Trust in the Machine

Enigma Machine

We talk about trust a lot these days. Trust is one of those things that the scarcer it becomes, the more you find yourself thinking and talking about it. Trust is also a very fragile entity and it comes in a continuum of shades and magnitudes, from trusting your alarm clock, to implicitly trusting your mom, to trusting in God. Trust can be based on exact understanding (the alarm clock), reinforced by repeated experience (mom), or a result of pure faith (God). Trust that needs to be verified is no trust at all. Trust usually comes into play when one willingly relinquishes control over certain outcomes to a trusted entity. The amount of trust involved is in direct proportion to the importance of expected outcomes, thus “I trust you with my life” is very different than “I trust you to deliver my mail”.

For a while now, I have been following ONC’s efforts to build public trust in health information exchange and electronic health records, summarized in this appeal to patients: “If your health information is sent or used electronically, it's important that you trust the systems that protect it.” Yes, it is important and it is also not much different than trusting the United States Postal Service (USPS) to deliver your mail. You trust that your letter will be delivered in a timely manner to the intended recipient and nobody else. You trust that it will reach its destination in one piece, that nobody will open and read your letter in transit to “provide you with better service” and that the USPS will not make copies of your letters and otherwise use them or sell them to the highest bidder. Tampering with other people’s mail is a Federal offense subject to fines and jail time. Pretty good start, if you ask me.

When we advanced from paper letters to email, we paid a price for the associated convenience and instant delivery. Regular electronic mail has no envelope. Your email service providers reserves the right to read all your emails and use the content any way they see fit. Unless you take special precautions, anyone could intercept your mail and derive some joy from reading it too. Interestingly enough, most people became impervious to the loss of privacy. Now we are contemplating the exchange of health information through similar mechanisms, and we are being told that we should really use envelopes for exchanging health care information.

Our health care providers have been exchanging information about us for quite some time and much of this is done over the Internet now, but instead of using a public postal service, they established private networks and secured those as best they can, thus obviating the need for envelopes. This is very much like the diplomatic pouch system, where the channel itself is secured, but each secret document inside it is not necessarily locked down.  To be fair, the amount of data exchanged between health care providers (prescriptions, lab results, claims, radiology images, etc.) is so massive that it would be rather inefficient and expensive to start putting each message in its own separate envelope. The individual envelope system does make sense for exchanging small pieces of information with patients, and even for some small health care providers when they communicate amongst themselves and with larger ones infrequently.

But this is not just about envelopes. It is also about making sure that our messages go to the intended recipient and that we are certain that the sender is who he said he is. The last part is a bit tricky and the USPS, for example, never purported to verify the sender’s identity, maybe because mail fraud is punishable by up to 30 years imprisonment.  In lieu of similar laws for health information exchange over the Internet, we are being told that technology exists to protect us just as well. These technologies consist of software tools for proper authentication, non-repudiation, integrity, availability, confidentiality and the associated paraphernalia of cryptography, ciphers, encryption, public keys infrastructure, passwords, biometrics, tokens and networks of machines to support this mathematical infrastructure.

There may be value in explaining the technology to people, but even the most technology-challenged folks amongst us know enough to trust the machine, just like we know enough to trust that the alarm clock will go off in the morning, or that the TV will turn on when we push the power button. And we do understand that a certain rate of failure is to be expected. But, and there is always a “but”, we are not the ones pushing the buttons here. All these wondrous technologies are applied by an intermediary. Basically, we are delegating the stuffing and opening of envelopes to someone else and that someone else is not your trusted secretary of 25 years. It is a complete stranger, and if we are to comfortably exchange our secrets over the Internet, we must somehow trust that those intermediary folks are not reading our messages for entertainment in the lunch room, or making copies to read later or to sell to interested parties. It’s not about technology. It’s not about trust in the machine. It’s about trust in the operators, and we know next to nothing about those operators and their interests other than that they are called Health Internet Service Providers (HISP) and could be large clearinghouses like Surescripts, or your own EHR vendor, or a local health information exchange organization, or an independent technology firm, or anybody else selling electronic envelope stuffing and opening services.

There is of course HIPAA, and there are all the new regulations specifying what needs to be encrypted, how and when it should be exchanged, who gets to be the keeper of the keys, and the process by which we choose to participate or not. People have an expectation of privacy when seeing a doctor, although with the advent of health insurance, those expectations have been greatly diminished. We have come to accept that certain data about us is not private, but we are still holding on to the notion that other, very personal, things need not be shared outside the exam room. Doctors don’t usually report to insurers how much alcohol we consume, whether we are sexually active and in what manner, what we eat, which illegal drugs we use, how we sleep and all other intimate fears or dilemmas shared with a doctor. Your doctor is entering all this information in a computer now, giving it a life of its own, and since sooner or later this information will be leaving the doctor’s computer, it may end up in unexpected places, not because the system was breached, but because the “system” sent it there. Will it end up on Google? This is the trust issue that needs to be addressed. Or perhaps it doesn’t.

In a world where most folks are just fine with seeing targeted adds on every browser page based on the contents of their gmail messages, maybe it makes no difference to us if Google “knows” that our last A1c was >9 and a flurry of diabetic adds are unleashed when we browse the Internet. In a world unperturbed by having every smart phone equipped with what amounts to a keylogger, where the Internet Service Provider and the phone manufacturer, along with the keylogger vendor, read every text message you send, perhaps sharing your overactive bladder issues with these folks is also a nonevent. And if that’s the case, why would we even bother with triple DES or AES or Blowfish or Twofish encryption and PKI and certificates? Let’s just cut through the chase, do me and Google a favor and post the stuff to my Facebook page and maybe Tweet a quick clinical summary for my 5000 most trusted friends.

Arguments for a Universal Health Record – Part II

All animals can exchange information when in proximity to each other. Humans advanced this useful exchange to occur when the interacting parties are far apart, which makes the human animal quite unique. First came human couriers carrying verbal information, followed by human couriers carrying written missives, then came technology. Technology in the form of transportation vehicles, and technology in the form of unmanned transport of sounds and symbolic characters, changed the world. Telephones and computers on the Internet rendered the travel time of information from any point on the globe to any other point to milliseconds or less, but did not change the age old paradigm of information physically moving from one place to another. Until now.

This is the age of social media. Those of us who remember licking envelopes and stamps are often tempted to dismiss social media as a superficial waste of time better suited to perpetually distracted kids than any serious endeavor. When you think about Facebook, Twitter, Google+, Farmville and such, it is hard to believe otherwise. Ignoring the actual activities currently occurring on social media platforms, and looking exclusively at the mode of communication, one is forced to acknowledge that a change in paradigm has occurred, and we are reverting to exchanging information when we are in close proximity to each other, only this time around proximity is virtual, not physical. Information ceased to travel virtually, and instead, we do.

When we “go to” Google+ and engage in a lengthy discussion regarding Universal Health Records, we are creating and consuming content which resides in one virtual location – Google’s network of servers. If you want to participate in such conversation, you have to “come to” Google+, just like you had to come to Town Hall in days gone by, if you wanted to debate matters of importance. Unlike exchanging information by horse, train, telegraph or email, this communication paradigm is once again social, but flexible enough to occur in real time or at a time of your own choosing.

Back to medical records. Today most medical records are stored in physical format (paper) at various physical locations (brick and mortar facilities). Health information exchange is occurring mostly through courier, whether manned (patient, snail mail) or unmanned (fax). Those who advocate for electronic medical records desire to change the format of the record from physical to virtual, leaving the storage of virtual records pretty much as it is today. Once the content is computerized, it can also be exchanged by computer couriers, such as email and Electronic Data Interchange (EDI). This is supposed to make medical records “liquid” and the data can then flow from one computer to the other in a network of rivers and rivulets spanning the entire nation. Since such a complex system of waterways can be useful only if 100% clean water is allowed to flow through, as opposed to a mixture of seawater, oils, spirits, and other beverages, much care must be exercised at every medical records repository to transform whatever is released out into the public system to clean water. As discussed in part one of this series, ensuring water purity and building canals, dams and other infrastructure is expensive, fraught with peril, and assuming such system can be built, it is also obsolete right out of the box.

What problem are we attempting to solve by computerizing medical records? The customary answer to this question is that medical care has become extremely complex, it requires scores of professionals working together and, to foster better outcomes, they should all have the most accurate pertinent information at their disposal. Now, if we could bring all these professionals into one room filled with books and journals, and sit them down around one table, we would be just fine with old fashioned verbal information exchange. Since this type of physical proximity is becoming less and less likely, we find ourselves in need of a solution to allow disparate teams to collaborate on one project. We can do this the old way, and arrange for virtual information to flow electronically between team members, or we can do this the social media way, and arrange for team members to meet in one virtual space and work in virtual proximity.  But wait, there is more... In health care, our projects are longitudinal. Each episode of care builds on all previous ones and also informs all episodes to come. This in a nutshell is why the entire medical record must be an open and shared resource.

Given the realities of our health system of systems, I am being told that such selfless collaboration at the data level is very unlikely, and given the real and manufactured concerns with privacy and government oversight, having a universal comprehensive data store is politically impossible in health care. Nobody objected to the technical soundness of the proposed solution. Granted, health care is much more complex than Google+ or Google Docs, and we will need more data, more definition and a much bigger and more sophisticated transactional database structure. As much as I would like to, we cannot flip a switch and begin accumulating universal health records overnight. So how would we go about starting to move in this direction? 

One very promising idea comes from Dr. David Kibbe and the Collaborative Health Consortium. The notion of a health care collaboration platform, or clinical groupware, could do for health care what Google+ and Facebook did for virtual social interaction, but it stops short of providing a longitudinal and open medical record. If you were an avid Facebook user and recently tried to switch to Google+, you probably already encountered the big tall wall surrounding that particular platform. While this may be a minor nuisance when it comes to social media, and fully understandable from a software, or platform, vendor business perspective, it is not so minor when it comes to medical records, as every doctor who tried to switch EMRs can tell you. Every business should have the right to erect walls around its platform, its innovation and its intellectual property. No business should have the right to monopolize patient data, even if it was created by services and tools of a proprietary platform. The data layer must be separated from the service platform layer, because the data layer belongs to individuals and, in aggregate, it is a public good. 

Another suggestion was that initiating standardized information exchange may lead to the eventual creation of local and later regional data stores. Perhaps the various State HIE organizations would grow into such data repositories. Perhaps the ever expanding integrated health systems would accomplish something similar. Eventually, we may be able to connect all these repositories into a federated model of national health records. All this is possible of course, but this rudderless experiment strikes me as a major waste of time and resources. So here is a small suggestion. There are several billions of dollars appropriated for a VA/DoD joint EHR which is supposed to be open source. Presumably, such effort will yield a database schema sooner rather later. Let’s use that. Let’s define a minimum set of data, not much different than what is required to be exchanged for Meaningful Use, and begin populating a national database. It will take time before this becomes the authoritative version, but it will happen. Initially, we can mandate certified EHRs to use the national database to retrieve and update this modest dataset in real time. This should not be a very difficult task for EHR vendors. At the same time, we should allow new products to be developed against this new and open schema. What would be the cost of building a simple user interface to the Universal Health Record to display an accurate list of problems, meds, allergies, immunizations and lab results? Hint: very close to zero. What value would physicians, and patients, derive from the ability to access such definitive lists for any patient, any time, from any browser, on any device? You decide.

Health Information Exchange is an outdated paradigm. It is based on understanding the Internet to be an improved version of the Pony Express system. The Internet has evolved into something completely different and unless we evolve with it, we are doomed to be arming heavily for a war that has concluded and it will never be fought again.

Arguments for a Universal Health Record

We passed the one thousand mark on products certified as EHR technologies for ambulatory care and the five hundred mark for inpatient care, and there is no relief in sight. In addition, there are multiple other software products that are routinely used in health care, such as standalone practice management and billing systems, claim processing software, pharmacy programs, lab, imaging and other diagnostics software, personal health records products, and more recently a veritable explosion in mobile applications ranging from monitoring your heart to evaluating your happiness. I don’t know of any other industry where so many disparate software packages are able to communicate and cooperate with each other seamlessly, and yet this is the goal of the gargantuan effort of those who develop interoperability standards in health care. If you’ve ever been involved in software systems integration, you probably know all too well that the weakest and most unstable link is always at the interface between products, even those built by the same vendor, regardless of the agreed upon standard. When it comes to seamless operations and cost effectiveness, nothing beats true database level integration.

For those who read this and have an irresistible kneejerk reaction tempting them to cite examples such as ATM networks, telephone networks, Google or email, please understand that this is an apples to unicorns comparison. Assuming that our ultimate goal is to have all health records for all people available at all geographic locations at all times, is weaving a web of rickety interfaces between thousands of products, really the best option? It is, if you sell existing, or enabling, technology for this arrangement, and it is not, if you intend to use, or pay for, the end solution.

The usual arguments against a Universal Health Record, and its scary database in the sky, are that we must build on existing infrastructure; that rip-and-replace is cost prohibitive; that a free market should provide as many choices as possible; and that privacy is best served by keeping data close to home, and certainly out of the hands of Big Government. Sounds pretty reasonable. What if we dig a bit below the surface though?

Medical Records

• Assumption: At any given moment in time there can be only one correct version of a complete medical record for any one person
• Fact: Currently, various parts of the medical record are stored at various locations, by various organizations, in various formats
• Fact: Most organizations possess unique content, but also content overlapping with what others store, containing multiple discrepancies and various errors
• Observation: Using partial medical records for provision of care could be desirable, inconsequential, dangerous or lethal, depending on which parts are missing
• Observation: There is conceptually no reliable way to know whether parts of the medical record are missing at the point of care, let alone ascertain the criticality of missing parts

Health Information Exchange (HIE), as its name indicates, is intended to shuffle fragments of the medical record from one organization to another just in time to inform the provision of care. The government and various other organizations are diligently working on standardizing the contents, the format and the means by which medical records data is communicated. Since the thousands of software programs deployed in health care all store data in different formats, using different data dictionaries, different storage systems and different terminology, it is envisioned that each system will have some sort of transformer at its edge that will translate the inner workings of the system before sending information out, and execute the reverse procedure before letting outside information in. Once the standards are finalized, all technology vendors will be building (or buying) such “transformers” and everybody will be communicating seamlessly. Could it really be that simple?

Reconciliation

Unlike banking, where managing a checking account at your local bank does not require immediate information on your Cayman Islands holdings, medical care operates on a single record set of data elements. Since this record set is being altered at various care facilities, health information exchange must continuously reconcile the data elements. So for example, let’s say that you visit your primary care doctor complaining of chest pain and he diagnoses gastrointestinal disease and prescribes antacids, but you are still concerned and decide to see a cardiologist in the city, who diagnoses angina. Shortly after visiting the cardiologist office you get hit by a bus and end up in the local ER. Was your cardiologist aware that you have been complaining of chest pain for the last 20 years, angina was repeatedly ruled out in spite of your concerns that Aunt Mary also has angina, and antacids always worked for you? Is the ER aware that you just got diagnosed with angina and have a shiny NitroMist sample in the backseat of your car? Is your primary care doc going to be appraised of your adventures?  In a world of perfect information exchange the answer is yes to all questions.

However, perfect information exchange in this case requires that your primary care physician pushed your medical records out to the cardiologist, including your fixation with angina and Big Macs, or that the cardiologist was able to locate your primary care records and pull the information in. It also requires that the ER was able to obtain your primary care records from back home, any other medical records from other providers and also the very recent cardiologist records and combine all those data points in one authoritative record set. This reconciliation process would occur every time you seek care and every time you, or other diagnostic facilities and eventually devices, update your records in any fashion. And these transactions will have to execute without a unique patient identifier just for you, and while processing and propagating privacy rules which may differ between various care providers and exchange intermediaries.

Now imagine millions of people with similar needs, and you have many millions of transactions flying around back and forth between thousands of software programs executing in hundreds of thousands of locations, from industrial strength data centers to the lonely Dell server under the printer in a doctor’s office. Yes, the contents will be standardized by those edge transformers, but every relay, every handshake, every acknowledgement and every translation back and forth to the native software program constitutes a point of possible failure, and every reconciliation of multiple messages from disparate sources is an error waiting to happen.  In computer land errors don’t usually wait for too long before they happen, and this has nothing to do with lack of standards. Sending applications lose connectivity intermittently and go into a peculiar state of limbo. Receiving applications often get stuck on one bad message, creating huge processing queues on the other end. Messages mysteriously disappear only to be found in a log file or another patient’s chart. Every new release is always an adventure. This is how things are today, with only a fraction of the envisioned number of transactions in the brave new world of a seamlessly connected health care system.

The Power of One

The alternative to having a flimsy system with a multitude of moving parts is to have one unified database system, with one architecture and one schema definition. This does not necessarily mean one EHR. We could of course have a single EHR built on top of this database system, but for those concerned with innovation, free markets and with the problematic one size fits all approach, by all means, let’s build thousands of EHRs with user interfaces and functionality to fit every individual preference, all accessing the same exact database, containing the same exact records. This Universal Health Record will be, by definition, complete and correct at all times, since all health care applications will be built on top of this database, much like browsers are built on top of the World Wide Web. Switching EHRs should be as simple and straightforward as changing from Firefox to Chrome, not to mention how happy the folks advocating substitutable applications instead of walled gardens would be. Oh, and the sum total of investment in a homogeneous data infrastructure is dwarfed by the various other public and private initiatives, all ultimately funded by tax payers.

The 800 pounds gorilla in the room is of course privacy and to a much lesser extent security. A medical database system of this magnitude would have to be built and administered by the Federal Government. Patients would have to be uniquely identified in the system. Granted such Universal Health Record would accessorize well with a universal health care system, but let’s face it, if you are on Medicare or Medicaid, the government already has your medical records. Private payers have mega databases chockfull of medical records and so do EHR companies and pharmacies. Your data is being constantly de-identified, sold, re-identified and exploited for financial profit. Once the planned information exchange network kicks in, a host of State and private agencies will also begin building their own repositories of medical records. The privacy horse has left the barn, and the best we can do now is regulate the use of what was once private. At a minimum, the Universal Health Records database will ensure that you can see everything everybody else is seeing and have some say in its accuracy and utilization, which is orders of magnitudes better that the alternative.

Commedia dell'Arte

"Accountability is something that is left when responsibility has been subtracted."
-- Pasi Sahlberg


This was the year when America turned on its doctors, and on itself. Not the 300 million citizens who are busy with other existential threats, but the elite 1% that effectively runs America, and the cadres of intellectuals who provide grant funded scientific cover to our leaders no matter how misguided they seem to be.  Health care is a fiscal mess and someone, other than policy makers, must be held accountable. The greedy little doctors who are over treating us to enrich themselves are a good target and so are all of us greedy little people who refuse to go peacefully and expediently into the night. The same strategy is being applied to education, with the pathetic self-serving teachers obsessed with their benefits and the misfit children who ought to be cleaning toilets instead of learning, identified as the culprits for our educational fiasco. Mind you, the elite 1% is not experiencing either education failures for their children, or health care difficulties for their families. For them, this is not personal, it’s business, and they are about to make us an offer we can’t refuse.

A hundred years ago, give or take a couple of decades, America delegated the responsibility for taking care of the sick to the medical profession, and as science advanced by leaps and bounds, people were greatly rewarded with better health and longer life, and doctors were rewarded with prestige and financial prosperity.  Some say too much prosperity, some say too little, but all in all, fewer than 10 cents of each health care dollar go to physicians. Professional responsibility for sick-care does not require one to be a saint and it is not necessarily incompatible with seeking higher remunerations for one’s services. However, something went very wrong along the way. Ever so gradually doctors have lost control of their profession to the rising corporate and public interests in health care who acquired complete jurisdiction over physicians’ reimbursements. Doctors became the servants of two masters, responsible for one and accountable to the other.

This obviously unworkable situation caused enormous problems during Managed Care I (the HMO). On the eve of Managed Care II (the ACO), our leaders are proposing, on behalf of the people, to release the medical profession from the moral and ethical responsibility which formed the foundation of the patient-doctor relationship and replace it with uniformly measurable accountability to public and private payers. Patients are advised to reject the old ways of paternalistic physician managed care, in favor of the empowerment afforded by payer, health system or employer managed care, which is certain to bring about better health care at lower costs everywhere except in Connecticut. Physicians, who enter apprenticeship as teenagers and graduate somewhere in their thirties, are having difficulty letting go of the historic burden of responsibility. Patients seem not to have read the official memo, and most are still expecting doctors to uphold their end of the ancient bargain. There are of course well publicized and well marketed exceptions.

While responsibility is entrusted, accountability must be managed, monitored and acted upon. From a patient’s perspective, the locus of trust must shift from the doctor to monitoring organizations. While the old trust was based on long term relationships, word of mouth or gut feelings, introducing much variability in outcomes, the new trust is based on facts, calculations and objective data, hence the controversial importance of Electronic Health Records (EHR), which are increasingly fitted to facilitate the transition from old to new.  EHRs too are the servants of two masters, used by one and governed by the other.

Early EHRs were built and sold to doctors as tools to enhance practice revenue and personal income. Interestingly enough, very few physicians found that proposition enticing, and EHRs did not sell very well. Today’s EHRs are prescriptive data collection tools, with budding capabilities for reporting and exchanging information, and largely promissory abilities to deliver relevant evidence based protocols at the point of care. As the Meaningful Use incentives program enters its second year, physicians are increasingly purchasing and using EHRs. A minority is truly excited about a digital future, but the majority of EHR users, and practically all those still sitting on the sidelines seem to be asking the same question: how does this help with patient care? Well, it does, and it doesn’t, depending on what one means by patient care.

Most physicians are looking at EHRs as tools to help them do a better job. These doctors are still under the impression that they are at the center of health care delivery and EHRs are tools to assist them discharge their responsibilities to their patients. They are looking to computers to help search a medical record in intelligent ways, abstract all pertinent information and no more, manage repetitive tasks on their behalf, deliver timely reminders, provide advice upon request and become invisible when not needed - in short, the perfect butler. This is about hands-on patient care, one patient at a time.

Those who govern EHRs are continuously harmonizing them, through the Meaningful Use regulatory system, to promote accountability of EHR users. They need data. They need boxes to be clicked, numeric values to be captured and buttons to be pushed, and they need everything compiled and transported out to analytics engines to assess performance or lack thereof. They don’t need to know about Mary’s Lasix trouble, but they do need to calculate the p value from paired t-tests for the average change in percentages between baseline and subsequent years across patients qualifying for the measures. This is about standardized patient care at the population level.

Today’s EHRs have some features serving their users, but most development is geared to serve the governors and as a result, EHRs are not able to please either one of their masters. As Managed Care II blooms and the doctors for the 99% transition to accountability regimens, minding their p-values and t-tests, EHRs will become fabulous engines for enterprise data collection and processing. When the powers to be come to the realization that government intervention based on the assumption that people are irresponsible, greedy, dimwitted and largely inconsequential is doomed to fail, and Managed Care II joins its predecessor in the annals of failed policy, EHRs will finally become slick, intelligent and nonintrusive servants to both responsible doctors and their patients, helping deliver better health care at lower costs, one patient at a time, and by definition across the sum total of the people, because technology is not the limiting factor. Responsibility is.

The F Words of Health Care

Vassily Kandinsky, 1923

Fragmentation, Fee-for-service and Futile care are the trifecta of what is supposedly ailing our health care system, or non-system, as it is fashionably described nowadays. Modern health care has reached its crisis point not due to hordes of people keeling over and dying in the streets, as they did during historical health care crises brought on by plagues and famine, but due to exploding costs of delivering decent care to all people. Since the issue now is mostly financial, health care as a discipline is attracting the interests of those who practice the dismal science of Economics. Over the last two centuries, economists have successfully addressed the F words in other industries with spectacular results in developed countries, so why not apply lessons learned to health care? 

The obvious reason to treat economists with suspicion in health care is the quintessential argument that people are not widgets, but there is another problem. Most tried-and-true solutions for increasing availability and quality while lowering costs of products are not accounting for the other explosion occurring as we speak – the Internet.  How can this assertion be true when we are in the midst of a government sponsored spending spree to computerize medical records and adopt Health Information Technology (HIT)? Apparently, even those who lead and define the HIT revolution are reluctant (or unable) to grasp its full implication, thus they are consistently underestimating the power of the Internet to serve the individual, and as a result are hedging their bets on technology with classic industrial models from days gone by.

In a 2008 Health Affairs article, Dr. Donald Berwick has defined what has become the official goal of policy making for the Secretary of Health and Human Services. Better known as the Triple Aim, the goals are to create better health, provide better care and lower costs of care. If you look at health care as just another industry, the Triple Aim translates into a better product with a better process at a lower cost. Well, when put this way, the solution is pretty obvious and it has been obvious for over two centuries. We must address the F words: eliminate Fragmentation by aggregating independent artisans in one physical location, stop paying Fee-for-service (piecework) and pay salaries instead, and most important, eliminate Futile work by standardizing the process. In short, apply the industrial revolution to health care and realize the economies of scale that brought prosperity and happiness to the developed world. Except that for some strange reason, this solution doesn’t quite work in health care.

Case in point: Federally Qualified Health Centers (FQHC). FQHCs started out in the early 1960s as community run clinics to provide medical care to the poor. By the mid-nineties, and with the best of intentions, the Federal government and the Centers for Medicare & Medicaid Services (CMS), created funding grants and reimbursement methods to support these clinics. Today there are thousands of FQHCs of different types, operating in health care shortage areas and providing team-based comprehensive care including preventative care, basic primary care, behavioral care, dental care, lab and pharmacy services, mostly to Medicaid beneficiaries and the uninsured, but also to small numbers of Medicare and privately insured patients. FQHCs must use mid-levels to provide and coordinate care and must report on quality measures. In return, FQHCs receive millions of dollars in grants for building and improvements, have access to cost effective workforce, can obtain free malpractice protection, are tax exempt and are paid more than double what a private practice is paid for Medicaid services. By all accounts, FQHC are addressing the triple Fs of health care rather well, but how are they doing against the Triple Aim objectives?

Studies are mixed regarding quality of care provided by FQHCs, and patients cared for by FQHC are largely sicker than those seen in private practice. Interestingly enough, neither Medicare, nor privately insured patients are flocking to FQHCs, in spite of the financial advantages offered, particularly to Medicare patients, and in spite of the spiffy state of the art facilities. This may, or may not be, an indicator for perceived quality of care. How about lowering costs? Do FQHCs provide care at a lower cost than, say, an independent solo private practice?  Adding direct reimbursement rates, grants, tax breaks and other benefits, FQHCs visits cost more than twice the amount paid by Medicaid to private practices, which cannot compete with FQHCs and all but disappeared from areas where FQHCs operate. What would have been the results if twenty years ago CMS would have decided to increase Medicaid fees and pay for uninsured visits to independent practices, instead of exclusively backing the creation and operations of a separate but equal clinic system for the poor? We may never know for sure.

FQHCs are only a small example* of why economies of scale are not easily achieved in health care. Large hospital organizations and even fully integrated health systems, which may be providing better care (or not) seem equally incapable of reducing costs in spite of attacking all three Fs, or seeming to do so, and there are two reasons for this failure: a) larger health care facilities have disproportionately larger overhead costs and b) large systems are better equipped to charge more for services, which renders their efficiency efforts less urgent. And this is not a matter of opinion. CMS acknowledges this built-in inefficiency as evident in the physician fee schedule which pays an additional “facility fee” for services provided in hospital owned outpatient clinics, presumably to cover the extra overhead. Surprisingly, CMS is consistently creating incentives and regulations to accelerate provider consolidation into these big inefficient and expensive systems. The only possible explanation would be that CMS is betting that elimination of the last two Fs (Fee-for-service and Futile care) will be easier in a consolidated environment and the gains will ultimately exceed the losses from doing away with independent practice (Fragmentation). What about information technology? Well, it is supposed to help with process standardization, data collection and performance measurements, similar to what computers do in every other industry.

We have all seen the infomercials for high-tech hospitals, where a bunch of doctors are seated around a conference room table, each holding a laptop or tablet, presumably discussing patients in a team environment. There is something very wrong with these pictures. First, it costs us a fortune to have all these physicians in one room. Second, there is almost no added utility for them to be using computers instead of passing around a piece of paper, and computers are expensive. Third, there is no patient in the room. Now let’s imagine a different picture: a primary care physician sitting in his office, with a patient next to him, both interacting with a computer on which a Skype conference is taking place with an oncologist sitting in his own office thirty miles away, a surgeon in a hospital lounge in the city and perhaps a radiologist half a continent away. Everybody on the call has access to the same electronic medical record, appointments can be made in real time, literature can be consulted and shared, prescriptions can be changed and a common care plan agreed upon by all and understood by all can be created and by using intelligent predictive analytics tools various options can be explored. Perhaps a family member in a different country is conferenced in and perhaps the patient is at home or in a break room at work. Perhaps there’s an electronic sign-up sheet for the oncologist, if the patient wants to ask something else later and have a physician friend in New Zealand listen in. And with one click on a PayPal button all doctors are paid for their time.

In this Internet age, manufacturing style physical consolidation is not only unnecessary, it is cost prohibitive. Modern lifestyles and modern medicine have created a need for doctors and patients to collaborate and the Internet is providing the means to accomplish such collaboration without having to physically gather everybody under one expensive roof. There is no need to obliterate the operational efficiencies of private practice and replace it with the bloated bureaucracy of large institutions, and there is no need to dispense with long lasting doctor-patient relationships in favor of computerized care coordination, and there is absolutely no need to substitute a bunch of numbers in a computer for a real patient. The Internet is decentralizing and individualizing everything from politics to manufacturing. Health care is, and always has been, decentralized, individualized and based on the local patient-doctor dyad. The resemblance is striking. We either embrace the fully aligned collaborative nature of the Internet to achieve better health, better care at lower costs, or engage in a doomed effort to impose an unnatural centralized command and control structure in health care just because it worked well for nineteen century steel manufacturing and because policy makers don’t truly understand the magnitude of the connectivity revolution.

* According to the Kaiser Family Foundation FQHCs had about $12.7 Billion in revenues in 2010, 75% of which came from Federal and State agencies. They served almost 19.5 million patients with over 77 million encounters. Simple math yields a cost of approximately $165 per encounter.

The Pin Factory EHR

In 1776 Adam Smith explained to posterity how specialization increases productivity using the now famous example of a pin factory. While one master pin maker could turn out anywhere between 1 and 20 pins each day, going through all the steps involved in making pins all by himself, a specialized army of laborers, each fulfilling one step in the pin making process, could increase productivity more than two hundred fold and turn out almost 5000 pins per person per day. This would have the triple benefit of enriching the factory owner, creating jobs and making pins both affordable and widely available for consumers. What happened to the master pin maker, who used to make a very nice living when pins were expensive and hard to come by? He would most likely be employed in the factory to supervise the smooth flow and quality of the new pin mass production system. He would make sure that each laborer works at a speed appropriate for feeding the next laborer in line and he would probably sample a few pins here and there to make sure they are as sharp and sturdy as the ones he used to make in the olden days. When the master pin maker passes away a new supervisor would be hired, most likely one that has never made an entire pin before, but instead has a much better understanding of the production process. The profession of pin coordinator has been born.

Although Adam Smith put forward the notion of specialized labor, Henry Ford is customarily credited with the invention of the modern assembly line. Interestingly, Ford is attributing his invention to the observation of Chicago’s meat packing industry. It seems that while no two cows are identical, the butchering of animal life lends itself rather well to disassembly line methodology. Today, manufacturing assembly lines use human labor where it is cheap and in abundant supply, and are staffed with robotic machinery where human labor is expensive and/or scarce. In all cases the process is orchestrated and controlled by sophisticated computer software. This is why we are all able to purchase a car, chat on our cell phones and enjoy perpetually fresh slices of white bread in plastic bags, amongst many other wonderful things, which were once only available to the wealthy few.

Modern medical care is increasingly out of reach of most people. It is expensive, and adequate resources are scarce in many areas. Medical care also varies widely in quality, and the costs of production are anybody’s best guess, depending on geography, time of year and even workers vacation and education schedules. This is very much the same as making pins in the eighteen century. In all fairness, some specialization of labor has already occurred in medicine, but there is no coherent method of placing each worker in his/her station of the continuum of care, and there is no standard process by which workers hand off work from station to station. According to experts, this lack of orderly processing, along with the absence of quality control, is creating a terrible waste of resources and a flurry of defects in the finished products. If the advanced methodologies of modern day manufacturing are working so well for everything from cars to pins to cows, wouldn’t it make sense that we should at least try them in medicine?

Fortunately, we already have several pieces of the puzzle in the works. As mentioned above, we do have a certain degree of specialization in medical practice. We also have hospitals, which could function very much like factories, but as Clayton Christensen observes, most have no well-defined assembly lines. And then, of course, we still have the independent small shops that take piece-work home and operate without any standardized quality control. We also have the beginnings of computerized control systems in the form of Electronic Health Records (EHRs), which, according to John Halamka, are quickly moving from just bookkeeping software to dynamic coordination of processes, complete with encyclopedic knowledge of medicine and a good measure of artificial intelligence to devise and “enforce automated care plans”. 

The only thing left to do is to lay out proper assembly lines, and we don’t really need to think outside the box too much, because manufacturing has solutions for this dilemma as well. In modern industry, there are practically no factories that start out with raw materials and end up with a finished product. Instead, some factories concentrate on producing parts and others are built to receive parts and assemble them into useful products. Exact specifications for each part, to be followed by production lines and relied upon by assembly lines, make this geographically dispersed process possible. In health care, the primary care homes will serve as production centers, where people are constantly measured, tracked, tested and evaluated, so when they are finally shipped to a hospital for a procedure, the hospital knows immediately which assembly line to place them on and the omniscient EHR will control the most minute detail in the process, from medication dosing to incision size and implantable device brand and model, thus reducing both errors and costs. Once the hospital’s work is done, patients are released back to evaluation and management in production centers, and here is where the cyclical nature of health care differs from a typical manufacturing process, and this is why it is extremely important that EHRs be interconnected and preferably Cloud based to achieve a high degree of omnipresence.

Yes, there are many more details to be worked out, like emergencies, accidents and the exact specifications that an EHR should contain on each type of person. We will have to establish quality feedback loops between hospitals and primary care centers to continuously refine processes for both entity types, so basically the EHR will need to be able to adapt to, and learn from, new information, in a manner similar to IBM’s Watson software. Since people are not pins or even cars, the tolerance levels (allowed deviation from specs) will be high initially, so line workers will need to be highly skilled as well. In all likelihood physicians will be working those lines for the foreseeable future. As the learning control system improves, portions of work would be offloaded to less skilled resources and eventually to machines, and more significantly, entire tasks could be packaged into deterministic protocols and pushed out from expensive hospitals to the less skilled primary care production centers, which will further push the most trivial tasks to consumer owned devices.

Obviously, EHRs will prove to be the heart, brain and circulatory system, of the health care industry. As we speak, EHRs are increasingly being tasked with care coordination activities (not to be confused with continuity of care, or longitudinal care), which are the precursor to the industrial line controller. Folks wondering why they should use EHRs that are not ready for prime time, should understand that we have to have an EHR in every practice, so that the system can have visibility into current processes to learn, adapt, grow and devise new methods of providing care. After all, you cannot control that which you cannot see. 

If you think this is all farfetched and disastrous, please find a senior citizen that lived through the Great Depression and ask her what she thinks about dinner being prepared moths in advance in computer controlled industrial vats, thousands of miles away from home, pumped full of preserving chemicals, freeze dried, shrink wrapped and delivered by airplane to a football field size department store, with minimal human intervention, ending up in a small irradiation chamber in your home before it hits your dining table (or couch). Yet we all buy the stuff and feed it to our kids with no apologies, because it is cheaper, faster and more convenient than tenderly preparing beef stroganoff and baking pot pie at home, after work, every day. And neither grandma nor you can even fathom the handcrafting of pins by master artisans. Is health care really that much different?

2011 EHR Adoption Rates

On Wednesday, November 30, the Centers for Disease Control and Prevention (CDC) released the results of its yearly survey on Electronic Health Records (EHR) adoption for office-based physicians. No surprises. Generally speaking, the majority of physicians in ambulatory practice are now using an EHR, and over half of surveyed doctors say that they intend to seek Meaningful Use incentives. The report is also presenting results broken down by State, so you can learn what folks are doing in your immediate vicinity. The more instructive exercise is to compare last year’s survey results [Fig. 1] to this year’s estimated EHR adoption numbers [Fig. 2].

Figure 1: Percentage of office-based based physicians with EHR - 2010

Figure 2: Percentage of office-based physicians with EHR - 2011

The most immediate observation is that 6.2% of physicians have adopted an EHR in 2011, thus returning to EHR growth rates preceding the 2009 -2010 slowdown, which was largely due to the confusion created by Meaningful Use regulations. The next observation is that the percentage of docs that have at least a basic EHR has gone up by 8.9% in 2011. A basic EHR is one that has “patient history and demographics, patient problem list, physician clinical notes, comprehensive list of patient's medications and allergies, computerized orders for prescriptions, and ability to view laboratory and imaging results electronically”. Although the survey instrument in 2011 did ask about more advanced functionality, and is practically identical to the 2010 instrument, the CDC did not publish a separate number for those with fully functional systems in 2011. Although I cannot be certain, I would assume that most of the growth in 2011 was fueled by Certified EHRs, which by definition should be fully functional. So if I had to guess, and I hope CDC will release the numbers so I don’t have to, I would estimate that in 2011 we have at least 20% of physicians using fully functional systems, which is roughly double what we had in 2010.

Another interesting trend that has been holding since around 2007 is that about a quarter of office-based doctors have some type of bare bones software in their office and they are not upgrading to even a basic EHR. Considering that over half of those surveyed intend to apply for Meaningful Use incentives, this trend is bound to change in 2012.  Some of these folks may have purchased a fully featured EHR, but chose to either not turn features on or chose not to keep up with upgrades to newer versions. For ambulatory EHR vendors these numbers translate into a market opportunity ranging from 50% of the market to a full 80% of ambulatory physicians.

It would be very beneficial if CDC released the complete data set from this survey (anonymised, of course), so we could gain a better understanding of EHR adoption patterns by practice type, size and location. Although it is widely acknowledged that larger practices and employed physicians are further along the curve, the rich details provided by the survey instrument should help both vendors and various organizations engaged in efforts to spur technology adoption, better target their work, and it could also illuminate any disparities which may affect quality of care for vulnerable populations and physicians who serve them.

In summary, the new CDC survey is showing a stable growth in technology use by office-based physicians, modestly improved by government initiatives over the last two years, and well positioned to further improve in 2012 and beyond.