And now, for something much more to the point:
And now, for something much more to the point:
Implementation of the Electronic Medical Record in a hospital or office setting carries inherent risk. Like all medical technology, there is danger as well as benefit to the patient. Unfortunately, this aspect of EMR implementation has been woefully neglected. Our approach has been to consider the EMR no more risky than the doctor’s pen it replaces. This could not be farther from the truth. I refer the interested reader to the following blog post as an introduction to this fact:
Next time, we’ll explore this aspect of EMR implementation further.
It is not difficult to envision a better EMR than already exists on the market, because they all suck. And they all suck for the exact same reason: They were designed by software developers, not health care professionals. They should all carry a label–Caution: This product was never intended for actual use by a doctor or nurse in the care of a live patient. This fact is obvious to anyone who is forced to use one of these programs while sitting in front of a patient, as opposed to running a demo of the product in front of a group of naive physicians. Gee, we all think, it looked like it worked when the guy was clicking through all those “uniquely designed, personally templated, specialty specific and easily customized” screens. Which, in case you still haven’t figured it out, weren’t really running the program but were just carefully choreographed prescripted, preloaded, dummy demo’s. You figured that out after you bought it, right?
Step One and Only in designing a real EMR: Make the EMR fit the health care professional, not the other way around. Doctors do not sit in front of fixed monitors with a mouse and keyboard. They sit in front of patients. We reach out and touch the patients occasionally (well, some of us still do, anyway). We make eye contact with the patient and family members, not with a computer screen while we ask questions of the patient over our shoulder. Doctors and nurses do not click radio buttons. We don’t do “drop-downs.” I decide what fields are mandatory, not some hospital IT guy who gets told by the CFO that the hospital reimbursement will go up if we can make the doctors document the reason the patient is in the hospital on every note, every day, every time he tries to type anything into the medical record. The medical record is to serve the health care professional’s goal to care for the patient. Let’s repeat that: The medical record is to serve the health care professional’s goal to care for the patient. It is not, therefore, to be designed to maximize hospital reimbursement, regulatory compliance foibles, or translate efficiently into some midlevel administrator’s Excel spreadsheet program to make his quarterly report easier to format. Because you are screwing up the patient’s quality of care just so you can make it easier to run a report on how screwed up the quality of care has now become. Are you with me here?
The following has been proven by multiple, reproducible, double-blind, multicenter clinical trials published in the best peer-reviewed journals: There has never been a system more efficient in accurately documenting patient care than a doctor or nurse dictating their findings and plans into a hand held recording device that they carry with them, the results of which get accurately transcribed, reviewed, and signed. This is the way we document our office visits with patients. This is the way we document discharge and transfer notes. This is the way we document our operative care. It is not helpful, accurate, or appropriate to replace this with a single check-off box that describes my operation as “Repair hernia, inguinal.” But now we can do even better than a simple dictaphone and transcriptionist. Every nurse, doctor, and health care professional in practice for more than a month falls into certain documentary habits, certain turns of phrase, that he uses over and over again. These can be templated, triggered by certain verbal or tactile cues, then easily augmented and modified on the fly to suit the individual patient experience. Reviewed, proofed, and corrected in real time, then electronically signed, sealed, and delivered. Poof! The technology exists. It has existed for twenty years. It’s just that nobody cares enough about the way we actually do things to design this form of input into the multimillion dollar boondoggle that is your hospital EMR. They need to implement this form of input. Mouse clicks and managed care don’t mix. Are you a seventy-year old primary care physician who has always handwritten his notes in illuminated manuscript grade calligraphy with a gold-nibbed fountain pen? Fine, use your nib on a handwriting recognition tablet running Evernote. We’re not trying to put a man on Mars, here. This isn’t a technology problem. It’s an attitutude problem. And the EMR providers attitude is “Screw it, Jack, just write the code. They have to buy it because the government says so.” I’m sorry, I digress.
Doctors and nurses don’t sit down much. We can’t wait until we find a convenient open work station to write the orders that will save the patient’s life. “I’m a doctor, Jim, not a damn transcriptionist!” Or something like that. Your hospital ER will never have enough computer workstations to allow you to provide timely care to your patient when the patients are lined up in the hallways on a busy Saturday night. I need to do my doctor thing on my handy, personalized tablet that I carry with me whenever I am playing doctor. It contains all my personal professional documenting tidbits. It has wireless access to the hospital medical records database, the laboratory database, the radiology database with images available for my review when the radiologist reading of my patient’s CT scan sounds a little–oh, I don’t know, a little ‘intoxicated’ perhaps–and access to my office server with all my personal patient records. It turns on and trusts that I am me because it recognizes my fingerprint, so I don’t have to keep coming up with a dozen new passwords every ninety days to keep some IT guy happily HIPAA compliant. It even can play me the basketball game that I’m missing while I wait for the nurses to come in to do my emergency case. I can talk to it, and I can write on it, and I can tap on it–I can do things that make my care of the patient easy,and efficient, and better. And I can decide how I want to do that, not the hospital administrator who got taken to that “conference” in Vegas on EMR implementation just before deciding which vendor to sign with. Do I sound bitter? Sorry.
I can write orders on my personal little tablet and they get sent to the hospital order computer thing. No more verbal orders. No more orders appearing in my electronic inbox screaming to be signed when I’m sure I was no where near drunk enough to order that stuff last night. But hell, got to sign it or it’ll just bounce back anyway. If it isn’t on my personal tablet, I didn’t do it. So there. My customized order sets, entered with a finger swoosh. Because I write the same orders for patients over and over again. I don’t want to have to make my own special sauce at every hospital, ambulatory surgery center, clinic, nursing home, homeless shelter, and my office. Then rewrite my personal note templates and order sets every time the IT guy decides to trip over the mainframe plug or do “mandatory system maintenance.” I’ll just keep that stuff with me, thank you. You figure out how to make it easy for me to jack my pad into your system, or I’ll go to a hospital that will.
All of this new electronic, streamlined patient-centric data input is to be automatically, consistently, and reliably entered into my personal database, not just yours, hospital administrator. I need to keep an accurate record of how many gastric resections I performed in the past 10 years and how many of those patients died within 6 months of surgery. I need to be able to easily upload my data to the various Boards, Colleges, and regulatory agencies that demand I prove that I’m a good doctor on a frequent basis. The data automatically and easily transfers to my billing software, not just yours. I’m the one generating the data, I think I should be benefitting from the effort. Harrumph.
I could go on. I’m sure you have lots of even better ideas. It’s a start, at least.
The almost over half-dozen regular visitors to this blog know that I much prefer “snarky” over “wonky.” However, this is serious stuff (see parts 1 and part 2 of this discussion), and I’m sorry to say that we’ll be eschewing the alliteratively amusing yet enlightening parable on this one for the tediously technical. Non-physicians are encouraged to move on to this entertaining dog video:
Seriously, it’s a great video, where these dogs are dressed up as US Supreme Court Justices. Even if you think you’re going to read this wonkishly tiresome tirade, take a moment first. It’ll make the process more palatable. Or whiskey. That works, too.
As you may have gleaned from Parts 1 and 2, I believe the current state of the EMR mandated implementation to be no less than a public health crisis. Over the top? I don’t think so. Please ask any physician, nurse, or health care practitioner in the country. I’ll wait.
See? We have to take meaningful, immediate actions to save patient lives and stop the documentary disease that is degrading medical record keeping even as we speak. My recommendations, therefore, are simple and immediate. Here they are:
Step 1: CMS is to immediately suspend the “meaningful use” mandate. Financial bonuses for those practitioners who implemented EMR should, of course, remain in place. These folks laid out a lot of good money based on this expectation. But we should immediately lift the threat of financial penalties for those practices that have not yet fallen down this well. This will also come as a great relief to the hundreds of practices that paid tens of thousands of dollars to install an EMR and have, in practice, chucked it into the closet because they found it unworkable, and therefore have not been able to meet meaningful use criteria.
Step 2: Immediately charge the Bureau of Weights and Measures to develop, in a timely fashion, minimum standards for electronic medical record keeping. Believe me, I would never have thought this up on my own–this was the plan back in the day when we were all talking about how great it would be to have an EMR. Seriously, the Bureau of Weights and Measures. These standards should be developed with input from the hospital industry, physicians, midlevel providers, nurses, and database engineers specializing in intercommunicability and security. This standards-making process should actively exclude representatives from the current EMR companies. The process should pay no mind for backward compatability concerns. These are the assholes who jumped the gun in the first place and forced us to buy dangerous, flawed products at exorbitant prices. Screw ’em.
Step 3. Require commercially offered EMR programs to be tested and certified by the FDA as a medical device. You think I’m kidding? I’m not kidding. The EMR is a medical device, a device that affects every single patient care interaction. A bad EMR is dangerous to patients. If the FDA requires that my felt tip skin marker be certified before I use it to mark a patient, yeah, I think they should check to make sure that the EMR doesn’t systematically screw up my orders. Not such a crazy concept. Oh, and the current multibillion dollar products already on the market should be retroactively required to pass this testing. No grandfather clause here–these programs are flawed and dangerous. Let the FDA sort it out by soliciting open comments from the consumers (that would be us). They will be swamped with helpful information.
Step 4. Governmental funding of a standard EMR to be made available to any and all practioners/healthcare facilities at no charge. This whole EMR idea began when a bright young (or maybe not so young, I wasn’t there) physician stood up (might have just raised his hand, not sure) and pointed out in a committee meeting during the early ’90s that the VA system had this pretty cool computer system that was better than paper charting. Say what you want about the VA, they were the first and the best at this EMR stuff. Then we all said, yeah, that sounds great, let’s do for the country what the VA is doing for our vets. Which, of course, we completely screwed up. So I propose that the federal government pay the money to improve upon the VA program to develop a simple, effective, broadly applicable, safe EMR. Then the federal government should make this program available as the minimum default option for any and all health care practitioners and facilities for free. Because we paid for it already with our tax dollars through VA funding. This program, of course, wouldn’t be the “bells and whistles” pretty thing that private industry will provide for profit. But having this free product would establish a baseline that all the other manufacturers will have to exceed if they expect us to lay out a lot of our hard-earned cash. Because right now we’re paying for a grossly inferior product.
I can already hear my Malthusiastic colleagues and trade group lobbyists saying that these recommendations are bogus, just more government regulation/interference with the free market, blah, blah, blah. Hey guys–we tried that, and we got this. So no, it didn’t work. Mandating purchase of a product immediately disrupts the free market. Believe me, if we weren’t forced into it, very few of us would’ve purchased the grossly defective products currently on the market at extortionist prices. The free market will be effective at leading to competitive pricing of an improved product if we take away the threat of near-term penalties and establish standards for this medical instrument. Then, and only then, can the free market work its magic.
I urge my colleagues, representatives, and medical societies to take up this challenge. Nothing in our current practice of health care is more important.
Sequel to Automated Malpractice and Digital Dishonesty, the first in a series dealing with the disaster which has befallen our health care system.
The single, most powerful tool employed by every physician in the care of patients is the medical record. It is the tool that impacts the care of every single patient. It is the tool that makes difficult diagnoses possible. It organizes and makes possible treatment modalities of every type. It allows collaboration between multiple health care practitioners. It prevents redundancies, harmful treatment interactions, errors in therapy. It makes possible quality improvement. It is the basis of clinical research and the datum upon which discoveries are made and ineffective treatments eliminated.
The medical record is the fundamental instrument of all patient care.
Take away the medical record and you might as well eliminate the use of antibiotics, of vaccines, of anesthesia, of blood typing–because every aspect of these treatments and their safe implementation requires an accurate, accessible medical record.
We are currently in the process of destroying the patient medical record as an accurate, dependable tool in the care of patients. As a direct consequence of our current actions, patients are being harmed. Quality assurance programs are being compromised. Future research will be flawed.
This phenomenon is being witnessed across every aspect of medical practice and in every patient care venue in this country. It is being experienced by every physician, nurse, and allied health care provider. It is incontrovertible. It is, quite simply, a national disgrace and a health care emergency.
The previous system of writing notes and orders with a pen in a paper chart was flawed, of course. It had the drawback of being, on occasion, uninterpretable. This led to errors. It had the drawback of being difficult to access. This led to inadequate communication, redundant testing, inadequate treatment. There was a definite and defined need for a better system. No one would argue differently. The new system needed to be:
i. accurate, ie., legible
ii. accessible to all patient care providers, including physician offices, emergency departments, hospitals, clinics, researchers, auditers
iii. systematically better than a bunch of pieces of paper stapled together: that is, it should intrinsically avoid dosage errors, medication conflicts, redundancies, etc.
Obviously, the new electronic medical record should be better than paper. Paper required health care practitioners in the form of nurses, PA’s, and pharmacists to interpret a physician’s desires and catch our mistakes before they reached the patient. If I ordered an antibiotic for a patient and the patient had a history of allergy to that antibiotic, a key patient fact that I was not aware of at the time of treatment, I relied on a clinically astute nurse or a pharmacist with the patient’s allergy flag to catch it and stop me before I injured the patient. Obviously, this was a problem, because some nurses and pharmacists, just like some physicians, are more experienced, conscientious, knowledgable, and savvy than others. Better to have a system that automatically and consistently checks every order against the patient’s documented record for such a life-threatening conflict.
The intent, at the national/governmental level, was to create and mandate the implementation of just such a system, a system to address the known failings of the existing paper-based system and improve patient care. But the reality rapidly superceded the intent. Through a tortured labyrinth of governmental committees and corporate boardrooms, we allowed the private, profit-motivated sector to leap into the gap between good standards and mandated implementation. In retrospect, we all should have seen this coming from many miles away. A government mandate that would require the purchase of a product by every doctor, hospital, and clinic in America? How much was that worth? Obviously, a great deal. So why wait for standards? Why develop a good system, a worthy product, when we can beat the competiton and sell a cheap, untested, poor product and require these guys to buy it whether they like or not, whether it works or not, even if it’s intrinsically dangerous?
Which is how we arrived to the point we find ourselves at now. Hospitals, physican practices, and clinics have been forced to expend millions of dollars to purchase poorly designed electronic medical record systems. Physicians across the country are now required to spend dozens of hours training up on a variety of unstandardized systems. They are required to become facile in the care of patients utilizing deeply flawed instruments that do not approach the minimum level of quality to allow safe patient care. The problems exhibited by the current state of the governmentally mandated art range from the simply irritating, to the amazingly stupid, and all the way to the point of the systematically murderous. I do not exaggerate. Allow me to provide examples:
The simply irritating: Each hospital at which I care for patients has its own brand of electronic medical record (EMR). Each EMR requires a unique user name and password. But some of these hospitals also have a separate system for charting in their Emergency Department. Many have a separate system to access electronic ordering, or another system to view radiology images, and another system to complete medical records, and yet another system to access labarotory results. Additionally, a separate system exists for electronic prescribing (also mandated), not to mention the required State Narcotic Abuse Database Access, also with a unique user ID and password. None of these systems, user names, or passwords are centralized or mutual in any way. And in the name of patient security, every one of these systems requires that I change my password every thirty to ninety days. But not in the same thirty day cycle. And not to anything that remotely resembled my previous password. Which leads to the constant juggling of multiple passwords with no chance of providing anything near efficient patient care. Maddening.
In those practices that require efficient clinical care, the implementation of the EMR has been uniformly disastrous. Patient office visits are now characterized by a screen interposed between the patient and the clinician, eye contact being a thing of past. Documentation is terse, inadequate, and often plain untrue. Most opthamologists and many other physicians have been forced to hire a new category of health care worker, the EMR documentician, to follow behind the physician and enter information into the computer. This has led to enormous additional expense and a new avenue for inaccuracy. I routinely hear from my colleagues how they have to spend their evenings at home doing EMR documentation from the patients that they saw during the office hours earlier that day, as if anyone could expect to recall the subtleties of dozens of patient interactions six hours later. Simply insane.
Amazingly stupid: Two of the hospitals I work in utilize the Allscripts EMR system. This system features the amazingly stupid feature of not superceding previous orders when such a feature is obvious and required. The diet order is a case in point. When a physician enters a diet for a patient, the previous diet order should be stopped. But no, this is not the case. So my patients routinely have several, conflicting diet orders running contemporaneously. I have discussed this on over a dozen occasions with the IT department, the physican IT laison, and presented to the IT/Patient services committee. No patient, I explain repeatedly, ever needs more than one diet at a time. They all patiently explain to me that such a change is not possible in this system. So my patients routinely receive trays for clear liquid diets, full liquid diets, regular diets, specialty diets, all at the same time. We depend on the dieticians to sort this all out on a daily basis. The waste in effort and food, not to mention the costs associated with cancelled procedures and the morbidity of inappropriately administered diets, is intolerable, but apparently unfixable with the current state of technology.
Systematically murderous: As I have mentioned, the key to good medical care is the conscientious application of patient information to allow informed decision making. The old system of paper charting was deficient in this aspect. A new, electronic system, should permit artificial intelligence applications to prevent many common errors, such as administration of inappropriate medications, wrong dosage, or duplication of treatments or testing. The system should be designed to enhance patient care and safety. Instead, we have a system that actually decreases patient safety and requires constant vigilance on the part of all health care providers to prevent system-incurred errors. This is the exact opposite of what is needed and appropriate, but it is happening constantly. In the postoperative setting, needed DVT prophylaxis medication is often ordered, but the system fails to reliably discontinue the same prophylaxis order from the preoperative period. This results in two or three doses of the blood-thinning medication arriving on the floor for administration to the patient. Usually, the nurse is sharp enough to detect the duplication and send the extra doses back to the pharmacy, but if there has been a shift change or breakdown in communication, extra doses may be administered with resulting complications. The system must be designed to make us better, not require us to police new, intrinsic challenges to good patient care. The current situation is unacceptable.
Not to mention, medicolegally compromising: One of the nationally implemented EMR systems fails to close out patient encounters when the patient is discharged. This leads to the occasional situation wherein a treating physician accidentally places multiple orders into the system for a patient, not realizing that the orders are being entered into a patient encounter that is no longer appropriate. There is no systemic prevention for such an error. So in the middle of the night when an emergency patient requires admission to the ICU or is to be taken emergently to the OR, dozens of orders are entered, but not implemented. The physician is told that his orders are missing, so he or his associate/resident/PA re-enters all the orders. But now there exist a slew of orders added to a past encounter that were never acknowledged or administered. Lord only knows what will be made of that when the chart is eventually reviewed for quality or legal concerns. Good luck with that in a court five or six years after the occurrence.
What is obvious from the foregoing tirade is that the current implementation of the mandated EMR fails to meet the minimum requirements that I laid out in the beginning of this post. It is legible, but it is inaccurate. It is not accessible, due to a complete lack of interoperability standards. As a system, it fails to improve patient care, but rather introduces an entire new class of systemic challenges to good, safe patient care. The current system is simply unacceptable. In the next post, we will discuss the remarkably simple solution to the current situation.