture.”The patient centric side of me is also very dismayed at the lack of a robust *patient portal* on most systems.
Long story short, I hope smart customer-centered innovators storm in & create something you love bc it’s so USEFUL. What’s your view of the EMR use at the big integrated systems? You seem to have just given another perfect depiction of how silo care is totally dysfunctional, giving nobody what they wanted from the relationship. Do we all (providers & pts alike) need to reject the silage and go ClevelandMayoKaiserGeisinger?Heh, maybe we should resurrect those Sixties pro-integration songs… 🙂
@epatientDave @gfry You would fall off your chair if you saw the platforms we use at some of the institutions I work at. They are the least intuitive, most horribly complicated systems you can imagine. And they cost ALOT!!!Right now EMRs (the vast majority of users) are simply creating digital, siloed, and tethered data collections that emulate the paper chart the physician once had. The data is useless just sitting there. The Mayo’s and Kaisers of the world have spent a lot of time, money and effort into making it work. Someone is going to have to light the fire that aligns the other 98% of us who do not practice in these esteemed institutions.
For what it’s worth, as you may know, Practice Fusion is a free practice management system, and at THAT price, it BETTER be worthy and easy to use, or who’ll bother? Plus, they’re guaranteeing you WILL get the incentive bucks. So yeah, it’s free PLUS the stimulus money.If the thing actually has any value to it, the big-iron vendors ought to be quaking in their boots. But, if they’re like all dominant technologies when disruption starts, they’ll find it impossible to believe.
Nothing in life is free 🙂 Without a catch! For their *free* EMR you need to give them your billing/ collections. ’nuff said.
Ahhhhhh, that’s interesting – the publicity doesn’t talk about that. So, what’s the commission amount to? Like an EMR?
Gilles, sometimes disruption takes root when a small class cares enough to do something outside the mainstream. IOW, sometimes a solution starts without the current majority giving a hoot.I didn’t think of it when I started writing this but ACOR might be an example, if it (or its inheritors) ultimately undercut the established order. Hm.
Howard you are correct that noting is free. FYI, to the others on the feed that there are other systems that have the same pricing model as Practice Fusion. The SAAS model that Practice Fusion is based on is where everything in software is going, to the Cloud. “The network is the computer” S. McNellyIt is interesting that some of the big iron products have been around for a long time and many have been bought from other vendors. Many systems in use today are basically desktop systems. Vendors are trying to wedge some of these into enterprise solution. Not so easy, that’s why they don’t talk to each other. Enterprise software has to start with an enterprise design on a enterprise platform.Jeff Brandtwww.comsi.com
It is easy to criticize health care for the slow digital transformation and believe me many innovators have stormed in and just as quickly stormed out after their venture capital was consumed. Many billing companies and practice management software companies felt they had the essential knowledge and relationships to carry the day when it came time to offer EHRs to their customer base only to faint at the complexity. Medicine in general is an undefined and fuzzy knowledge domain coupled with bizarre and every changing business rules (not to mention specialty specific practice and reimbursement logic) that leads to considerable programming difficulty. We can digitally transform the paper chart simply with scanning and document management (and there are vendors that preach this design); this would entirely replicate the paper process. And offer little improvement except better access and decreased file storage. The real reason behind the tremendous investment in digital transformation is to improve care with decision support, which is an entirely new concept to include in the care process. The above noted TEPR conference had a special session based on a time based competition between products, a meaningless exercise measuring the wrong metric, which should be improvement in decision making over time and patient outcomes. Time is relevant only because it is currently the critical limiting factor for physician income, especially for primary care. If we do not find a way for physicians to maintain income but take more time with patients we will not improve outcomes. To achieve better outcomes we need to 1. Easily find and monitor the at risk populations 2. Provide evidence at the time of decision making 3. Provide a way for the clinician to review and see progress 4. Decrease the amount of unnecessary care (particularly polypharmacy) 5. Increase the amount of appropriate care (health maintenance and drug monitoring) 6. Provide (and prove) adequate support and education at the proper level of health literacy to the patient populations we serve. Last time I looked none of these were included in the 99213 E/M code. Trying to implement EHRs in a care environment that is not designed to reward excellence in chronic care, Bayesian logic, or wellness in general and then relying on the opinions of clinicians trying to use electronic systems like paper charts is an exercise in futility – much like measuring physician satisfaction under capitation where clinicians, tortured like Tantalus, are expected deny unnecessary care patients consider an entitlement while simultaneously maintaining their trust and high opinion. EHR developers have been asked to create comprehensive systems that are all things to all types of providers (specialty, age, training program, computer skill, and ability to manage change), integrate care across vigorously defended silos, use partial, nonexisting and often conflicting standards, appease CCHIT, predict meaningful use and future reimbursement pyramid schemes all while marketing to appropriately cynical and short sighted physicians that have seen a thousand trends come and go. That being said, and as noted above, EHRs can and do provide a customizable toolkit for those organizations that are determined to improve care using evidence, data and a strong will to improve. There are plenty of innovators in this market place – from documentation management solutions, voice recognition systems, tablets, templaters, etc to fast-growing full-bodied low cost systems like eClinical Works. And yes even Epic – the successful vendor currently under popular attack. Epic has successfully created a scalable solution that is based on a single data repository with full integration of the inpatient and outpatient world, uses applicable standards where available, has a working vendor specific HIE that is pretty good (Care Everywhere) and is the principle reason that most HIMSS level 7 hospitals are Epic clients. Yet it’s too expensive for many and it remains to be seen if it will ever be available to small practices.
You’re all nailing the problems, esp grapmag’s comment. For example, our CCHIT work group just finished putting together the draft criteria for the Behavioral Health EHR (public comment ends Dec 11). It was quite an effort to develop consensus on what we thought an EHR should do. Some of the challenge was the gulf between what clinician-types wanted and what developer-types said was practical. Some was due to the very different clinical models that some of us come from. We also needed to have criteria that a sizable portion of vendors could hit in a year’s time. Throw in some “meaningful use” ideas and it gets even messier. But we all figure you have to start somewhere, so you set a target and improve from there. It will be interesting to see how other certification orgs approach this.The quickest way to get to robust, intuitive, truly useful EHR applications in a rapid fashion, IMO, is to have a large number of folks developing small pieces of it, with a low cost of entry. Think Darwin and survival of the fittest. And open source. I don’t see how the current reliance on large, locked-in, hard-to-modify apps will ever get us to where we should be in a timely manner. If millions of patients were the consumers/purchasers/users of these applications, rather than thousands of hospitals and groups, then we could probably get there 2-3 orders of magnitude more quickly.But no one has developed, AFAIK, a model that would enable this to happen. Maybe Google and Microsoft are coming closer? Again, I think it comes back to putting patients front and center and having a say about their own data. Anything that gets them interested in learning about their health should improve costs and overall health.
This is a great discussion.I have little to add, since most of the people here are aware of the issues. (And also to mention to Dave – Yes, your observations of healthcare are very accurate – The fax, sadly, is our only widespread standard of communication. HIPAA and a whole lot of other things have kept us from using anything else.)With regards to EHR/EMR adoption, though, I’d like to offer some insight. And I only offer this as a person who’s trilingual.I’ve seen a lot of the discussion about “how to make the EMR do what the docs want”. I’ve been to LOTS of meetings about this.What I often see is the docs making certain requests: “Make it hit the road!”.Then I see the IT folks build a machine with a glove and a hand on an arm, and a little square foot of pavement, and the hand hits the road repeatedly.Then the IT folks bring the machine back to the docs and say “Voila!”, at which point the docs say, “That is SO not what we wanted!”.What I’m getting at here is that the implementation issues really fit a linguistic model better than any standard organizational model.(If any of you are multilingual, or are married to a professional interpreter, you’ll understand what I mean.)Essentially : 100% pure interpretation is a myth. Some people will argue that “This translation is 100% accurate”. They’re lying. Or they’re not multilingual.In reality, two people speaking the same language will have about 85% comprehension rate of eachother. Our brains fill in the details, and in most instances, that’s enough to get the message across. Still, sometimes, my wife will correct me : “I meant for you to do the laundry NOW.”But when you cross cultures : Comprehension goes down. A BAD interpreter will try the word-for-word translation approach. Ever try babelfish or google translate? That’s what you get. Sort of a mess that SORT of makes sense but NOT REALLY.Good UN Interpreters realize this and are trained at navigating both cultures. When the American says “Hit the road”, they certainly don’t translate that into the French “Frapper la rue!”. What they do is consider the cultural divide. They consider that what the American is really saying is “Please leave”, but with a certain colloquialism. And they’ll make a judgement call about the best way to communicate that.And in the end, they’re acutely aware of how frail communication really is. (I often wonder how they handled the pressures of the Cold War.)ANYWAY – How this applies to Clinical Informatics : Part of the reason the docs generally hate the software is because the clinical translation is so poor.Getting workflows properly analyzed also requires an investment that most vendors don’t tell the hospitals. (Who wants to say, “Oh yeah, our system is going to cost $20 million, but you might actually put aside $40 million to handle the informatics support you’ll need!”)Anyway, so these are very human issues – Financial competition, poor translation, fear, and misunderstanding – Which lead to these poor implementations.So no wonder the situation is still a mess. And no wonder we’re still mainly on fax machines – Paper is the most flexible medium to communicate by. You can leave all of the errors on the page and still transmit the page to someone else.(Sigh!)
Great discussion… As a non-IT, front end user (albeit, slightly more *digitally savvy* than many of my colleagues) I am very grateful for the follow-up input of the author, Dr. Vaughn and others. It’s also very interesting to see how the process is viewed from the outside in by @epatientDave. No surprises there. Dr. Vaughn is absolutely correct…. the end game is better care. “…To achieve better outcomes we need to 1. Easily find and monitor the at risk populations 2. Provide evidence at the time of decision making 3. Provide a way for the clinician to review and see progress 4. Decrease the amount of unnecessary care (particularly polypharmacy) 5. Increase the amount of appropriate care (health maintenance and drug monitoring) 6. Provide (and prove) adequate support and education at the proper level of health literacy to the patient populations we serve…”My fear is who will *control* the decision making protocols *built in* to the platforms. I also believe, and please correct me if I am wrong, but these decision making protocols will be based solely on EBM??? EBM does not always provide the *best* answer in many situations. I think the patient *controlled* clinical trials that are now underway might prove this (in 5 years or so). I also imagine (ever the skeptic) that Big Pharma has a tremendous role in the direction of many clinical pathways that exist in primary care… HTN, CHF, DM, etc. Who will oversee these clinical decision making tools? Who will they answer to? The concept of improving the health of our patients should be the main driver behind EMR adoption, but I’m willing to be that that rationale is not being debated in board meetings across the nation— It is still $ and cents… ROI, costs, disruption, etc that will keep the adoption curve flat for years to come. For the most part, after discussing this with many, many physicians, they do not see why they should shoulder the entire (and substanital) economic burden of an EMR roll-out, if the end-game is to improve the health care *experience* for everyone — patients, payors, govt, etc…. “…The moment skepticism is abandoned for orthodoxy, scientific inquiry degenerates into pseudo-science…”
Unfortunately most EHR systems have little ‘out of the box’ decision support, probably for some very good reasons: 1. lack of consensus 2. the fear that it will change 3. the complexity of medicine. The best you can hope for under these circumstances is that the EHR provides a robust, flexible rules engine that allows you to modify the decision support to meet your needs. This is where Dirk is absolutely correct – if you don’t budget for the time and human resources needed to create advanced clinical decision support, then you are unlikely to get much back from your EHR investment in terms of improved care. The current implementations of EHR are a far cry from ‘forcing EBM’ – its more like ‘you can do EBM if you want, but its a lot harder than you ever imagined’. Avoiding alert fatigue, inserting ingenious hard stops that are ALWAYS proper for any and all clinical scenarios, creating health maintenance that is evidenced based yet flexible enough to recognize multiple interfering conditions (colonoscopy recommendations for those with adenomas or cancer are more aggressive than for those that don’t have those conditions), and maintaining drug-drug interaction alerting that doesn’t lead to a 98% override rate requires a very flexible rules engine, very creative people that understand the multiplicity of clinical scenarios and the EHR software and the reporting tools available and most importantly a care delivery organization that can recruit adequate ongoing clinician input into the design and repeated redesign required to see progress in providing all the care that should be delivered to each patient population. SAAS has been mentioned above and the question has been raised as to who should create the EBM CDS rules; as someone facing these questions on a daily basis and wondering how smaller organizations will ever find the resources to perform these tasks I hold out the hope that 1. Our leading academic institutions will become more adept at creating decision support rules and recommendations based on best practice and EBM 2. The work of those academicians will be abstracted into a consumable form (“Son of Arden”?) 3. End users will be able to subscribe and import those rules with a minimum of adjustment. That’s a lot of simplification. That’s my ‘I have a dream’ speech for EHR and I fully recognize the inherent difficulty of requesting complex rules for systems that vary widely (even with certification standards), have proprietary architecture and are far from standardized in design.
Good points. I look forward to the day when EHRs live up to your “I have a dream” criteria.
One comment on the fax machine regarding security – a fax machine is not a secure method of communication.In my previous incarnation as medical records director for a long term care facility in the 90’s we routinely received faxes meant for other facilities. There was a great deal of private information each day that needed to be shredded due to someone hitting the wrong speed dial button. This is still quite common today…Sending the message electronically, while not eliminating this issue, will certainly help, and also allow a clearer audit trail than is currently possible using fax.
Brian – You’re absolutely right. The Fax machine has a lot of flaws.Unfortunately, it’s currently our only standard for information interchange. Why?1. Everyone uses the same standard.2. Low cost of implementation : Generally costs you about $100 to start your information pipeline.3. Low training cost : Most office staff know how to use it.4. Low informational cost : Don’t have to edit documents/orders before you fax them.Unfortunately, the cost to the patient : Delay in their medical information getting across. Security issues when someone presses the wrong “Autofax” number.By the way, I love GrapMag’s comments – All dead-on. Decision support is MUCH more complicated than most people realize. This is why EMR implementations have such a high failure rate and poor compliance.Until we start budgeting for the resources to address these issues, having the right software will still fail.I was asked today by a doctor friend who I’m trying to get on the “Informatics bandwagon”. He asked me, “Won’t the insurance companies use this information for their own purposes?”And I pointed out to him : There are a LOT of people who are going to use healthcare IT for their own interests :1. Insurance companies will use the system for their financial interests.2. Malpractice lawyers will use the system for their financial interests.3. Software vendors will use the system for their financial interests.4. Hospital administrators will use the system for their financial interests.So yes, it’s very easy, as a doctor, to say, “Maybe this is all a bad idea.”The only problem : As doctors, we took an oath : To put our *patients* interests ahead of *our own*. That means, we have to be prepared to deal with #1-4, if we want to help get the right medical data to the right practitioners.So yes, the same system that could give ePatientDave an accurate medical history, or give an ED doc a list of the patient’s meds on the outside, has ENORMOUS potential for causing problems in medicine. Eventually, as those of us who still practice know, we’re seeing a lot of doctors wanting to leave the field.This is why I strongly believe : If doctors don’t step up to the plate in terms of leadership, Healthcare IT is going to shift medical culture in a really unacceptable way.Someone in government needs to start examining these big issues. So far I haven’t seen Dr. Blumenthal discuss these effects of HITECH. In my humble opinion, it sounds great on paper, but we haven’t had a national discussion of the costs/benefits of Healthcare IT implementation, and I think that’s the reason a lot of us feel confused about “Why is this so hard?”.It also scares me because we’ve already passed a law that’s going to impact every single doctor, practice, and hospital, regardless of their ability to handle these issues. Oy vey – People are worried about healthcare reform debate? This one has already happened, and the ink is about to dry. Only most people don’t yet see the big picture implications… I only hope we (in this discussion) can help guide our country towards some sanity.- Dirk ;)(By the way, I love this discussion – It seems this is attracting the 10-15 people in the country who really grasp this… How do we get this discussion into the national arena?)
@dirkstanley I suggest we ping our networks and have them join in…. this is one of the best discussions I’ve seen on EMR/EHR’s etc.
The diagnosis of the problems is dead on, and I do think people are starting to get it.Here’s my takeaways and some perspective. *Yes, the system is designed for high cost and low quality. This is too often reflected in the IT systems built to support it. Systems that work (Mayo, Kaiser, etc.) operate under different incentives (cultural and economic) and cultures than most of the health care system.*ARRA/HITECH may, in fact be supporting and further ingraining these broken systems.*High cost often comes from complexity and wide array of workflows.*Usability testing has been historically non-existent in health care.Meaningful use light appears to be thin in order to get users on the program and try not to discourage innovation. This is good news for the SaaS providers and the other innovators. See Halamka’s blog on the principles of the HIT standards committee for evidence.http://geekdoctor.blogspot.com/2009/11/guiding-principles-for-hit-standards.htmlI'm optimistic about ARRA and HITECH because the spotlight and focus on the industry should be a good thing and should increase competition. The competition should help reduce prices and increase usability. This competition will hopefully keep us from throwing good money after bad.None of the systems described fulfill Christensen’s criteria for disruption: simpler, more convenient, less expensive. I postulate that Christensen’s criteria create a large part of the user experience, and it takes a lot of work and talent to make things simpler on the surface. SaaS may be our best bet at providing the less expensive part. Might not work for all, but it will work for many underserved markets out there (primary care, small clinics).EHRs have historically been in the way of physicians doing what we (as consumers) want them to do: make us healthier.Still, I do see some hope here. User experience (UX) is getting some much-needed attention in health care. It seemed to be the phrase of the day (along with mobility and collaboration) at HIT conferences this fall. iPhone is one example. Some estimates are that it has 32% penetration with physicians now. No incentives needed. Paid for largely out of pocket. Why? Applications (like Epocrates) that are useful on the spot in the clinic. When user experience and intense competition among solution providers exist, physicians will adopt even if they have to pay. Effective solutions are those that make the job of physicians easier. They reduce workload. There are people who understand that working with physicians and getting their input and feedback is imperative. The best way to overcome communication barriers is to increase the level of communication, and there are a few of us multilingual (technology, business, health care) people working in the field and focused on user experience. Perhaps it’s a small group, but I would love to see it grow with your help. (BTW, I am testing the waters on building a group dedicated to UX in health care. Let me know if you would like to get involved on twiitter (leonardkish) or commenting here.)The wide array of workflows can be a tough nut to crack, and providers might want to look at the move to electronic records as an opportunity to reexamine their workflows. Same has happened in the ERP space, and eventually these things work their way out. If there’s enough demand, there will eventually be enough solutions that fit.Want to help change things? Please check out http://www.clinicalgroupwarecollaborative.org for more that get it as far as using the web as the platform and modular design in health care. People involved here, like Kibbe and Kuraitis get all the issues described here as well. I’m involved with this group as well, so feel free to contact me with any interest as well. There’s a lot more to say about some great apps getting one job done (look at airstrip), and we will see more of these incorporated into larger solutions via partnerships and acquisitions. Too much to say in too little space.
@leonardkish @dirkstanley @epatientdave @ gfry and others. gr8 comment… great suggestion for f/u so we can all collaborate on making this work.
Very interesting and encouraging that Practice Fusion released a few weeks back PHR recently. This may be the much needed bridge that can help consumers, via PHR use, help drive EHR use on the clinical side. Hasn’t been that connection before. Hopefully this will also mean we can quit using these silly, confusing acronyms. It’s about data and authorizations, not records. Not allowing patients access to their seems to be akin to not letting users examine the systems prior to deployment. Participatory medicine requires a agile development approach. We can only lower costs if patients/consumers participate heavily in improving their own health and keeping themselves (as much as possible) out of the clinic in the first place. Here, too, we need the right incentives.
In response to Howard’s ping… ;)I’ll go contrarian a bit. Meaningful use (MU) per se has nothing to do with quality. At best it is a surrogate marker for what the dominate constituencies in health information industries (HII) believe to pass for quality. In many respects HII is a very small group, out on the leading cusp of healthcare as it transitions from paper to digital representations. The point is that their notions of this surrogate, like many other surrogates of quality (Joint Commission, CAAS, etc.), may not be generally adoptable or ultimately believable. Perception among the rank-and-file likely to be just another obligatory hoop — one they must deal with to survive. Because “funding under MU,” like many other “programs,” is all about the appearance of an inducement up front with a penalty on the back side for the stragglers — reality is a variation on a zero-sum game.Why MU? Because gov doesn’t give money (incentives) without an obligatory carrot. You must have a weir on the dam — or the disbursement of gov money will be based on first-come first-serve — not acceptable as fair.Will MU have long-term impact beyond the limited incentive period? We are seeing the entrenchment of parts of HII in the equivalent of the Joint Commission’s role for healthcare in general. The boarder question is do pseudo-governmental “quality” entities bring more quality? Remember our present healthcare system has risen under the watchful eyes of the Joint Commission.There is a notion expressed in this threaded discussion about MU’s impact (catalyst for change) on the move from quantity-incentivization (Fee-for-service) to quality-incentivization. We as a nation cannot, under our current incentivizations, meet the demands for healthcare (various approximations around 45 million un-/under- insured in the US). It is inconceivable, outside a political context, that the costly transitions from paper to digital representations will of itself tip the balance to quality over quantity and save in overall costs AND cover those disenfranchised from healthcare so far. Political rhetoric + HII hype will have a very short reality shelf-life.Dave touches upon FFS reimbursement, it is and will remain a profound problem. But the alternative is to incentive with what? The history of HMOs, certainly in California, is when you cap at a fixed fee then anything above a certain level of provider-productive gets shunted into other “cost centers” — out-of-network services of ERs. I think it is a fair statement that the failures of HMOs in California directly traces the rise in ER utilizations across the state. FFS brings more cost, but also brings more productively. Captitation of fee (e.g., salary) brings its own problems — chief of which is how to incentive productivity. MU really has nothing to do with how we incentivize the work force, but cuts at the twisting and complexity surrounding the selling of MU.Dave also touches upon the disruptive nature of PHRs — total agreement there. There is a trade off that I believe is seriously in need of consideration. We are presently in the process of formalizing the creation of thousands, if not million, of health information siloes. Yes, I’m aware of MU’s push for interoperability. But in reality, other than the first few that do receive gov money — the vast majority of HI installs over the next several years will see no money — because the pot will vanish very quickly. Gov incentivizes and does not sustain innovations! (or they’ll be on the back side where the penalties reside or the zero-sum game starts culling the bucks)These thousands of siloes will have varying degrees of interoperability for years to come (we’re creating the mother of all legacy structures). Will there be interoperability? Yes 10-20 years downstream (IMHO). Cutting back to the PHR theme now — the disruptive innovation is to create 1 silo (or a very few) right now. HI moves to a service model (PaaS) and the patients/clients are brought in as partners with sufficient constituency pressure to drive this. Provider-based and gov HI funding goes to creating very few siloes and the support infrastructure (minimal at the provider level, web-based). Dream? Perhaps, but the alternative is going to be the chumming that’s starting now. The glory days are now with the big established prominent leaders of HI amongst providers extolling the virtues, the horrible days are coming when you consider the dragging that will be required to bring all single, local, public, safety-net, etc. of providers up to the expected level of participation the HII leaders envision.If you allow every provider to have their variation of a silo — it will be years, decades, ??? for full participation. If you create very few siloes (PaaS), beyond the auspices of providers as we know them, then everyone is ratcheted up to the same playing field. On that playing field (where interoperability is not a requirement, but a basal condition) then we may see MU move from the hypothetical (we see now) to more of a reality. Of course, we got to cover all AND figure out (or settle, as least-evil, how do we incentivize health provider productivity/reimbursement).
There’s some typos in my first commented post (I’ll correct prn), but an issue I forgot to include is one that I’m (and colleagues) are seeing with our CPOE (4 months in). We’ve had a rudimentary amalgam of EHRs since 1997, and just added a new install (layer, floating on a see of legacy) of an EHR 8 months ago with the CPOE “go live” thereafter. That issue is the degree of human involvement/discretion creating perceived and unperceived problems with the “software” (and more importantly the engineer’s concept of workflow and fault tolerance). Many factors at play, the dwell isn’t on the design and implementation, but rather on the concept of the thousands of intents, designs, implementations, modification, etc. that will be “allowed” to occur to facilitate adoption and use. Staggering!!!When you consider using software, lets say a desktop email client — you start using and you adapt to the constraints of the program. When you transition to a web-based client for email — the very same happens. You don’t like how it does something, you try another rendering of the same functions. Some are quite happy with adapting, other are never happy — but the bottom line is email happens. It happens within narrow degrees of utility and acceptability — and most rendering of email function converge around the leading renditions in the field. Accommodation in utility becomes an added/deleted feature downstream only under sufficient mass review. (I’m leaving out the cost comparison between email apps and EHR installs, but stipulate several orders of magnitude difference depending on size). Rather the point being the trade-offs between wide-adoption low-cost utility v. unique-adoption high-cost utility v. time (a 3D plot). PaaS v. unique installs v. time. And then lay in some Newtonian mechanics — unique installs will have tremendous perpetual aggregate inertia v. PaaS will having a high initial inertia (and incremental inertial resistance thereafter). Where inertia is the resistance to change or human (individual) accommodation.HI should be conceptualized as a public utility, the need for realtime total HI in our mobile society is real. The realities of changing health plans and those without health coverage are here for the foreseeable future. HI needs to flow and needs to be ubiquitous not unlike the ubiquity of email (not to trivialize the content). Allowing massive customization and human accommodation as long as at some level interoperability occurs doesn’t foster this ubiquity — it forestalls and adds to tremendous downstream costs. Email works because of the limits of customization and accommodation. Perhaps an MU measure should be the per install cost for a unit of interoperability. PaaS would drive this towards zero (as it has for email).Back to the issues we are seeing with our CPOE — there is a tremendous disconnect between notions of utility and the ease thereof by end users AND the engineering of the UI (user interface) and other aspects of functionality (well beyond the trivial issues of BID and q12h) — I stagger a guess at the complexity of customization/accommodation that are human-driven that will impact interoperability.Didn’t touch upon HIEs, suffice it to say, I find them anachronistic — a solution to the problem of interoperability about 5 years ago. It is a temporary/incomplete solution to what is needed — ubiquity! We aggregate (“silofication”) thousands to millions of separate silos into hundreds and thousands across this country. We all no the problems with multi-generational copies or faxes — what about the issues inherent in multi-generational data? Silofication is also anther obstacle for patients and their access to a whole realtime rendering of their PHR!
Can’t believe I did that type, last paragraph “We all no” should be “We all know” 😉
@HJLuks &all: I’m slammed (can’t play) but this thread’s potent. I’m pleased decision support’s being emphasized!
Great stuff TIm… all you #HIT #EMR folks… don’t miss this phenomenal thread… some awefully smart people offering some very useful advice
Great comments, Tim. “HI should be conceptualized as a public utility, the need for realtime total HI in our mobile society is real.”There’s been a lot of talk recently about the “health internet,” but maybe the “health grid” is a better analogy.
Nice Post, and I second that Howard that many of the HIT system’s I’ve utilized over last 10 years of practice are expensive, involve massively outdated an expensive systems of hardware maintenence. I continue to maintain that I as a practicing phsician am the best to determine “meaningful use” meaning… makes my patients happy, eliminates paper completely from my life, eliminates ( most ) office overhead including all the staff that search for paper an the folks that manage the staff and beg third party payors for money., It’s ecommerce, folks. It’s cloud. It’s physician developed and used HIT. This is not new stuff here… Every business aside from healthcare lives and breathes by these technologies. And the PHR/EMR link to ecommerce for primary care is the final link of the puzzle to eliminate staff. ( And E-Patient Dave and our customer patients) So as we face unprecedented challenges in healthcare reimbursement, costs of dealing with third party payors and government, keep in mind that we are talking about livlihood here for many of us physicians who see patients for a living. Reform is too slow and expensive for us to wait. Natalie Hodge MD FAAPCo-Founder and Medical Directorwww.personalmedicineinternational.com@nataliehodge
@leonardkish… I like the health grid analogy!
First of all, just let me thank Howard and Dr. Vaughn for providing the platform for such a stimulating and informative discussion. This is important. I’ll go as far as to say that the proper implementation of EHR with true meaningful use criteria will be more important than any kind of “health reform” that comes out of congress any time soon.The criticisms of current EHR’s have been enumerated here and are (as far as I’m concerned) absolutely true.ePatientDave and otheres have hit the nail on the head in terms of the most salient points. The problem with the current batch of EHRs is that they1. are inefficient (they are based onthe current model of FFS and quantity over quality), 2. they do not incentivize the kind of innovation required for true population based public health improvement, 3. they don’t communicate with each other (at all) and (most importantly) 4. they represent a closed market that offers an exceedingly difficult obstacle for new innovators to invade. The current setup appears impenetrable and unchangeable.And *that* is why I am enthusiastic and encouraged that the current system is not only *ripe* for disruption but that it will occur sooner than any of us realize. The stars are aligning in a way that will allow rapid and sustainable change in the medical information industry. I truly believe that it will occur rapidly and change the way we look at things for the forseeable future. I’d be interested to see how we view this discussion in 5-10 years. What are the stars that are aligning? 1. Physician dissatisfaction with the current crop of EHRs and the inability to do what we want to do based on mandates from a wide variety of (non-medical) actors.2. Empowered patients who are realizing that their health data can and should be digitally available to them. Period. No questions asked.3. The entire open-source movement that has had quite a track record in capturing hearts, minds, and now market-share by providing open platforms that compete with and increasingly soundly defeat current monopolies and entrenched behemoths. (Think Linux, Wikipedia, Android, Google, etc…)This space is ripe for disruption. I truly believe it is already happening. Historical precedent clearly indicates that the factors I’ve enumerated above will not be stopped. That is why I feel so confident.Thanks again for the fascinating discussion.
@leonardkish… perhaps #HIG Health Information Grid … if it’s conceptualized as a utility perhaps we qualify for TARP/Baitout funds 🙂
As an enduser of fully deployed EPIC system, I see the good and bad of the currently deployed systems daily. Several brief comments are in order.First, let’s not fool ourselves. Money is driving the bus here. Anything that hospitals and clinics can do to increase revenue streams will ALWAYS get priority when software builds are requested. EMR’s are currently designed with collections in mind before patient care. Why else are friendly dot phrases included in EPIC? They add TONS of unnecessary documentation to notes repeatedly. Why? Because without all the parts of stupid Medicare Evaluation and Management criteria, no one gets paid by Uncle Sam. This is clearly NOT about providing streamlined or efficient care. I can’t tell you how many suscinct 1-2 sentence notes from surgeons convey TONS more pertinent information than pages of regurgitated medication and problem lists.Second, information overload and the physician.Since no one gets paid unless the doctors “sign off on it,” I am amazed at the THOUSANDS of orders I sign weekly for who knows what. Every encounter with an ancillary care specialist must also route throught the doctor for electronic “signature.” Can I really be expected to electronically sign several hundred orders a day, review and erify every result, pacemaker and defibrillator check, one by one, while seeing patients in clinic or doing procedures, e-mail reponses to questions that are sent to us each day and still get home at a reasonable hour? Well, crafty programmers have allowed me to highlight all the orders and just right click on “Done.” Or if I don’t get it all done, do it from home. While this is efficient, is it quality? It depends on whether we are interested in checking off national guideline “benchmarks” or really interested in how closely we attend to each item that affects our patients. So here, too, the benefits of EMR’s appear skewed toward that which improves better revenues. Bottom line: information overload is also going to kill end users and I’m amazed at how few EMR experts are talking about this reality. Distributed systemsDespite what others may think, the “cloud” has arrived to nearly every functional EMR system out there. Why? Because increasingly we are seeing companies house patient information relative to their technology on their own servers. This serves two roles: (1) revenue (people must pay for their service) and (2) facilitates regulatory follow-up of medical devices. Nowhere is this seen more clearly than cardiology, with its pacemaker and defibrillator databases housed on separate corporate servers that (may) link back to the EMR. Event recorders are similar. Soon, a patient’s medical record will be housed in hundreds of different places, so development nerds should understand this. Frankly, the ultimate owner of the patient’s data should be the patient, since they have a vested interest in its maintenance. Only problem is, they have no clue how to maintain it, so they’ll remain at a distinct disadvantage for years. Conflicts imposed by guidelinesOne only needs to look to recent medical device industry efforts in the cardiology community to make sure patients “know their heart’s ejection fraction” (a measure of muscle heart function) so they can receive “optimal” care. Never mind that “optimal care” might require an outpatient surgery hospital bill of (I’m not making this up) $175,000 to implant a biventricular defibrillator. While for the guideline followers this might be just the right thing to do (and scores big on quality monitoring websites) the ethical conflicts of “driving business” using electronic “reminders” presents real challenges to ethicists. How do we use an EMR that “guides” us this way and NOT think there’s an ulterior motive?Finally, I would be remiss to not mention the improvements in interphysician communication afforded by EMR’s as it relates to patient care. Our system integrates outpatient AND inpatient systems, so continuity of care is evident, as long as the patients stays in our system. E-prescribing also has improved over the last year, but I resent that I still have to use a paper script for controlled drugs. Why?Anyway, there you have a few thoughts. Thanks for this great discussion. I hope I added a bit from a different (yet similar) perspective.-Wes
Thx @doctorwes. Interesting take from someone using an integrated inpatient/outpatient system. I agree with your comments about the fact that money and net revenue is and will likely remain the driver in decisions made by the HIT folks for the foreseeable future. As I said in a previous comment I am also concerned about some of the clinical pathways that may find their way into the decision making modules. We’ll never really know who’s driving that train. Perhaps it will be the larger Mayo Clinics, and Cleveland Clinics of the world who will help us refine those pathways and vet the data that will ultimately guide the decision making of 100’s of thousands of physicians… but I remain skeptical. Pharma, and the Medtronics of the world will have something to say about these pathways and decisions making criteria… IMHO!Thanks again to everyone… this has to be one of the most informative threads I have ever read.
This has been an informative discussion, but I must take exception to a comment made early in the string by Howard Luks, which is patently false. Users of Practice Fusion’s free, Web-based EHR are free to stay with their own biller or billing service, and keep those contractual arrangements in place…and many do just that. That said, Practice Fusion has been working to integrate certain billing products and services with our EHR to streamline office management processes, improve data validity and so on. We look at this as a value-add for our users. In fact, the idea for this integration came from our users.Luks must have mistaken us for another vendor.
Glenn… appreciate you clearing up the “patently” false statement. thx for your input.
Wishing I had something to add to this wonderful discussion.
wow wow wow!!! Howard look at what you started! This is one of the best discussions on EMR/EHR I have seen. I do not have a clinical background so can not speak to the usability of the tools from that perspective but from what I have seen (inquisitively) I would be frustrated with majority of the tools that are available out there (very broad assumption here). The question arises then (and I am curious if anyone here has any hard numbers) what is the ROI for a hospital for ex. to replace the clunky EHR? What is the incentive for a physician office to change their workflow once again to move to something “potentially” better? I hate saying this but at the end of the day these are small businesses whose profit margins are being squeezed daily…One other thing that I have not seen in this discussion (unless I glossed over it) is the cost of converting and digitizing the filing cabinets? T0 (time zero) going forward will benefit us patients and help us (given data portability) to manage our health, to gather our own data and extract it all from the silos but what happens to the last 10,20,30,80 years of my life? What is the cost to convert and more importantly make sense of this data?
Hi Eugene, I agree with you. You don’t see these kinds of conversations at HIMMS, RHIO’s or the multiple other HIT bodies, because they are only focused on the initial implementation (and sale) of their wares. CMIO’s are focused on the initial implementation of paperless systems. The question you pose is an entireley different issue. You are speaking about constant iteration of the platform a system is using. You are speaking about disruption. Real continuous innovation. And that is much more difficult because large health systems are boggy and slow. I was recently asked to Speak at The Mayo Clinic about the customer experience of making house calls, and the web platform that came out of the customer centric approach. Clayton Christensen of HBS has an eloquent analogy in his talk regarding health care that you should watch. Not many large multi-billion dollar systems are going to fork out the dough for epic, then turn around quickly and iterate into another system, because it’s too freaking expensive and slow. The quick iteration comes from startup players, smaller nimble organizations that can iterate code create new paperless business process and turn on a dime. Our organization has made the kind of replacements that you suggest. The risk is high. We started out with off the shelf, old clunky delphi player with local server backups many docs in primary care are using. Then we replaced that the next year with a delphi player that was rewriting in c++ with a patient portal, still hosted locally with all the server maintenence. Then we replaced that with a full on web platform ( ours) for ecommerce ( our platform is cash pay or concierge or direct practice) connected to cloud based web ap EMR with a PHR that connects to our web back end. Our current platform is on a weekly iteration cycle that results from our physician customers feedback. Not many docs in primary care are going to make these kind of quick changes and capital expense. This requires business partners, sometimes investors and a scalable business model. But the organizations that move through the exercise create incredible value for their customers ( we make house calls, who can compete with that? ) and a resulting overhead of reduction of 80% for our physician users. That’s remarkable. There is such a disconnect between emerging Health 2.0, consumer internet industry ( what customers want) and the physician facing hospital admin facing side of health care (what we do to get paid) . Our platform is one of the first to put the two on a collision course. Hope that helps. I would love to hear others feedback on this issue. Thanks, Howard for the great discussion. Natalie Hodge MD FAAP Medical Director Personal Medicine http://www.personalmedicineinternational.comPersonal Medicine International http://bit.ly/41665P
Hi Eugene, I agree with you. You don’t see these kinds of conversations at HIMMS, RHIO’s or the multiple other HIT bodies, because they are only focused on the initial implementation (and sale) of their wares. CMIO’s are focused on the initial implementation of paperless systems. The question you pose is an entireley different issue. You are speaking about constant iteration of the platform a system is using. You are speaking about disruption. Real continuous innovation. And that is much more difficult because large health systems are boggy and slow. I was recently asked to Speak at The Mayo Clinic about the customer experience of making house calls, and the web platform that came out of the customer centric approach. Clayton Christensen of HBS has an eloquent analogy in his talk regarding health care that you should watch. Not many large multi-billion dollar systems are going to fork out the dough for epic, then turn around quickly and iterate into another system, because it’s too freaking expensive and slow. The quick iteration comes from startup players, smaller nimble organizations that can iterate code create new paperless business process and turn on a dime. Our organization has made the kind of replacements that you suggest. The risk is high. We started out with off the shelf, old clunky delphi player with local server backups many docs in primary care are using. Then we replaced that the next year with a delphi player that was rewriting in c++ with a patient portal, still hosted locally with all the server maintenence. Then we replaced that with a full on web platform ( ours) for ecommerce ( our platform is cash pay or concierge or direct practice) connected to cloud based web ap EMR with a PHR that connects to our web back end. Our current platform is on a weekly iteration cycle that results from our physician customers feedback. Not many docs in primary care are going to make these kind of quick changes and capital expense. This requires business partners, sometimes investors and a scalable business model. But the organizations that move through the exercise create incredible value for their customers ( we make house calls, who can compete with that? ) and a resulting overhead of reduction of 80% for our physician users. That’s remarkable. There is such a disconnect between emerging Health 2.0, consumer internet industry ( what customers want) and the physician facing hospital admin facing side of health care (what we do to get paid) . Our platform is one of the first to put the two on a collision course. Hope that helps. I would love to hear others feedback on this issue. Thanks, Howard for the great discussion.Natalie Hodge MD FAAPMedical Director Personal Medicinewww.personalmedicineinternational.comPersonal Medicine International http://bit.ly/41665P
Great post by Rick and a great discussion. Hard to add anything except personal experience. We have implemented Epic EHR and CPOE with almost totally voluntary medical staff and have achieved over 80% CPOE at all 12 live sites within a rapid period of time (the last hospital within 2weeks). We feel our success is based on having deep clinician involvement in build, design and support of the system.We designed the system, not primarily with financial goals in mind, but with clinical goals as our focus-patient safety, clinician efficiency and acceptance, and getting all of our clinicians on a single platform. We cannot expect that we would see or experience all of the various benefits expected by every stakeholder, and that is why the clinical ones need to be the right ones.Part of why it is difficult to demonstrate benefits is that the data on legacy processes are fairly thin, and that includes medication errors, etc. With these platforms, we can finally start to wrap our arms around and to surface the underlying system issues, and we are seeing this.One of the keys to success and the barriers is the ability to make rapid changes to the system to improve usability and safety. But change is a demon that is hard to manage.Our users request changes, but don’t want to deal with change. For someone who has worked clinically for over 20 years and work with and design these systems, I see the value, but that does not mean the road is not rocky, and that we don’t make mistakes along the way. But as those who argue for healthcare reform say, doing nothing is not an option.Great discussion, and Rick, nice job.
Luis, Your organizations accomplishments are to be lauded and I certainly understand the hard work it has taken to get this far – and how much work remains to be done as you so rightly state in your response. * The chasm that must be leaped to go from paper to basic EHR for a hospital is quite large and under appreciated by critics. * Luis is spot on – ‘pre EHR’ data on med errors and process failures is hard to come by. The only way to get the data of interest was manual chart review or scavenging what fragmented electronic data was available in relevant silos. That costs money if you don’t have research grants. Remember the EHR benefits promised by proponents have been extrapolated from places like the Brigham where they have been playing in the sandbox with their self built toys for 20 years. * The lack of financial outcomes (initially over-hyped by those in power as an offset to health care costs) is the source of the latest media and critic hand-wringing – but as Luis reminds us all – that wasn’t the reason that we bought the software. As silly as it may seem we really believe it is about having a single source of information about the patient that will allow us to empower rapid improvement cycles for quality, safety, satisfaction and efficiency. * The fact that we are able to achieve high levels of utilization with independent medical staff physicians is testament to the planning, hard work and yes to the quality and usability of the software, which is far from perfect but good enough apparently despite the all the noise about how the current crop of EHRs is unusable, designed by suits and nerds that don’t practice, etc. * It remains to be seen, as Luis notes, whether we can really achieve the ‘Brent James’ model of change now that we have the data, especially considering that community based hospitals with independent medical staffs are quite different political structures with different financial incentives. Smaller hospitals can’t afford complex systems; if they could they couldn’t afford the informatics trained clinicians that are needed to continuously update and improve the system (and yes I understand that this has been completely solved by proposing as yet untested, unavailable and yet to be business modeled clouds and SAAS solutions). Thanks to everyone that has posted and commented, it’s been a great conversation so far. I can’t wait to see what the Santa’s ONC-elf will deliver in time to make this a ‘meaningful’ holiday season!
Whew! Very good read. My own opinion is not that the 600+ EMRs are badly designed but that they don’t connect to anything. The reason Kaiser, Cleveland Clinic, Mayo, etc. are successful is that a great majority of the time an individual physician will open up a new patient’s chart and have meaningful historical data on that patient there. In short, the EMR provides a value add.As long as that value add is there so the physician (or their own clinic) don’t have to spend inordinate amount of time re-entering patient’s historical data based on the patient’s memory it’s surprising how useful crappy interfaces can be.Secondly, too many offices that implement EMR’s implement them with paper processes and don’t leverage the electronic nature of the record. Without changing the work flow in a clinic to leverage the ubiquitous availability of the information it’s hard to get the expected ROI.Third, and probably the most important, I’ve found that we’ve gained the most efficiency out of our system by making sure over 80% of our patients have access to and are continuously encouraged to use the web portal. Most importantly we really encourage them to communicate directly with the physicians as opposed to dealing with the nurses. Strange as it may seem this has actually reduced the number of messages in our inboxes, streamlined and improved patient care and saved each nurse about 8 hours of phone time a week. This allows them to be much more productive in the clinic.Fourth we made sure that all of us in the clinic use the same documentation tools. This means that when the physician goes in the room and opens up the chart half of the note is written for them already. They can review the information collected with the patient at the point of care and concentrate on the exam, medical decision making, patient education, order entry and future plans of care.Lastly, doing all of this in front of the patient in the room has helped to improve transparency, patient involvement, compliance and satisfaction.You just can’t do this in the paper world (that is where multiple people are looking at the same data at the same time from different locations and sharing with the patient on large monitors). It just changes the paradigm and has absolutely nothing with the vendor. In my own experience the correlation coefficient of a successful implementation of an EMR and a specific vendor is about the same as the correlation coefficient of the time it takes to drive across town and the maker of the car. The REAL variables are creative workflow changes, the other “traffic” and all agreeing to standardized ways of handling the data and communicating.I could never figure out why when an medical office building was built why a single practice management system and an EMR weren’t included in the infrastructure like the utilities. Imagine the chaos if every office had to arrange for their own wiring and plumbing with different utilities! It’s no different with the information infrastructure. You want success? Make sure as many physicians as possible (or preferably everyone in the same “disease” or “patient” watershed uses the same system. Wouldn’t matter how bad it was, the value of the information alone would be huge.Somehow we physicians have to agree to work together rather than as mom-and-pop shops. Our patients expect it of us just like we expect the banks to cooperate and schlep our money around wherever we go without that much effort on our part.The good news is we can do this. My big regret is that the current administration wants to spend up to $40K per physician to help implement EMRs. What they should do is completely underwrite any and all interfaces period. Most physicians would gladly spend $40-$100K of their own to know everything about that next NEW patient that walks in the door and eliminate all of the costs of connecting to the world.
I have been reading all the comments and got ready to add one more.I understand that according to all the experts (which have always been part of the system, since there have always been the licensing requirements to be called an expert) setting up EHR is very complex. But has any large company or large hospital ever hired someone like Dave for an extended period of time in order to first assess the product and then help in improving/rebuilding it? In the end, we (the bodies/patients) will own all the content in the EHRs anyway. The smarter companies should rush to get their hand on the best expert patients and use their new and still untarnished expertise to create better/simpler/more usable products. I am amazed it has not happened yet. John Halamka, this is a message FOR you! You have the best expert patient at your disposal. What are you waiting for?
This is a good thread by all means. EHRs have evolved from the day years back when I wrote a simple one by today’s terms and integrated it with billing software, again all on the desktop. We have data overload and mining it to get to what we need is an issue too. I have to add one comment, there’s a heck of of code out there! The more code you write, it has to work and communicate with other code, and do it right so now that we are in the “spin” cycle of the wash we have data that has been entered into various systems for years doing just that, spinning. I remember the day that I tested out my Quest HL7 interface and it actually worked and I about fell out of my chair when it worked, but I’m in southern California and Quest had just purchased another lab company who had a whole different set of lab codes, Unilab, so I didn’t get the miles out of it that I thought I would as it appeared a big majority of offices used Unilab and it was a few years before they were combined. I used the above as an example of how sometimes the best made plans don’t always come out the way you want them to work and right now with bringing all the elements together with technology throwing us a left curve every day, it’s a mess and the next issue to work through are the devices and what meaningful use they will have and how they are implemented as most have a short financial life as being backed with venture capital money and need to sell product to stay in business, thus proper and human like use and implementation gets tossed by the wayside at times and we lack laws or there are big gaping holes on how and where the information is and will be used, more code that has to work with other code. As Dave found out making use of “old” established data and mapping it to a format where it can be used constructively is a challenge, talked about that not too long ago with an IPA and the first word out of his mouth was what about the errors in the data, no denial here. One thing for sure it’s an opportunity for clean up, if we have the time. In short, nobody has a clean and perfect data house and transparency is certainly making all of this very clear.So how do you really find meaningful use with all of this and ensure we have not only usable data, but in formats that are user friendly to doctors, patients and everyone else? With the new budget money going out to all the colleges, those students are going to have to learn every EHR out there to be equipped to train, you think? With all the data, drug choices, diseases, treatment plans that exist today that were not around 15 years ago, it is tough and I think we need to simplify this system so we, as humans can work with it, and at least have the same/similar screens available for the clinicians and some standards there in too. It makes me tired to try and help and support all of this too. I think if we had more folks in higher levels of authority become participants in their own healthcare and experience some of this themselves, we might be seeing a few different ideas and approaches here, where’s the hands on role models we need?
Again – This is such a great discussion. We should all pack up in a van and head down to Washington DC and just have a polite discussion with the ONC. A few follow-up thoughts :1. Re : Natalie’s comment “CMIO’s are focused on the initial implementation of paperless systems.” – Huh?? Am I missing something? I’m focused on a lot more than that. (Or am I in the wrong job title?)2. So I’m going to throw this out there, since we have so many smart peeps here, highly skilled with IT :While I appreciate patient involvement and participatory medicine, I will tell you that of all the patients I admitted in the last 24h, I think almost none of them would log in to check their medical care. While there are clearly people who are interested in their medical care, the other side of the bell curve may not share the enthusiasm. But still, I think the ANSWER to our interoperability problems, is patients. – If the government asks us all to speak “Vulcan” to eachother electronically (so we can share our data), then we’ll all just debate it, and the software industry will try to make their version of Vulcan. – BUT… If the PATIENTS start showing up, saying, “Dr. Stanley, does your hospital speak Vulcan?” – And they ask their PCP, “Dr. Jones, does your office speak Vulcan?” – … TRUST ME, we’re all going to start speaking Vulcan. So THEN, I wonder… (as I look into my crystal ball)… What are patients going to accept, in terms of a “Medical GRID”? It seems there are two ways to distribute your data in the future :1. CENTRAL DATA STORAGE = 1 big database somewhere, and every doctor puts there data there so patients can read it – E.g. The PracticeFusion model, “Cloud” central. :)2. DISTRIBUTED DATA STORAGE = Many, many databases all over.Every office, every hospital, gets to keep their own EMR. (Then you have to get them all to speak Vulcan if we’re really going to start sharing data.)So here’s my question : Will the public ACCEPT PracticeFusion (or anyone, for that matter) making the “UEBER-database” of medical data? What if PracticeFusion had 100% market penetrance? Or at that point, will you start to get the public upset about someone storing all the data?I recently started this discussion with some computer hacker buddies of mine, who warned me: “Dirk, don’t assume that you can’t have both CENTRAL and SECURE data storage”, but that’s not the point… I think in our country, where George Orwell has been read by every kid by age 15, that the public may in fact not be happy about a central massive storage tank for medical data, regardless of how “secure” it is.The OTHER option, distributed data locations (e.g. every office/hospital gets to keep their data, we all just need to share it better) then STILL has problems because of the PATIENT IDENTIFIER problem – And so far, the Universal Patient Identifier has gone over like a lead balloon. I can see my Ayn Rand-loving friends throwing a fit about “What?!?! A new identifier that the government gives out to everyone?!? I’m moving to SeaLand!”So which option do you think the [vocal] public would support more? (I’ll pass the mike back to Dave about this.) 🙂
Dr. Voran is exactly right. Their is no magic EHR that is enormously intuitive and effortless. For primary care one of the biggest problems is the ’empty record’ as David points out. Primary care doctors start seeing more efficiencies the longer they are on the system – with the easing of pain starting at about 3 months – when the follow up chronic care patients come back in and the doctor can finally use the one click refills and use the complete problem list, allergies, etc. I absolutely agree that far too many organizations lack the vision to leverage the electronic nature of the chart – they are trapped in ‘paper thinking’ Please stop by David’s Blog to check out his golden rules of Healthcare IT: http://bit.ly/6GhCT8 Rick
Dirk, I think the answer to connectivity is right in front of us. There’s no way I can see a global central clinical database working. Likewise the current models of distributed databases are chaotic and prohibitively expensive (monetarily and politically). Instead my best guess is that we’ll evolve into a social media integrative model.For example, Twitter, Facebook, this blog along with other blogs can be configured to share information seamlessly. It shouldn’t be too long before an EMR installation would include the same type of processes where a list of other systems would be available in which to simultaneously post patient registration data (what I call person-level data) along with visit specific information (visit-level data).In this model when a patient shows up that comes from another office that is using an EMR we’d search to see if that EMR could receive incoming posts. If so we’d select it and import the patient registration data, update it if need be (which would post back to the original contributing system). As the conclusion of the visit the new visit-level data would post back to the other system at the same time it posts to ours. Subscribing systems would see the note as an “outside note” but integrated with the medical record so if the patient returned the record would always be up-to-date.It would be the responsibility of everyone to make sure of the accuracy and integrity of the patient data. EMRs that have junk in them would very quickly be “unsubscribed” and their users would suffer.I could see where the patient could configure their own PHR to “subscribe” to all of their physician’s EMRs as well as retail clinics like Minute Clinic or other labs, pharmacies, etc.This viral approach would dramatically escalate the adoption of EMRs because physicians would spend money to eliminate local work (I hope) and their practices “capital” would increase along with the number of their subscribers.The cost of this? A small fraction of the cost of one-to-one-to-many interface engines.Yeah, there’s a lot of heavy lifting to do to make sure systems use HL7, SNOMED-CT or other nomenclatures for data elements but the standards are out there and there’s a market for a “subscription” enabling engine that wouldn’t store any data but convert data between co-subscribing systems that could either be billed back on a transaction basis very much like eRx transactions today. This would completely eliminate the heavy upfront barriers to interfacing two systems that prevents small practices from sharing information. It goes on and on but I look at our rapid adoption of social media as the model for how clinical information would flow in the future.
Dr. Voran, brilliant! I think you’ve just described the beginnings of a whole new economy of patient information. Lots of opportunity here. In order for this model to take off, the contributors of information will have to be web-service enabled. So far, there aren’t that many, but it just shows that those that are (mostly SaaS providers now) will have a distinct advantage if this sort of model begins to take hold.
Leonard,Just because the native application may not be web enabled doesn’t mean the database isn’t web enabled. I don’t know of an EMR vendor whose underlying database can’t be configured to be web enabled. As an example, go look at the teller’s terminal in any bank. That application natively is pretty much a mainframe application with a windows skin on it. However the database is accessible via the web for the customers.I’m much more optimistic than others in that this conversion can happen sooner than later. There’s a tipping point coming in the future and the speed of that conversion is going to be rapid. The question is when will it begin?A lot depends on the voice of the patient.
I agree with David… and Leonard in that his concept is *brilliant*. Perhaps @epatientdave or @gfry could comment re: *health grid* concept? It is very likely that patients will drive the “health grid” concept to fruition much faster then anyone else will be able to conceptualize, build and execute a solitary national repository of all health care data.
I think David’s concept is fantastic! And works well with the idea behind #myhealthdata. Just look how easily one can share content of this blog OR many other types of web-based content with a seemingly simple app like AddThis. I had a long conversation with the very young founder of that company a couple of months ago and thought at the time his app/widget should be replicated for use in HC. David’s has just confirmed it! The tipping point will only come when enough people are going to DEMAND access to their data. That’s why we created HealthDataRights.org. Please help us spread the word!
While I appreciate all of the “viral” and “open-source” ideas being tossed around here, I’d like to comment from the perspective of a physician who is up to his ears in healthcare IT. As Dr. Voran above correctly stated, “…the answer to connectivity is right in front of us.” While it is true that Facebook and Twitter have served as great examples of data sharing, and yes, I can now share my family photos with my family in Germany right from my cell phone, unfortunately, there are unique characteristics of healthcare which make a lot of those options unappealing, and thus I don’t share the optimism that social media is going to rescue healthcare IT. (At least not until we have some major cultural shifts in our country.) For example, I have been working hard to develop a communications mechanism to let specialists in our area know when a consult has been requested. You quickly find out how hard this is to do : – If we try simple email as an option : I hear “I don’t use email” or “I don’t use THAT email” or “My nephew changed my email to a different service” or “Is email really secure?” (which it isn’t). – If we try chat-rooms : I hear “That means I have to log in every time to find out if there’s a consult?” and “How will I know if there’s an emergency?” or “Is this really secure?” (which it isn’t). – If we try TXT messages via cell phone : I hear “I don’t have a cell phone” or “Is the hospital going to pay for my cell phone?” or “What if I lend my cell phone to someone, is this really secure?” (which it isn’t). The solutions which appear to work in healthcare, to meet all of the organizational and privacy issues, are just not appealing to doctors. “You mean I have to go out of my way to log into the hospital computer, put in my account and password every time, and click through two menus to find out what’s waiting for me?” Still, I remain optimistic about patients REALLY organizing to help healthcare develop standards for all of this. For example, I have what I call my “Wilford Brimley Healthcare Integration Plan”, which sounds sort of funny, but it sort of explains the premise under which I think patients could help straighten all this out. (It stems from the paradigm shift I think we need : Instead of the government investing through government programs, I think we need to send the $19 billion to Madison Avenue in New York, where advertising and marketing miracles are created.) Here’s the “Wilford Brimley Healthcare Integration” plan.1. Wilford Brimley, star of Oatmeal and Diabetus commercials (which appeal to the AARP crowd) does a commercial on the SUPERBOWL in which he basically states : “Wow, I just got out of the doctor’s office, and I found out they almost made a big medication error because they didn’t know what my cardiologist was doing! And they ordered extra tests I didn’t need because they didn’t know what they did in the emergency department last week! But NOW, with GOOGLE HEALTH which speaks FLOWER, all of my doctors can share my healthcare information easily!” – Why Wilford Brimley? Because he appeals to the AARP crowd. – Why AARP? Because they’re politically influential, and have time to write their politicians, AND spend more time in doctor’s offices. – What is “Google Health”? The PHR, self-explanitory. – What is “FLOWER?” – The protocol by which all doctors/hospitals will share medical information (e.g. .CCD, or .CCR, or .MML, or whatever standard we agree to share information on – I have ideas exactly how to do this… We just have to come up with a patient-friendly catchy name…) 2. Patients start showing up at their doctors office, asking, “Dr. Smith, does your office speak this FLOWER thing?” and at hospitals asking, “Dr. Stanley, does this hospital speak FLOWER?” 3. Dr. Smith starts asking her EMR vendor, “Does our software speak FLOWER?” and Dr. Stanley starts asking his software vendor, “Does our software speak FLOWER?” 4. The vendors suddenly get a lot of press/buzz from doctors/administrators, asking, “Does our software speak FLOWER?” 5. As hospitals start getting upgrades that handle FLOWER, then the next trick : The Universal Patient Identifier. As patients register in a doctor’s office or hospital, the registration clerk asks this : “What is your name?” “What is your date of birth?” “What is your address?” “What is your insurance information?” “Do you have a health portal you would like us to send all of your data to?” – IF YES, then the clerk gives the person a keyboard : – The patient types in their Google Health account and password. ** THIS NOW IDENTIFIES THE PATIENT TO THE OFFICE/HOSPITAL – So medical record number #123523 = “dirkstanley@gmail.com” 6. Google Health (PHR) becomes the portal AND IDENTIFIER by which offices/hospitals can share information. (Google Health has to make provisions to allow “licensed practitioners” to have emergency access.) – To support the patient’s privacy here, the government needs to pass a law with CIVIL penalties for “Unlawfully accessing a patient record” – – To support the patient’s privacy here, Google Health has to make it clear to every patient a “LIST OF WHO HAS ACCESSED THE INFORMATION IN YOUR HEALTH PORTAL” 7. For patients with limited interest in managing their own health : For the price of a GoogleHealth account and password, they can allow their doctors to share their medical information.8. For patients with more interest in managing their own health : They can log into Google Health and watch their records, and even, if they really don’t like a particular doctor, delete them from their health portal. 9. This helps solve the “Big Brother” problem : For patients afraid of a “National Health Grid”, this gives them the ability to “opt out” – None of this is mandatory – No messy politics of “Someone is storing all of our health data”. 10. This also helps solve the patient identifier problem : No messy politics of “The government is giving us all a number that they track us by” and “I’m just a number, not a human being”. 11. This also helps solve the Healthcare IT Investment problem : Offices get to keep their software, hospitals get to keep their software/hardware : No messy “We just invested $5 million, now we have to rip everything out?!?!” politics. 12. This lets offices/hospitals keep their own individual patient identifiers.13. This also creates a distributed model of healthcare storage : If one system goes down, at least other systems could potentially be working to provide SOME data. (See : Why the Internet works as opposed to Compuserve).Anyway, for what it’s worth, I think that’s a better way of approaching standardization. We can all worry about SNOMED and ICD-10, which are important protocols too, but without a good groundwork on which to lay things, we’re never going to finally connect the pieces. (Perhaps SNOMED and ICD-10 could be part of the FLOWER protocol.)
Dirk, “Flower” Power! Uh, you must be my age. I love it!
LOL – I’m 39 – Generation X here, but all Generation Xers had hippy teachers. 🙂 Well, you know, you need some catchy marketing term for this Geneva-convention-like protocol that all doctors/hospitals are going to use to safely trade information. (We want patients to ask for it by name – If the government mandates it, there’s going to be all kinds of messy political strings…America is not socialist enough to handle that.)”Flower” is my rough idea, but I’m sure someone on Madison Avenue could come up with something better.
Whatever the name, I think the security and privacy could be built into the “sharing widget” described above with the system to system private/public tokens that are established when its installed. The installation process could involve any number of identifiers upfront (UPIN numbers, TAX IDs, etc) to validate the underlying system in realtime like we do now with eligibility checking on registration. On the patient’s side the widget could include individual ID’s with validation against existing records and accounts. If there was any government role it would be to be as one of several validation agents required before the widget could be installed or used. All the patient would need is their private token to access their data everywhere. The contributing systems would be automatically sharing data in the background or the patient would have to enter their private token upon registration for that system to collect the patient’s data.I really think all of the tools are present to develop a very secure virtual healthcare data exchange without having to reinvent the wheel or create some monolithic network to manage this data.
If the focus of the EHR becomes the patient, instead of the sacrosaint protection of professional, non-sharable data the entire system will become MUCH simpler. It is only because of the antediluvian requirements when the doc had a quasi-religious interaction with “his” patients that all the energy is spent on the false protection of people privacy. On this, I am in full agreement with Jamie Heywood. Privacy is a fallacy. The sooner we agree on this, the faster we will have highly efficient and useful EHRs. Much education needs to be done.
I’ve been following this (remarkable) thread tangentially for a few days now and was planning on giving a more substantive response to some of the points listed above over the weekend…. But I had to quickly chime in and endorse the recent statement by Gilles.We as a medical community need to wrap our minds around what the issues of “privacy” and “data rights” really mean in a new digital age. It is clear that patients will no longer accept the previous top-down rules for access to their own medical data. I also concur that the idea of privacy that many of us understand is based on a (vanishing) economy of scarcity. In the old paradigm, doctors have the (a) know how, (b) the answers and the (c) physical charts. Patient’s access these items at the behest of a physician gate-keeper and with considerable difficulty. The digital age, democratization of information and technology for instant access to information changes that in ways that we are only beginning to grasp. In an economy of abundance (rather than scarcity) the patient will have options regarding how to handle their own data. The EHR/PHR hold the key to either allowing the physician a partner or an obsolete bit player in this process. For those not familiar with Jamie Heywood, I would recommend the following: http://www.wired.com/techbiz/people/magazine/17-10/ff_smartlist_heywood for context .
Chukwuma, couldn’t agree with you more on several points. As a physician I’ve found it very important that patients see everything in their chart. Transparency is a foundation for quality. Back before our charts were electronic I would leave the paper chart with the patient asking them to browse through it while I finished with another patient. Often upon returning to the room the patient would have several pages turned over. As we reviewed them they would be lab reports or even visit notes of OTHER patients that were misfiled. In short, by doing this we were turning over the chart audit to the patient who always did a better job than the medical records technicians we hired.Now that we’re electronic it’s harder to leave the patient free to browse through the “chart” because it’s not only theirs but everyone else’s chart. Our web portal can be configured to allow the patient to see their entire chart and while a minority of us strongly want this there are too many physicians that do go ballistic at the notion the patient would have access to their entire record so we’ve had to configure the portal to only show the patients results, labs, measurements … virtually everything except the physician’s notes. Those of us that want the patients to see our notes make them available electronically on demand in either secure electronic messages or by simply printing them out at the end of the encounter.When physicians know that 100% of their notes are going to be audited by the patient the notes become more accurate, easier to read and convey more information. Of course, it takes them longer to do this and that does interfere with productivity. The problem we face is no one is PAYING for quality these days. Lip service may be given to quality and care but at the end of the day 3rd party payers (including the government) pay primarily for quantity. So if a physician chooses to spend extra time writing accurate, informative and useful notes it’s usually at their expense. This has to change in order for us to really move forward.I’ve found that transparency of the medical record is a strong motivator for change but the inertia behind protecting the chart in the name of privacy, confidentiality, etc. is overwhelmingly strong and a very difficult for those of us in the minority to buck. Administrators are loathe to stand up for the minority since IMHO their primary motivation is to protect their jobs rather than take chances and move an organization into the future.Again, all of this would change overnight if the payment mechanisms would change.
Very interesting discussion, EAI has been around for a long time and many industries have made it work (e.g., banking). Integration of desperate systems can be accomplished if the parties involved want to make it happen. The problem starts with politics at the ground level inside health system. The ownership of data the next. One of the reasons RHIOs fail is that organization do not what to give up their customer data to competition. Renaming it to HIE is not going to overcome this problem. I consulted on an early RHIO project where the main stumbling block to success was the retention of patients by the primary stakeholder.I like the record banking concept where the patient selects a storage facility the same way we select a bank to store our money. EMR system get access to the patient data via permission granted by the owner, the patient. The patient also pays for the storage of the data. HealthVault or Google Health with some changes, could be that service. The EMR would access a copy of the data to be utilized during the visit. The EMR could store auxiliary information that is not part of the “legal record” in the EMR database. On the “Privacy is a fallacy”; in the late ninety I worked for the first online banking company, CyberCash. We had a saying; “Privacy is Perception”. Very little today is private but we keep telling the masses that things are private so the think they are safe. People need the security of their belief that they are safe and and have privacy. HIPAA basically spells this out as your best effort.This is a great conversation which idea can lead to innovation.Jeff Brandtwww.comsi.com
I agree with all that this thread is basically about – Which is why the “Wilford Brimley” plan, I think, is the best way to do this – Approach it from a marketing, patient-centric standpoint rather than a government top-down approach – The only thing that trumps the government in terms of power is the dollar. (Some call this cynicism – I call it realism – But the intent is really just to help the patients in the end.) I have been trying to get a dialogue with someone from Google Health or Microsoft about this – Since I don’t think they’ve fully appreciated how they could leverage themselves to gain a major marketshare in this – But so far, I’m just one doc, and most people look at me like I’m crazy when I talk about how Wilford Brimley could help integrate our healthcare IT. (The intelligent part of my writing is often lost in my humor.) 🙂 And of course, this plan (that we’re all stewing over here) would bring the cultural shifts that we need to make this all happen : 1. Patients need at least SOME involvement to help organize their healthcare info. (This plan lets that happen for the minimum investment of just creating a PHR account and password.) 2. We need to nationally shift away from the old “top-down” approach. 3. Every patient should decide for themselves : “Does privacy trump safety?” (This plan would let patients decide this.) 4. We need to shift away from the “Medical-charts-are-quasi-religious- in-nature-and-controlled-by-the-doctors”-type paradigm. 5. Patients need to demand a standard for information interchange.Other benefits of the Wilford Brimley plan (we should start to rename it, since I recognize my term “Wilford Brimley Plan” isn’t going to help sell this idea to anyone, so SOMEONE PLEASE HELP COME UP WITH ANOTHER NAME FOR IT) – 1. It would create jobs for people in the PHR industry – Could get good support from the software/IT industry. 2. Doctors wouldn’t FEEL the “loss-of-control” – They get to keep their software and their workflows and their charts and their medical record numbers. Less political pushback from doctors. (Patients trump government AND doctors in this game.) 3. Patients would start to see their hospitals/doctors/specialists offer to send their information to their PHR faster. 4. It sets up a nice mix of distributed and central computing – Having one center, from a data perspective, has enormous costs : – Easier to focus hacking efforts – When one system goes down, there at least would be SOME redundancy from other doctor’s charts (Read : “Why Compuserve is dead and the Internet is alive.”) Obviously, we still need to fix some of the issues about SNOMED and ICD-10 and “Who’s going to keep the master list of problems”, as @epatientdave can well inform people of, but at least having one central way of finding every doctor/ED office records, and then aggregating them, would be a baby step in the right direction. (This is why I think FLOWER would best be served by .MML rather than .CCD or .CCR, although I’m not technically adept enough to fill in those details… yet.) 🙂
So here’s another model (one that I’d proposed many years ago). Google Health or MS Vault are good but I’ve found them cumbersome to use as an individual and it’s relatively heavy work to convince all of the providers, pharmacies and other healthcare providers to upload my interactions with them to both of these entities. What I’d like to see in GH and HV is a system where I log in, identify my location and up comes a list of physicians, clinics, hospitals, pharmacies, etc. in my area. I check the ones that I’ve visited recently. GH/HV fax my signed consent for release of information to those entities. If they are able to export the data electronically instructions are provided. If they’re still paperbound then they fax the requested information back to GH/HV.As this information is received it is then reviewed by a GH/HV individual and the data is normalized and entered into the GH/HV database for me.Of course this is labor intensive but it’s essentially what goes on right now where my patients sign a consent for release of information and we obtain records from previous physicians. Paper stuff (>80%) is scanned into the record and the rest that comes back in CDs, CCR data files, etc. is imported.Unfortunately no two databases (even from the same vendor) are configured identically so most of us take the easy road and import this information as documents. No one has the time to sit down and enter this data into the appropriate database fields for longitudinal comparison.Having someone do this would be worth my time as a patient and is something for which I would pay. However, I’m not sure what we as patients would be willing to pay would cover the cost of normalization so I’m proposing another revenue stream. When I visit a new physician’s office they would log into my PHR and download the data they need. The physician’s office would pay a nominal fee for accessing and downloading this data. Most offices I think would be willing to do this since they’re already paying for the paper and staff time to do de-novo data entry of my data into their systems.Finally, if that’s not profit enough for GH/HV then there’s more revenue streams that could be leveraged, with a twist. I, as the patient would participate in the revenue stream.Here’s how I envision these other revenue streams:- I could chose to expose my healthcare data to researchers either as de-identified data (for which I’d get a discount on my subscription) or as person identifiable data (for which I’d actually get reimbursed by the researcher). Researchers could run reports to see if certain criteria are met on de-identified data for which they’d pay GH/HV. Nothing new here. This goes on all the time. But if they did find appropriate targets a message would come to me asking if I wanted to participate in the study at various levels. The more information I shared the greater the reimbursement to me.- I could chose to expose my medication lists to pharmas. In this scenario a small clickable link beside my medications would allow me to quickly link to the manufacturer of that medication. I would get the click-through payment. If I allowed direct marketing then I would be given discounts, coupons, etc.- I could choose to expose my data to other patients. This would allow me to find others with similar diseases to form peer-to-peer support groups online leveraging the record. Now all of a sudden the PHR becomes more and more like a social network linking the patient, providers, researchers, pharmas and any other entities. The key thing is that it’s all patient directed. Nothing happens without the consent and direction of the patient. But there’s a potential for financial reward to patients who do. Those who want their data hidden will be able to do so but they won’t be participating in the benefits that come with sharing their information.This is doable now but it does take some back end human labor. Any comments?
I think the comparison to social networks is the applicable model, but… The “but” centers on the distinctiveness between how social networks have evolved and your (David’s) suggestion on how participating providers should be recruited. I’ll digress to my comment (supra) regarding HI being a public utility (not a grid, we are in the process of considering that — grid is conceptually too rigid — we’re in the amorphous state still). A public utility or public transportation (similarly; my analogy here: http://symtym.net/2009/11/everything-new-is-old-again/) has two features: content and conveyance (transport layer) that are wedded. There is a public good being sought here — the ability for everyone to have a whole representation of their HI available everywhere to anyone (so authorized). That speaks to a level of engagement beyond mere voluntary participation (fax and faxing are so apropos to the severity of our national HI problem) — to a level of fundamental commitment. If we’re driving to a conceptual individual whole representation of HI, then we’re creating in essence something similar to a life estate — a property right that exists for the life of the individual. Providers become co-tenant (with definable rights) within this life estate. Rather like a member of a social network (majority of models have a wall-garden nidus and persistence, e.g., AOL, FB, LinkedIn, Twitter, etc.).At this point in time, data in and out of social networks is constrained (limited) by the varying degrees of dissimilarity between the database structures and willingness to allow your members to go wandering beyond their spheres’ of influence. But where there is ingress and egress there is alllowed a structured exchange of data (beyond the mere willing to participate, e.g., token = fax). Data portability (with commensurate standards adherence) is becoming (will become) the holy grail for social networks — my data, my way, always… Using the social network paradigm (yck cliche) for healthcare, we should not be interested (or encourage) mere passive (minimally active) participation (again with faxing) — but rather the minimal push should be “you’re in or you’re out” with structured data that is adherent to open standards utilized in our conceptual life estates. Not unlike an individual power utility or manufacture — you have to produce electricity with a certain current and voltage to travel our transmission lines or you need to package your manufacturing with certain dimensions and weight to use our rails…The win-win here would be every patient/client has a renderable whole representation of their HI realtime, all-the-time. And the providers have access to a applicable and contextual rendering of HI realtime, all-the-time.How I would handle security? Limit the permanent instances of HI to 1 or a few. HI is delivered realtime and remains “realtime” for the duration of a contextual process (the ability to do work, e.g., billing, coding, treatment, etc.). Like electicity delivered by a public utility — the need is contextual to a process, once the process is terminated there is an off state. We should secure HI by limiting the number of permanent instances and define strictly the temporary instances of HI storage. Contrary to this, we seem to be going and funding a direction where the incomplete instances of HI are truly going to be innumerable. Contrary to the evolutionary direction of social networks as well — where the trend is looking like we will be able to “take with us” a whole representational rendering of our social attentions between social networks.We have the conceptual (current) model (social networks) and we have the historical models (railroads, public utilities) — we need to desparately abandon the healthcare model that has NOT served us for so long.
@Loic was interviewed on French TV a few hours ago and said “If faut arrêter de protéger le passé à tout prix”. Simply translated, “We must stop protecting the past at any cost”.Be careful guys! When 2 frogs are saying that the old model is not worth protecting and that building the system of the future requires yanking out the myriad of preconceptions accumulated as the system aged and got sclerotic, we know we are in trouble.@loic comment is, I believe, on the spot. That is why I constantly make fun of @jhalamka who has demonstrated an absolute inability to say that much of the stuff they have tried just doesn’t work and has to be considered below junk grade. Until it happens, we will all suffer from this morbid wish to always be compatible with ALL of the past. We must take a lesson from Apple. A few years back, a decision was made that OS X would no longer be compatible with OS 9. Screaming ensued and all the defenders of OS orthodoxy fumed and warned of doom and worse. A year passed and no one ever heard another word of defiance from these diginosaurs, simply because in a fast moving world of software dropping hooks to the old days is a must.
Agree, deux grenouilles rendre droite. As a fellow (at least trained) emergency physician, I appreciate @halamka need to resuscitate…but there comes the need to recognize a prolongation of life v. a delay in death.
It is always easier to start over. But, I think we know that that’s not going to happen. Take the paper charts, what a pain to get them into a digital system. Our phone system is one of the best analogies. The entire system is backward compatible, My father still uses his analog dial phone. Mobile phones when through a transition a few years ago when they dropped analog phones in most area. At one time I said I would never want a totally digital phone because they didn’t work in a lot of areas, but everyone at the time supported analog cell phones. Today that to has changed. If you are going to make a disruptive change to a system it must be planed into the future. If planned, the disruption is very low (e.g., 2000 or digital TV).Jeff Brandtwww.comsi.com
I doubt there are any examples of a disruptive technology or innovation that has occurred in “the future” (where there is a perceptual planned awareness). All have been recognized only with retrospection and never planned — the very nature of disruption. Should there be a period of legacy — sure, but there’s a difference between tolerant allowance and engineered (indefinite) inclusion.
Tim. Great comments. A la Hernando de Soto (http://en.wikipedia.org/wiki/Hernando_de_Soto_Polar): thriving economies (in this case a medical information economy) are built on strong ownership rights such that people can trade with trust. Strong lines must be drawn with authorizations clear and enforceable without limiting sharing when it’s needed or when users don’t take an active role in managing their data.Sounds like a business model to be tested. A PHR/EHR whereby organizations can subscribe when someone licenses their information. Need to see if anyone will buy.If HV and GH won’t do it, why not build it?
Absolutely fascinating conversation. Thank you for your contributions (Howard thanks for leading). Knowing you are unbelievably busy but absolutely committed to pushing the envelope I would like to propose a Skype call to explore Dirk Stanley’s “FLOWER” plan as a starting point. Please DM your coordinates and I’ll make it happen.
LOL – I’m flattered. I’ve been trying to figure out how to get in contact with someone at Google or Microsoft just to discuss this informally to see if the “FLOWER” plan has merit. Anyway… Would be glad to engage in a skype conference with you guys, if you feel the idea should/could be fleshed out more.Areas that would still need to be discussed :1. Concepts how FLOWER would work. (I personally don’t understand enough about .CCR or .CCD to know whether they would function completely to serve this purpose, but I personally really like .MML, which never really caught on in America but is widely used in Japan.)2. A business model to attract Google/Microsoft (I think this is not too hard)3. A more detailed plan : This is why only Google or Microsoft could do this. If they develop FLOWER in conjunction with some big EMR (e.g. PracticeFusion or Meditech or whatever), they need to find a partner to do this with so when the commercial hits the Superbowl, at least some people in the country will have quicker access to Flower.4. A more detailed marketing plan : Part of the trouble with integrating all this on a national level is the issues are often targeted at one audience : Doctors, OR IT people, OR Administrators. Marketing needs to be developed that targets ALL of these players, making FLOWER understandable. (E.g. the problem with Informatics is, quite frankly, so few people understand what it is, and it’s almost impossible to explain in a few sentences… We want to avoid that with FLOWER.) Anyway, if you guys are interested, I’m certainly game.
I’m game too…I think we’ ve got something here. HJL914.715.9598hjluks@vzw.blackberry.nethjluks@gmail.com
Dirk, as a Computer scientist who has been working in HC for a while, a few suggestion as to get the Flowers off the ground from a design point of view. Your comment about the CCD CCR; You should not need to know the low level details. Domain experts (caregivers) need to tell IT (Software engineering architects) what they need. Let the geeks figure out how to get you what you need. One of the problems with HIPAA is that there are mandates on low level protocol in the law from 1996. From a technology point of view this is a big problem.First thing to do is: Develop a requirements specification, Use cases are one of the best. That is, from a users of the systems point(s) of view, what do you need the system to do. Example, you make a request and the system return the answer in the form that you need. I have worked in many domains and healthcare is one of the most challenging. There are many players (actors in use cases) with special needs. One size does not fit all.HIT system are no different than any other enterprise software in use today. They just need to be designed correctly utilizing proven software engineering methods. The requirement must come from the domain experts not engineers. Also remember that the term IT is highly overused, ( the person that sets up your PC is IT. Software engineers are what is needed to develop software. Great conversation,Jeff Brandtwww.comsi.com
Very interesting everyone, The thing to remember Dirk about Google/Microsoft in relation to Practice Fusion/ whichever emr, is that there are competing interests, Practice Fusion’s PHR game is eyeballs and advertising revenue ( Same as Google) The reason our house call affiliates are on Practice Fusion/Patient Fusion’s platform for house calls is that these two interface with our ecommerce platform, social media and marketing sites to provide complete transparency for consumers, and elimination of essentially all staff for primary care physician’s utilizing our platform. So PF’s product is COMPETITIVE with Microsoft /Google contingent for CONSUMER eyeballs. And each of these PHR products are racing to the finish line, for the most part without involvement of practicing physicians. ( The new physician workforce in primary care will START with transparency, PHR/EMR, participatory medicine and customer service, I predict) My other prediction is that by the time we all finish this post, PF will be acquired, who knows?? Dirk, I agree that taking marketing and consumer approach is the essential step in transforming health care. We ( providers, suits, HIT contingent) in health care have spent ( wasted ) so much time dealing with the third party payor systems we have all lost sight of our customers. The health 2.0 consumer movement is moving at light speed, and if we caregivers don’t pay attention, patients will be getting the majority of their health care needs from the web, and we will be looking around with empty pockets wondering how we all went out of business. ( hospitals included) With regard to reform, at the Mayo Clinic’s Transform in September I had the chance to chat with a contingent from medicare and when he commented ( regarding taxpayers paying for reform) he said ” We just write the checks” So I have long given up on reform… For our business the EMR/PHR being one and the SAME is a critical step. ( It’s the same information, right?) The connection of the PHR to our consumer facing site is the second step. Providing lots of great health content ( for consumers) and community ( for consumers) is important. Good luck everyone, these kinds of discussions spark great companies, and then the question becomes how are you going to create your business model, who is your customer, how are you going to execute? There is certainly room for more innovation in health care right now. Nataliewww.personalmedicineinternational.com http://www.linkedin.com/in/nataliehodgemd
Natallie, The stickiness of patients is a very good take in the PHRs . An interesting thing that I haven’t thought about is that the PHR could be the driving force to facilitating capturing and retaining patients as well as advertising of both internal and external products (e.g., pharma and wellness).Jeff Brandtwww.comsi.com
Jeff, interesting perspective but I font know if it should be taken literally. In my very biased view PHRs are all but tools to “capture and retain patients”. PHRs are freedom, empowerment and personal responsibility tools for people interested and engaged in their own health. Could you please expand and/or clarify? Thank you. Carlos
I would like to comment as a patient (although I have a background in Oracle database design and programming)… I have spent parts of several days reading this post and the comments following it. Frankly, this is the first time I’ve felt real HOPE something productive is happening. Until now, frankly, I’ve seen/heard a lot of “talk”, but there just hasn’t been much meat to go with it.So, to all who have participated in this discussion, I say a big THANK YOU. And most of all, I want to encourage you. If I can do that in any way, please say so. I’m not much of a baker, but I’ll even send cookies. 😉 On the other hand, what I need to know, as a patient is how I can get my local physicians involved as well as the “non-profit” almost-monopoly (http://www.roanoke.com/news/roanoke/wb/228316) which owns/runs most of the medical practices and hospitals here. I don’t have other choices in healthcare unless I travel (which, incidentally, I do…all the way to Los Angeles from Virginia), thus making the PHR a market tool to “capture and retain patients” is fairly meaningless here. One needs viable choices before that will work.
Carlos,I was reading Natallie’s post when a light when off in my head about patient stickiness,which can be a good thing. Yes, I agree that PHR do facilitate freedom and empowerment… My company is in the mobile PHR business. Companies like Google and Microsoft have PHRs services that are free to providers and patients. Their revenue has to come from advertising and data mining. Google, the leader in free tools captures their users by offering free tools and enticing them by adding new features and content. It doesn’t take long for the customer to see that it is easier to stay than leave, stickiness. The tool you know is a the best.By providing EMR and PHR that are linked together does provide more stickiness and if you control the content the relationship with the patient becomes stronger.When selection a PHR from the patients point of view the critical criteria is the ability to export the data from the service. If portability is not an option the service is not a true Personal Health Record.I like the health banking solution where the patient owns all of their health data. The patient is responsible for the cost of storage and access.Jeff Brandtwww.comsi.com@jeffbrandt
Yep, physicians in primary care don't realize the power, influence, and value of their eyeballs on the web. In our web platform for house calls, this is the mechanism of revenue for our EMR/PHR piece of the platform, hence, even LOWER practice costs for physicians. ( Our model results in 80% discount on operating overhead compared to MGMA averages in primary care. ) Patient Member access to their medical record through our site is just icing on the cake for our health consumer members. And once you've had your doctor in the home, you can't go back to the old way.
I sense a challenge of sorts is soon to emerge as our comments start to focus on various conceptual *platforms* (FLOWER) (Fluid Medical Grid) which could be white boarded and begin to take shape. With the help of some true fire starters and innovators, as well as the input of some high quality IT folks who have commented here, we have the team to push this further. Those who are willing to join the likes of Carlos, Dirk, Gyles, Dr David, eDave, Natalie, and myself who will try to morph this from a discussion to something more meaningful please ping us. First Skype this Thursday.
“patient is responsible for the cost” and “doctor in the home” are only solutions for very few patients — the viable universal PHR (lack of a better term, I prefer PHI) must be funded by sources other than the patient (you’ll disenfranchise millions) and must be unlinked to the notion of PCPs available in the home — cost and limited/diminishing numbers of PCPs in the US. If PHIs are considered a public (for all) utility, then cost is borne by a utilization fee (providers). Nonsense? Well I’ll bet the cost per utilization in a universal PHI construct (may have more than 1 vendor, but common platform) will be significantly less than provider coming up with their own solution that won’t be reimburseable. Linkage of PHR with EHR is only as good as the funding relationship (insurance, uninsured, job, etc.) — millions are disenfranchised and millions change funding relationship annually. Change in funding, as the root associated link between EHR and PHR, will always lead to fragmentation on PHI.
If the conceptualization of PHI (personal health information, a person’s whole health information set, obviously HIPAA’s version of PHI is an inclusive set) as a “life estate” (property interest for the life of the individual) has merit and legs in this discussion, then its existence is defined not by “relationship to” (to provider or payer) but by the “existence of” the individual. Participation (co-tenancy) within an a particular PHI is contextually defined by the provider, payer, etc. The default setting: you have a PHI (platform provisioning down-the-road discussion and ferreting out), default setting: you own the rights, default setting: to pariticpate in health system you must allow contextual access to the “co-tenants,” other default settings…Changing focus, funding for PHIs might be divided into early/start/ramp-up and later/sustaining. Early: advertising (a la GH and MV). Sustaining: secondary and tertiary data markets, savings on IT installs/maintenance etc. (in-house), with sufficient scale no need for HIEs and RHIOs, etc. HIEs/RHIOs supplanted by the health internet and device/browser maturations. We don’t need to create in software and hardware what already exists — retasking/repurposing/securing will be significantly cheaper than new developments, installs, and perpetual need for maintenance/upgrade/etc. A choice between always utilizing existing leading edge v. always amalgamating legacies.Again focus change, it would seem you drive adoption not at the provider level, but from the constituency (-ies) level — patients/clients and their advocacy groups (e.g., AARP). The providers are (except early adopters and academic centers) and will be dragged into EHR-era in a reactionary manner (carrots and sticks) with piecemeal implementation over indefinite period of time (ARRA/HITECH seeks to compress this, but does anyone truly believe the time frame for massive implementation is feasible?). The constituencies have a prime time of opportunity (IMHO) to take a different route and drive downstream adoption by providers. Wonder if there were pushes for the concept of a PHI, one that is a personal property, that exists independently from providers/payers, one that has privacy/security as fundamental constructs, and has the ability to be accessible now by anyone given access on devices already in existence? The components are already here (some agreement and assemble is needed): syntactically/semantically encoded content + TLS (HTTPS, XMPP) + browsers.
Perhaps this is stating the obvious, but there is at least 2 major visions (compatibility?) for PHRs (again I prefer PHI) in this comment stream. First, the vision of a the PHR as a dependent rendering from an existing EHR/EMR — always a “subset of” what constitute a contextually “whole” record (context = provider). For the sake of discussion a contexual PHR (cPHR). Second, the vision of a PHR that is the whole (or near whole) amalgamation of all EHRs/EMRs (patient is a “subject of”) + social + personal derived materials. For the sake of discussion a wholistic PHR (wPHR). In discussions and formlations cPHR is a derivative product and will always be a subset of the integrative product the wPHR (the true set).With this description it is easy to see multiple putative “products” arising from these discussions — encompassing cPHR and wPHR offerings. In my thinking the cPHR, long term, is just a variation on a commercial inducement. The wPHR, long term, is the economy and currency we should be shooting for.PHR math:wPHR >/= SUM cPHR
One of the problems we’re struggling with is that unlike a bank there is no medical equivalent of reconciliation and balance forward. This makes it much more difficult for one entity to transfer the “medical balance” to another entity. Banks can just send the balance instead of every transaction. We in medicine are relegated to sending the each transaction to others in order to get a “complete” picture.The second problem is the huge uncertainty factor. Each one of our systems (PHR, EHR, EMR) that we use have a place to put in a patients Diagnoses and Problems which some may consider a high-level summary of a patient’s health. Unfortunately we need to make diagnoses on each visit and in the ambulatory world up to 60% of the time we haven’t determined the diagnosis because we’ve just started the work-up process. So when someone reviews a list of diagnoses all sorts of interesting things that have been ruled out in subsequent visits are found.Thirdly, no one is paid to update and ensure the validity of these Problem lists. So some one may come in for skin rash, muscle aches, joint pain, abdominal pain (along with appropriate codes) for several visits and a year down the road a diagnosis of lupus erythematosis is made. Who’s going to back and update all or most of the previous diagnosis and problems, changing them to SLE? No one. Certainly not the physicians nor the patient (heck I use 3 PHRs and haven’t taken the time to reconcile them) and certainly not anybody in the back office. So the data in the systems is suspect. And besides our attorney’s and coders state that you really can’t do that as if you did go back and change the diagnosis you’d have to resubmit the bill or expose yourself to accusations of fraud and abuse.Having said that there is validity to longitudinal lists of diagnoses over time so maybe there is no need to go back and update the diagnoses. We may just need to link the previous diagnoses. That, in itself, maybe more diagnostic than a worked-over Problem List.I laid awake long hours trying to think of ways to get around these problems and haven’t yet. When I bring this up in our Physician Advisory Group we can’t get a majority of physicians to agree on what is a “Problem” and while we can get slightly more than a majority to agree on a “Diagnosis.”Our web portal allows patients to either create their own PHR (and then give their physicians access to this, look at the data in our EMR or check both options and see a combined view. In the last 2 years not one patient has bothered to create or maintain their own PHR (me included). Why? None of us have the time.We need a way to have information in one automatically flow into the others. Each system much be able to filter the data based on who entered it at a click of the button. The best way I can think of doing this is to have a virtual PHR that is a composite of all of the data from multiple EMRs that the patient, physicians and others can view, slice and dice at will rather than transferring data from one system to another.One thing I do know. Getting the patients eyes on the record does nothing but help. I can’t tell you how many times that on difficult cases the patients are the ones that have come up with the key finding, question or actual diagnosis that winds up being the “correct” answer. We just need to get to the point where the patients along with physicians, nurses, pathologists, technicians are seeing the same data and can communicate freely.
One of the longest — and smartest — threads I’ve seen in a while.Howard, please count me in on Thursday.Does GH and MV keep coming up as obvious PHI carriers (in the same way AT&T is a carrier of voice traffic) because they are already 800-lb gorillas? Any nonprofit models out there? PubMed, for instance? Or a creative commons model? Or eBay? Or even bittorrent?I mentioned early in the thread that licensing of one’s PHI seems like a good model (http://hitshrink.blogspot.com/2009/11/licensing-ones-own-health-care-data.html). That keeps the data under control so that it can’t be repurposed, resold, or reused without permission — an important assurance for consumer acceptance.I REALLY like Dirk’s grassroots model, the FLOWER or Brimley plan, which I think of as the Patient-Powered health record. Adding in social media aspects makes a lot of sense, sort of a Facebook for health care. What I particularly like is the idea that this large set of population data can be used BY PATIENTS to see where they fall within their peer group (however they define that) for various data elements. Join a group of people with similar medical problems for education/support? No problem. Let them access my records and give me ideas about my meds, side effects, activity level? No problem. (eg, PatientsLikeMe)But as a physician, I’m thinking “What would entice me to access someone’s PHR?” I’m an inpatient psychiatric consultant on the medical floors. I spend an inordinate amount of time playing Sherlock Holmes to find out what meds someone is taking, when they stopped, which symptom or side effect came first, talking to the patient’s roommate or spouse or daughter to find out what happened when. If even SOME of this were on a patient-powered PHR, that the patient gave me access to, it would (could/should/might) save me time; would certainly give me better info. Would I pay a dollar to access it? No. Not the money so much as the invasion–putting in my credit card info in a public PC on the floor. No thanks. No, this is a mutual benefit (I do my job better, pt benefits from that) so there is no need to exchange money for the patient to license to me temporary access to their data. But pharma and marketers and insurance companies? These are more one-sided relationships, so I as a patient can decide what I think it is worth for them to access my data, if at all. This can be done using an auction-type model.I think it was David who mentioned other revenue streams. Insurance companies giving premium or co-pay discounts to members who provide access to their PHR makes sense.And researchers, as David mentioned. Think of the pools of data that can be collected. Say a researcher wants to review longitudinal lipid/weight/BP/cardiac/glucose data on at least 2,000 people who were on olanzapine for at least 6 months over the last 3 years, stratified by family history. The researcher puts out a notice on the Research Board, which patients can subscribe to, or you get a message in your inbox based on your profile and you choose to participate. The researcher bids $1000 to access these records, and a group of patients pools their bids together, eBay-style, until there is a match. Researcher pays via P.O. or credit card, and the money gets split up evenly among the participants, minus a transaction fee for the processor.The same goes for the drug company, the device manufacturer, or other marketer. Finally, design an open-source API so that others could hook into this, either to add functionality or to build interoperability into the Epics and Cerners and Practice Fusions in the market. Maybe build in a credit system so that when data flows in from these other sources, a credit is provided. This credit can be used to buy access (license) to the other data sources for that (or another) patient, or for de-identified population data. Like good citizen, peer-to-peer network model, where the amount of data you get to download is related to the amount that you upload.Looking forward to talking more about how to build such a vision.=Steve Daviss MD (@HITshrink)
Steve there are the multiple reasons we have discussed for physicians to enable patients to access their PHR. In our model, one of the reasons our house call affiiates use this and it is integrated into our platform, is it enables us to work without staff. ( our patients have access to all of their health information all of the time. So they don't need to call for it. ( When is the last time you called your bank to check the balance? ) So a reduction of overhead of 80% compared to office primary care practice. It turns out that transparency eliminates the need for us to have paper pushers and people to manage the paper pushers. And although our marketing/ecommerce platform is cash pay by patients, there is not a single reason these cost reductions couldn't apply to medicare/medicaid primary care contingent. Frankly, the whole thing could be used as a backbone for a National Health system in which the government pays a salary to the primary care doc. If there are any takers out there, Mr Obama, are you reading this? We are happy to negotiate on behalf of our physicians. 80% reduced costs is a big deal for primary care. There are a few companies already working on a FB- like interface… I'm happy to discuss with anyone regarding this… Exciting discussion everyone. Natalie Hodge MD FAAP http://www.personalmedicineinternational.com
The discussion/planning seems to have moved over to Google Wave – We invite people who are interested to join over to the wave and see what’s going on.Just to wrap this up, I’ll copy-and-paste what I wrote as the origin of “FLOWER” – Thanks everyone for the awesome discussion, and we look forward to more thoughts/input in the future! :THE HISTORY OF FLOWER :—————————–About 4-5 months ago, I had an after-hours conversation with two of my colleagues/friends at work, where we are all up to our ears in Healthcare IT problems. Myself, Kipling Morris, and Nicholas Boisjolie.We started to discuss, “Why is healthcare IT such a mess?”And we saw that a lot of the chaos comes from the multiple private interests :1. Software vendors have their own interests.2. Hospitals have their own interests.3. Insurance companies have their own interests.4. Physicians have their own interests.As a result : There are almost NO interoperability standards. THe only standard we have today is the fax machine, for a whole bunch of reasons.ANYWAY – Then we thought about the “power hierarchy” of healthcare :1. The physicians are sometimes independent, sometimes controlled by group practices.2. The hospitals are mostly controlled by regulatory issues and insurance companies.3. The insurance companies are mostly controlled by regulatory issues.4. (It gets complex)The strange thing we noticed : There is one person here who trumps ALL OF THE ABOVE : The PATIENT.And then we asked the question : So if the PATIENT has the ultimate power, why doesn’t the patient control what’s going on?The reason : Most patients have NO IDEA what the problem is, and even if they did, even fewer would have any idea of how to fix the problem. And even if they did, how would you get them organized to drive a solution?So in our after-hours conversation, we started to ask : “What if patients could be steered to demand something?”And then we talked about the phenomenon of “Healthcare Marketing” – Any doctor will tell you the power of Drug ads – One commercial for “Requip” (Ropinirole), and suddenly patients show up at the door asking you to treat their restless legs syndrome.So then Nick, Kip, and I started to imagine : What if we had a commercial that emphasized, to patients, the importance of having their doctors share their information. We came up with the idea of Wilford Brimley delivering this commercial on the BIGGEST venue in America (e.g. the Superbowl), where he introduces several concepts to the American patient :”You know, my primary care doctor almost ordered something that interfered with something my cardiologist ordered – Because he didn’t know what my cardiologist was doing! And in the emergency department last week, they almost ordered a test I already had, because they didn’t know what my primary care doctor was doing! But now, with Google Health and FLOWER, all of my doctors can share their information!”This would introduce the American patient to some concepts :1. Medical errors and redundant tests – Unfortunately, this is a difficult subject to discuss – And for good reason. Nobody wants to create fear. 2. The value in avoiding medical errors and redundant tests.3. A way to avoid medical errors and redundant tests. (“FLOWER”)So then Nick, Kip, and myself wondered : What if every patient suddenly started to show up at their doctor’s offices, and hospitals, and asked, “Do you speak FLOWER here?”1. Office docs would start to ask their EMR Vendor, “Can we do this FLOWER thing?”2. Hospital docs would start to ask their administrators, “Can we do this FLOWER thing?”- Administrators would start to ask their EMR vendors, “Can we do this FLOWER thing?”3. Vendors would suddenly get swamped with phone calls, “Can we do this FLOWER thing?”4. Vendors, to stay competitive, would need to build FLOWER into their EMR.(We came up with the name “FLOWER” as a catchy, cutesy-type marketing name for the protocol that all EMRs could speak to a central PHR) – An “Esperanto”, or Lingua-Franca that could be developed as an add-on for virtually all EMRs (after the technical standards were built), so that virtually any EMR could easily transmit the patient’s medical information to their central PHR.Currently, Google Health and Microsoft Vault have some of this functionality, but it’s mostly with big university hospitals – E.g. Beth Israel Deacon, Mayo Clinic, Geisinger, etc – And that’s only because Google/Microsoft partnered with these centers to develop custom interfaces that work (generally via .CCD or .CCR) -The problem is that for the rest of America, they are faced with the “other way” of getting their records into Google Health/Microsoft Healthvault – Sending in their paper records to have them scanned by a company that will charge them. Who wants that?So not a lot of people currently use GH/HV for their PHR – And, as ePatientDave can tell you, even those who do are sometimes disappointed with the results.Now, this is where it gets complicated :1. There are legal issues (e.g. HIPAA) which make some of these electronic transactions difficult.2. There are technical issues that still need to be addressed :a. What protocol could be used to effectively transmit the data? .CCR? .CCD? I personally like .MML because of its simplicity, which is widely used in Japan, but nobody here in America seemed interested in it. Why .MML? (Medical Markup Language?) – Because almost any doctor can look at the actual code and figure out what’s going on with the patient. If the protocol is that transparent, I think it’s a good thing.b. How would a PHR aggregate the data effectively? (This is also difficult, since most people will tell you about the challenges of having a decent “problem list”) -So we don’t have all of the answers – yet – But then we thought, “Why hasn’t anyone else tried this approach to healthcare IT standardization?”.So after our chat, we all went back to our day jobs. And I tweeted about this a few times, under the “Wilford Brimley Healthcare IT plan” – Essentially, to test out the paradigm shift : to invest the money in Madison Avenue rather than ONCHIT – With, what potentially could be better results.The problem : My tweets went largely unnoticed – I don’t really have experience in making companies, and really my first thought was : Why wouldn’t Google/Microsoft be interested in this? A simple change in their perspective, and they could leverage the American patient to their benefit… And help organize healthcare in America.I tried, in vain, to find contacts in Google/Microsoft, to see if anyone was willing to talk about this – No luck. I have a bunch of “Thank-you-for-applying” letters from Google.Anyway, then came Howard’s discussion board, and people were having frank discussions about their problems with healthcare IT. So I re-posted the “FLOWER” concept.And for some reason, this time it stuck.And the cutesy-catchy “demo” name we came up with, “FLOWER”, seems to have struck a chord :1. Flower seems innocent enough – I can tell you that any solution that is going to tie together so many pieces of the puzzle and step on so many people’s interests has GOT to be politically neutral. Who can argue against a flower?2. Flower has a whole “blooming” concept – Plant the flower, and watch it grow. Branches and roots.3. Howard mentioned that the Baby Boomer generation (A BIG consumer population) will like the idea of “Flower Power”.So that’s why we’re here – To discuss Flower. (At least, until some marketing person comes up with a better name.) 🙂
Thank you for that explanation, Dirk! I read it prior to coffee and could still understand it. I believe there are a lot of patients who already are asking the right questions and will latch right on to FLOWER. I’ve been part of umpteen discussions online which circle all around this concept, but they fizzled out. I sure hope this doesn’t fizzle out. If you need a guinea pig patient, I’m willing.
If anyone needs a Google Wave invite, I think I have a few. Email me. =Steve
Me too. HJL914.715.9598hjluks@vzw.blackberry.nethjluks@gmail.com
I’ve got 2 invites to deliver. Have Google Wave but have had only one successful wave. The 1st two that leave me their addresses I’ll extend the invitations.
Can’t find the blog but wasn’t there a request for a Skype meeting on Thursday? Don’t know whether the logistics will work out but if they will I do have a free Skype account and have it loaded on my machines.
Yep, simple, and to the point. I am confident in the notion that directing the benefits of HIT to CONSUMERS is an important and overlooked piece of the HIT puzzle and that it will contribute to the success of our nation's heathcare. We ended up going entirely the consumer route and left our third party payor relationships entirely, so certainly we want to be involved and help the concept move forward. Dirk I would join eric ries " lean startup group" which follow's eric and Steve Blank's Customer Development model for business. It will help you clarify the model and who your customer is. Steve says the customer is the one who pays you the money. ( ecommerce, ad revenue, whatever) Strongly Consider a technical co founder to help bring your vision to life on the web. I'll float the Flower concept next time I talk to Bob R. ( CMO) and Ryan ( CEO) over at Practice Fusion, as well. Keep me posted as we have a group of doc's getting to house calls scattered over Chicago, Michigan, Florida, Kentucky, and we could easily tag Flower in with our direct to consumer marketing for our physicians. Best, Natalie
Well, I keep getting more invites to hand out. I can’t get rid of them all. I have 22 right now to hand out if you know folks who need them. I’m consistent… I’m staticnrg there, too. But, I believe Google is now giving them to just about anyone who asks.
Reviewed Dirk’s summary. Found one hole; in the US the patient isn’t the customer. Often patient and physician agree on a plan of care only to have it ammended by the payer. Concierge practices don’t have as much of this burden but at some point (when expensive tests are needed, hospitalizations, etc.) the payer butts in. We do need a system that returns the decision making to patients and clinicians INCLUDING frank discussions about the cost factor. Over the last 30 years medicine has really priced itself out of the market and we need to figure out how to correct this.How? Am not sure but somewhere along the line the real customer, the patient, have to be making financial decisions and begin bending the cost curve downward. I’m hopeful that companies like Med Express Labs (www.medexpresslabs.com) will begin lowering the cost of routine testing.A thought analogy would be to ask what would the price of eating out be if we had restaurant insurance and only had to pay a co-pay to go eat out?Concierge medicine does take the first step in realigning the transactions but most people would say it’s unsustainable unless subsidized by someone for large sections of the population. I can see scenarios where primary care is concierge based and emergent and acute care is all covered by catastrophic insurance but in this scenario I know one or two patients would still consume 70-80% of a primary care giver’s time leaving the majority of the practice underserved.I’ve exploited technology about as far as it goes but right now am being thrashed because I try to keep people out of my practice (thereby severly adversely affecting my WRVU’s and productivity). We’re working to correct that but only 3 of 26 payers will reimburse us for this type of activity. Managing the denial of claims from the 23 payers is eliminating any gains so we’re now looking at a retail approach to see if that’ll work in the midwest where the population density is pretty sparse compared to the coasts.FLOWER technology will make it easier for the patient and me to make decisions but I’m not sure it’ll do anything to cover the current and near future costs of care.
David wrote: “I’ve exploited technology about as far as it goes but right now am being thrashed because I try to keep people out of my practice (thereby severly adversely affecting my WRVU’s and productivity).”THAT is the crux of the problem: our current system incentivizes all of us who make a living providing health care to maximize the number of “high value” patients (where Value = Potential Gain / Work Effort, where Work Effort is a function of time and resource utilization) and to minimize the number of “low value” patients. It’s ugly, but that’s the reality. I should add that Potential Gain is mostly a function of payment but is also a function of intrinsic satisfaction, which the professions of medicine, nursing, and others instills to varying degrees of success within each of us. Dictionary.com defines “customer” as “someone who buys goods or services, especially from a shop.” To be able to buy a service or goods, one needs to know the price for the thing, know the price of the alternatives, and know the relative values which that thing provides. We have shielded patients from this information. Only by putting them in the drivers seat can they truly be a customer. They cannot learn to drive unless they are given the info the need to make decisions.”Flower” would give them the information upon which they can BEGIN to make “purchase” decisions. They can make decisions themselves, and they can ask for help. They can permit others to access their data, even groups of other people with similar health situations, using the wisdom of crowdsourcing.So, who would have thought that financial customers would actually PROVIDE access to their usernames and passwords of their banking accounts, their stock trading accounts, their 401k’s, to a site like Mint? Just so they can get better information and additional value by throwing their data into the pool? Yet they have.Why couldn’t the same thing happen for health care? Or, at least, let’s talk about the pros and cons.
lowered lab testing is helpful, I'll check out medexpress, and guess what there is a new clia waived lab analyzer that they use in the military that we are trialing with some of our affiliates, the machine costs 10G for the provider, but it essentially eliminates virtually every lab cost. Pretty cool… The thing about primary care is that we almost never need to do expensive tests, ( really). Direct Medical Practice affords the physician reduced malpractice risk and malpractice costs, which we feel translates to reduced " defensive medical testing". Frankly, it's the improved physician/patient relationship in direct practice that reduces the liabiity, improved communication. Not that we are better doc's or better clinicians… Why should payors NOT be left out in the cold on this issue? Why have hospitals and caregivers let themselves get to the brink of bankrupcy by adapting to the whims of insurers ( MEDICARE and MEDICAID INCLUDED) Why do physicians and hospitals continue to enable the pharmaceutical industry, with it's 300$ topical steroids? and expensive antibiotics? For care of the poor and elderly why couldn't a mini " National Health" platform be tested, in which government funding pays a salary to the doc in this drastically reduced overhead model? Dirk, we can help with promoting FLOWER to consumers through our site, easily, looking forward to learning more about how we can help there. Natalie Hodge MD FAAP http://www.personalmedicineinternational.com
It is already happening, It has happened, I should say. There is an entire web industry built around this notion It's called health 2.0. And is is moving very fast. As fast as EBAY, GOOGLE, and the rest. We're talking about the speed of the internet. Physicians must get on board with consumer transparency, customer service, consumer internet services, participatory medicine or we will be outsourced. We are losing our customers due to our third party payor relationships and rigamorale. Just look at Take Care Health and Minute Clinic… they have grabbed over half a billion in revenues from primary care doctors. WE WILL NOT GET THOSE REVENUES BACK if we keep doing what we are doing.
Dr. Hodge, am seeing it happen (as I’m also a collaborative physician with CVS’ Minute Clinic and am pushing many of my routine patients that way) but am worried about the applicability of some of the models as our generation gets older and sicker with chronic care issues (which MC, Take Care and other retail services shy away from).Theoretically we could drive down the cost of testing by dramatically increasing the use of point-of-care testing and volume (which is one of the conondrums of lowering the cost curve) it’s still not cheap. For example, many of the point-of-care testing equipment, while costing under $25K for the device, the cost of the test packets onto which the specimen is placed cost $12 – $60 each. So the cost to the provider for can be $50 – $100, even for point-of-care testing in order to return a profit for the provider and many labs are competitive with those costs.Another approach is to open up all of our non-invasive testing to the public. That should lower the per-unit cost of the tests but in order to do so we’re actually increasing the total medical expenditures from a national point-of-view.I suppose it doesn’t matter if the consumer’s choices are driving this but then where are we? Have we done anything to help the population health?Many would argue that we’d get better Healthcare bang for the buck by investing in education, workplace reform, green environmental initiatives.Of course, we could pay for everything if we’d just make all medications over-the-counter including scheduled drugs. I doubt of the local pusher could compete with Phizer and most people would rather buy clean narcotics at the drug store that on the corner or from a friend. The tax revenues here would more than cover the cost of many services. Furthermore, when OD’s appeared in the ED we’d have a better idea of what’s in their system (instead of the dealing with whatever the narcotic was cut with or contaminants in homebrew).Sometimes I go in circles and it’s easier to sit back and watch the natural evolution occur around.
I’ve been part of, or have followed some excellent discussions in the past. I’m going to link the ones I can remember because they might be (or might not be) pertinent to this discussion:Anne Marie Cunningham (@amcunningham) in the UK has an excellent discussion on her blog about why what one calls a “patient/consumer” is important: http://wishfulthinkinginmedicaleducation.blogspot.com/2009/10/why-i-think-i-am-patient-and-not.htmlThe discussion in comments of one of Dave’s outstanding posts: http://e-patients.net/archives/2009/04/imagine-if-someone-had-been-managing-your-data-and-then-you-looked.htmlPost, links, and discussion here (my blog): http://survivethejourney.blogspot.com/2009/06/my-health-care-records-my-property.html Sidenote: Dr. Rob Lamberts and I have had some fine discussions on twitter, on his blog (distractible.org) and on my blog. And probably other ways, too. He has taught me a lot about his end of things, which helps me as a patient understand more of what is going on.Doc Searls: patient as platform and “point of integration”http://e-patients.net/archives/2008/06/doc-searls-patient-as-platform-and-point-of-integration.htmlFantastic discussion on this short post: http://e-patients.net/archives/2009/05/participatory-medicine-what-is-it-for-you.htmlAnother short post with some good discussion: http://e-patients.net/archives/2009/04/the-i-in-it-stands-for-information.html
These are all such great posts – Everyone seems to be coming together to grow this Flower. Obviously, everyone wants to nurture it and see it blossom.The conversation has mostly moved to Google Wave – Although, there continue to be great posts here, as staticnrg and others demonstrate above. Unfortunately, we’re starting to reach chaos levels – But I still want to encourage participation by everyone here. (Heck, even coming to this page means you’re interested in helping to develop this concept.)At some point, we’re going to have to organize some, if we want to push the idea ahead, but UNTIL then, feel free to come along and join the Flower discussion on Google Wave. Just look me up (dirkstanley@googlewave.com), and I’ll hook you up to the brainstorming wave.(At some point in the future, when we start to see Flower grow more, we can start to organize the crowd by developing teams and such…)
Oh, and Natalie – I thank you so much for the offer, but right now we don’t have enough product… But I would say keep an ear to the ground and keep watching the skies, this may develop into something significant soon. We’ll see. 🙂
Just to memorialize it here, on January 26th at 12 Noon Eastern time, Dirk Stanley, Tim Sturgill and Steve Daviss will discuss the emergence of ‘Flower’ as conversational currency in the quest to stimulate a ‘bottoms-up’ but granular transformation of current HIT platforms, and functionally route both personal and electronic health information digital rivers via quasi ‘public utility’ infrastructure delivering on demand, real time, yet patient centric and rights authorized health information. The program link for the webcast is: http://www.blogtalkradio.com/2healthguru/2010/01/26/doctor-do-you-speak-flower
ahhh… transparency! Looking forward to it… Y'all keep us posted. I'd like to add a Flower definition to our patient and physician FAX page at http://www.personalmedicineinternational.com. Is there a good Patient Definition? and Distinct Physician Definition yet? Best, Natalie