Saturday, September 22, 2012

Shocking News – EHRs Work as Designed

The health care crowd is abuzz with The New York Times revelation that Medicare billing rates seem to have increased by billions of dollars in parallel with increased adoption of EHR technologies for both hospitals and ambulatory services. The culprit for this unexpected increase is the measly E&M code. Evaluation and Management (E&M) is the portion of a medical visit where the doctor listens to your description of the problem, takes a history of previous medical issues, inquires about relatives that suffered from various ailments, asks about social habits and circumstances, lets you describe your symptoms as they affect your various body parts, examines your persona and proceeds with diagnosing and treating the condition that brought you to his/her office or hospital. The more thorough this evaluation and management activity was, and the more complicated your problem is, and the more diagnostic tests are reviewed, and the more counseling the doctor gives you, the more money Medicare and all other insurers will pay your doctor. Makes perfect sense, doesn’t it?

In 1995 and again in 1997 Medicare has specified exactly how to measure a doctor’s thoroughness by creating 5 levels of visits and defining each level’s complexity in terms of an exact number of questions a doctor asks, and an exact number of organs and body parts that are addressed during a visit. The more sanctioned questions and body parts are addressed, the more money the doctor gets from the payer. During the olden paper days, no physician in his right mind would go to the trouble of actually writing down all these largely irrelevant things, and since Medicare always threatened to audit physician billings, most doctors practiced “defensive billing” and consistently charged less than they should have, because the hand written documentation was rarely indicative of the actual level of service. Enter Electronic Medical Records.

Since before the HITECH act and before the Meaningful Use epidemic, EHR vendors promised doctors an automated way of documenting a visit, so they can spend more time with the patient and not have to constantly write things down. Instead, a click on a couple of boxes would do that for them. Furthermore, physicians won’t have to waste money on expert coders to go through their scribbled notes and figure out a visit level. The software will automatically calculate the appropriate E&M code, based on boxes clicked. Structured data can be very useful for calculations. To make the entire process most efficient, three methods of documentation have been developed to replace hand writing and to efficiently minimize the need for extensive box-clicking.
  • Documentation by Exception – Every EHR has this “feature” allowing the documenter to click on ONE box usually at the top of the page which generates a professional sounding clinical sentence for each organ or body part stating that everything is perfectly normal, or that all your histories are unremarkable in any way. This is a great efficiency to be applied presumably after the interviewer ascertained that all is well with your past and present relatives and body parts. If something is wrong with one or two organs, the clinician can click the Normal button and then edit the exceptional few organs that are affected today, thus obtaining documentation for a complete review or examination of all your systems. Remember that every organ and family member documented is worth a few more dollars according to Medicare’s fee-for-documentation model of reimbursement. No wonder then that this is now a basic feature in every EHR.
  • Pre-filled Templates – These go by different names, but they are a huge time saver for simple and common problems and here is how they work: Let’s say you see a patient with an URI and it is flu season. You document the visit de novo starting from a blank URI template, use all the previously described efficiencies and generate a lovely visit note for this patient. It then dawns on you that you are likely to see hundreds of similar patients in the months to come, and that you always go about these things the same way asking the same questions and getting the same answers. You can save this visit note as a pre-filled template sans patient demographics and histories (really just the HPI, ROS, Exam and for the brave also Assessment and Plan) and when the next URI patient shows up, you can load this pre-filled template and edit exceptions, if any. Since technology is magical, EHRs will also load the patient specific histories and merge them into your brand new note automatically. Two or three clicks will get you enough documentation to allow your EHR to calculate a very nice E&M code and generate enough documentation to keep the payers at bay.
  • Bring Forward – This is really sweet for complex patients with chronic disease that come to see you every few months or so. We all know that not much changes in a few months and most likely everything you will be documenting today is exactly what you documented six months ago. Instead of starting from scratch every time, EHRs have created great efficiency by making it possible for the documenter to bring forward, or load, the previous visit note and allow him/her to edit and make changes based on today’s visit. This beats the old “copy & paste” by a mile, and with a click of a button you have all the organs and relatives and complexity of decision making documented in minute detail. You can now make a few changes here and there as necessary, and the EHR will calculate the appropriate E&M code.
There are other features in most EHRs that are designed to improve reimbursement, but these are the most popular. There are also administrative functions embedded in larger EHRs that allow those who employ physicians to ensure that the docs click on all the necessary things to ensure optimal billing and payment. It is very easy to be critical of clinicians in these scenarios, but let’s remember that if Medicare wouldn’t have defined the value of a doctor visit to be proportional to the amount of text generated during the visit, none of this would have happened.

So the “unintended consequences” of pushing physicians to use EHRs seem to consist of doctors actually using EHRs, as effectively as possible, to document all the little details Medicare wants to see. This can only surprise people who had no clue what EHRs are, how they work, and how they are used in everyday practice, which did not (does not) prevent said people from proclaiming themselves as health care experts, best suited to set the national agenda for EHR design and adoption.

Bonus Tip: Now that everybody has been properly shocked by the E&M coding efficiency introduced by EHRs, I would suggest examining the efficiencies introduced by the variety of “smart” order-sets.

Monday, September 17, 2012

Verbal Wizardry

In his dissenting opinion on the health care law, my least favorite Justice, Antonin Scalia, argued that Chief Justice Robert’s opinion stating that the “individual mandate” is simultaneously a tax and not a tax “carries verbal wizardry too far, deep into the forbidden land of the sophists”. Perhaps this is unusual for the legal system in general and the Supreme Court in particular, but in everyday health care conversations verbal wizardry is now the preferred method of communications. However, health care is much more complicated than the law (with deepest apologies to my attorney friends and family), and health care lacks a supreme authoritative source of truth, thus our verbal wizardry cannot be carried out by proclamation alone. Persistence, as they say, is the most important requirement for success, so in health care we are resorting to the tried and true method of repeatedly employing our verbal wizardry in conversation and in writing until it is wizardry no more.

But verbal wizardry is not a random act of confusion. In the Supreme Court opinion, it was most likely a deliberate construct to avoid calamity. In everyday health care discourse, the sustained verbal wizardry seems to serve as a tool for resetting our expectations from a profit driven health care system run amok. When it comes to medical care, most folks have very simple expectations. We want to stay as far away as possible from medical establishments when we feel fine, and we want someone to fix what’s broken when we are sick. It seems that this humble desire is untenable in our current system. It seems that we must frequent health care facilities when we are healthy, but should temper our expectations regarding medical care when afflicted with disease.

And it seems that this is all for our own good, because how would we know that we are fat, out of shape and depressed if the clinical team on duty didn’t educate us about these things. Armed with electronic prescriptions for generic statins and anti-depressants, an armful of personalized education materials and our damn data, we should finally understand what our 48 size jeans and our skinny bank accounts have been trying to tell us all along. Properly medicated and educated we should now take charge of our health. And if we fail somehow and get sick anyway, we should make allowances for this cost-effective and well-meaning new health care system and stop insisting that the doctor fix everything that’s wrong, because health care is about being healthy not about sick care.

Overriding grandma’s advice to eat an apple a day, in favor of regular visits to a job-creator sponsored clinic to assess your current and future productivity, and maybe even check your teeth, is easier said than done. And it is practically impossible to convince folks that demanding sick care beyond certain limitations is unfair, ungrateful and downright immoral. Here is where well-coordinated verbal wizardry comes in handy. There are many linguistic acrobatics in what seems like a perpetual marketing campaign from government officials, semi-officials and all sorts of stakeholders deriving direct and indirect incomes from health services, but the following three terms seem to be at the cold hearted center of it all.

Patient Centered

Everybody seems hell bent to put the patient at the center of something they sell, but the term “patient-centered”, coined some thirty years ago (thank you for the reference, Rick) was really limited to describing a preferred style for the doctor-patient interview, where doctors actively solicit patient opinions and input. The study showed that patients were more satisfied and were more likely to take their meds after a patient-centered interview with their family doctor. Twenty-five years later, Dr. Berwick imported the term to the US, giving it a more expansive definition to include “choice in all matters, without exception” for each patient. Since then, the term patient-centered has been prepended to whatever is on the agenda du jour. We have patient-centered legislation from the left, patient-centered market-based health reform from the right and our path to prosperity seems to depend on patient-centered reform. You want your bill to look good? Name it ‘‘Patient Centered Healthcare Savings Act of 2011’’. You want to be elected to public office? Start by promising that “I won’t stop working on patient-centered solutions to our health care crisis”. Seeking funding from the brand new Patient Centered Outcomes Research Institute? All you need to do is “adapt recommender systems frameworks that are widely used by innovative businesses outside of healthcare” to “an adaptive computer system that will assess a patient's individual perspective, understand the patient's preferences for health messages, and provide personalized, persuasive health communication relevant to the individual patient” [emphasis added]. Basically as long as you can “identify patient-centered factors” and come up with something that “measures patient-centered constructs”, the funding is yours. To round things up, we now have “patient-centered health insurance”, “patient-centered clinical trials” and everything from health care consumerism to health IT is proudly “patient-centered”.
Like a colorful hefty piƱata, the patient finds itself now surrounded by well-wishing guests speaking softly and carrying big sticks. The only end in sight is when the last dime finally drops to the floor and there is nothing left to extract from the patient hanging at the center.

Fee for Service

Unless it’s something your mom does for you, the expectation is and always has been that a fee should be paid for work other people do for you. Sometimes you buy services in bulk, such as subscriptions or yearly contracts, but in all cases the fee is calculated, and discounted for volume, based on a preset number of services. Another way to obtain services is to purchase warranties and insurance. In this model there is no preset number of services owed to you and you are paying for the seller to assume your risk, and as you probably know, those contracts are full of caveats and small print, because nobody wants to provide you services worth more than your payment.

Historically doctors were paid a fee (or a chicken) each time they did something for you. They were paid directly by you and later you could buy insurance to pay you back a portion of medical expenses. The latter model is still very much alive in European countries with strict price controls and better and cheaper health care systems. In the US, we decided to leave the fees to insurers and doctors to figure out on their own. For a while, when most hospitals were run by selfless nuns and most insurers were mission driven non-profit organizations, this worked well and we were relieved of all that tedious paperwork. As medical care became more advanced, the increasingly large flow of money between all parties attracted the interest of corporate America. Doctors became “providers” and patients became “consumers”, prices started climbing towards what the market can bear and insurers began looking for ways to minimize what they have to pay out in medical loss. Since there is a limitation on how much risk can be returned to the consumer before insurance becomes meaningless, the next best thing was to discharge some risk to the new “providers”. For that to happen, the small doctor shops that are not financially capable of taking risks, had to go away and the remaining “providers” needed to start selling insurers warranties on their “populations”. Where the doctor was once mostly aligned with patient interests, the “provider” is now mostly aligned with the insurance plan with which it shares the risk. I wonder at what point “providers” will begin referring to their operating costs as “medical loss”.

To achieve this wonderful outcome, insurers (public and private) needed to convince the populations involved that fee-for-service is the source of all evil and that their trusted doctors are not to be trusted anymore. Thus we must stop paying for volume and instead pay for outcomes. It makes perfect sense because all of us know that paying the hair stylist on a fee-for-service basis, for example, leads to an explosion in the number of haircuts and we only pay the salon if most patrons like their haircuts. The fact that fee-for-service is the most common model of health care payment in developed countries, and the fact that small practice is both more efficient and much preferred by patients, and also widely used in countries with better and more cost-effective systems, is largely irrelevant to the business model of America’s health care industry, which is now salivating at the prospect of shared “savings”, i.e. splitting what they manage to accountably squeeze from patients through value-based and patient-centered cuts to quality and utilization.

Finite Resources

The final piece in this shell game is to impress upon common folks that in spite of the best patient-centered warranties for our health, there simply isn’t enough money to honor these contracts, and warranties will have to be severely limited. To that end, every self-respecting health economist is somberly announcing that our “resources” are “finite” and sooner or later we will have to responsibly ration medical care (for the poor, of course). Unfortunately, the term "finite" is a mathematically defined term. Only the universe, time itself and God’s wisdom are infinite. All resources on this flying rock are finite. The rock itself is finite and so is the life sustaining energy from its sun. The number of grains of sand on a beach is also finite. The number of cucumbers in your fridge is finite, but that doesn’t mean you don’t have enough cucumbers for a large salad, unless of course your potbelly pig pet is helping himself to a heap of fresh veggies as we speak.

Providing for all medical needs of a finite number of people, with a finite number of organs, who live a finite number of years does not require infinite resources. It does require adequate resources. It may be that almost $3 trillion in “resources” is not adequate to meet the needs of all US residents, although it is a bit unlikely considering that this is double of what other rich and pampered countries spend on their medical care. The explanation to this puzzle is that only a portion of those seemingly lavish resources is going to actual medical care in the US. The rest is going elsewhere, and when our potbelly buddies are done taking their unearned share off the top, we are left with insufficient amounts of cucumbers to make a decent health care salad. 

The Verbal Wizardry Dictionary and Thesaurus – Health Care Edition will be available soon at

Tuesday, September 4, 2012

The Bionic Medicine of Programmable People

A couple of weeks ago I wrote a post about the clueless, but endearing, enthusiasm of technology people as applied to solving the health care problem. A few days ago Dr. Davis Liu published a post on The Health Care Blog describing the vision of Vinod Khosla, the famed venture capital maven, of replacing doctors with machines. It turns out that Mr. Khosla wrote a series of three articles at the beginning of the year in a technology publication describing how his pioneering vision will replace people in industries where either he or his wife are investing capital. Venture capitalists (VCs), although I’m sure they wouldn’t agree with this assessment, are a combination of professional gamblers and loan sharks. The secret to success is pure luck and ruthlessness, and when the combination works and the ball lands on the exact number on the spinning roulette, venture capitalists make lots of money. This is very different than running a business ala Warren Buffet or even Mitt Romney, let alone inventing a business like Apple or Microsoft. In return for risking funds, venture capital gets its juiciest pound of flesh when the funded business sells itself to the public, hopefully for more than it is really worth, and hopefully for a lot more than the venture capitalist risked. For that to happen, you have to create demand for whatever your fund is investing in at the moment. This is why you find VCs shedding tears at the mere thought of global warming, or telling us that the future is all about “I Robot”, or miraculous genetic “I am Legend” drugs , or “gamification” or whatever happens to make up their current investment portfolio. The problem with letting venture capital dictate humanity’s agenda is that the globe is getting warmer; people are getting poorer, sicker and dumber while a few VCs are getting richer.

So what is Mr. Khosla selling us now? It seems that his machines, outfitted with “bionic” software are set to replace the 80% of “middling” doctors and also 80% of equally “middling” teachers. Only physicians like Dr. House will remain standing (for a short time) so they can be “leveraged” to create even more “bionic” software, and may I suggest that Albus Dumbledore could be used to illustrate the surviving human teachers of the bionic era. If you are a little bit familiar with the startup world, then you probably know that a business based on services provided by people is not an appealing investment gamble, because it doesn’t scale well, i.e. revenues and EBIDTA cannot go simultaneously through the roof at the same incendiary rate, because people need to be paid for labor. The trick is to find a business model where no labor is required or to find laborers who don’t require payment. A couple of centuries ago we “imported” such laborers from Africa. Today we are “exporting” labor to where those laborers naturally reside. For tomorrow, we are proposing to make machines that work for free. And this in a nutshell is Mr. Khosla’s vision.

Let’s pause for a moment and address the technorati among us. This is not about medical technology for Dr. Leonard McCoy and Dr. Beverly Crusher or about the holographic doctor in Star Trek Voyager, all very awesome and super cool. This is not about stardate 43632.8 or a galaxy far far away. This is about a time frame of “five or ten years” for “bionic” software and “a decade or two” for an army of “Dr. Algorithm” practicing independently. It should be obvious even to the most ardent believers that, in such a short period of time, neither the science of medicine nor technology will be anywhere near the creation of the medical droid that administered to Padme in childbirth. However, these time frames are short enough to appear on corporate financial projections and startup pro-forma budgets. This is about real money and about some flimsy machinery deployed to play doctor to the poor (China and India are mentioned by Mr. Khosla, but the Mississippi Delta can’t be far behind), and if folks are harmed in the process, oh well…. (See below).

Now, how about this “bionic” software that Mr. Khosla envisions to be a temporary bridge between human cognitive endeavor and full machine control in “a decade or two”? Turns out that the term was introduced by a serial entrepreneur in the social media/marketing, hotter than molten lava, sector.  Later adopted by the O'Reilly AlphaTech Ventures folks who invest in clicks and links and data as well, “bionic” software has something to do with “programmable people” and “it has the potential to unlock a massive amount of unrealized human potential”, presumably as programmed and networked by VC funded programmers. According to Dennis K. Berman, a WSJ journalist writing about technology and “corporate scandals”, also cited by Mr. Khosla, we should accept the “rise of the machines” just like we accept earthquakes and hurricanes because it is simply inevitable. And if we still have any doubts regarding the superior intelligence of Jeopardy winning machines, we are reminded that “systems are now of such scale that they can analyze the value of tens of thousands of mortgage-backed securities by picking apart the ongoing, dynamic creditworthiness of tens of millions of individual homeowners. Just such a system has already been built for Wall Street traders”. I am so totally convinced now, but if you need more real life examples, you should read about the $108 million in venture funds going to “bionic” software helping pharmaceutical companies track social media activity to identify doctors more likely to influence their peers, or the $84 million venture investment in “bionic” software to spy on customers and make them buy more fatty food stuff.

According to Mr. Khosla, soon we will advance beyond simplistic “bionic assistance” to “lazy” doctors, and we will no longer be “free to be stupid or political” and “reject” the “cost optimization” served by “Doctor Algorithm” in its medical practice and none of us will need those multitudes of “average” doctors stuck in the “18th century tradition of “first do no harm””. Since VCs are only interested in engineering, as opposed to social engineering, their off the cuff suggestions for the “middling” 80% in any profession or occupation, where wages are proposed to be eliminated from the expenses column, is usually something vapid like “empathy, advice and caring”. Of course “empathy, advice and caring” doesn’t pay anything like actual doctoring, and it is precisely those large wages that need to be eliminated. Highly educated professional workers are also much harder to transform into “programmable people”, like say, the young girls working in one of those infernal laptop assembly lines in China. So once the new troves of captive cheaper-than-machine labor pools have been secured, all that is left in the quest for zero COGS, within the short cash horizons of venture capital, is to eliminate those expensive and volatile knowledge workers from the balance sheet. And since Mr. Khosla’s educated estimate is that “medical diagnosis or 90% of it is an easier task than Jeopardy”, and since he urges us “not to extrapolate the past and what has or has not worked”, it should be easy as pie to imagine a future almost completely free of professional physicians, classroom teachers and non-programmable people in general.

Thus, in Mr. Khosla’s hospital of the future, medicine will be practiced by thinking machines, while floor scrubbing, sheets changing and bed-pan emptying will be left to humans. Eventually, the medical machines, or the few human lords remaining, will probably take on the last yard of efficiency and create menial machines for janitorial purposes, at which point the only human beings in a hospital will be the patients in the beds, and venture capital’s ROI will be inching towards infinity. Unfortunately for Mr. Khosla, or future generations thereof, the transactional volume will be trending to absolute zero, since the “middling” 80%, after advancing to the mindless 80%, are now mostly extinct or have reverted to gathering wild berries where vegetation still remains. I hope being the Supreme Ruler of nothing at all proves to be a very satisfying experience for Khosla Ventures.