The EHR Debate: Fighting the Last War?

Fighting the Last War?

…Yes, we’re having the wrong fight by focusing on old problems. The EMRs that are producing the studies we’re fighting about are the current equivalent of 1990s EPR implementations. In general they’re hard to use and require lots of money and training to produce halfway decent results. The real improvements from IT came when user-centered tools came to consumers and then to business with Web 2.0 and new devices like the iPhone.

It may take months of training on Epic or Cerner to get a doctor or nurse to be three-quarters as productive as they used to be, but my two-year-old daughter can fire up an iPad and play games and watch videos with no training.What we’re seeing every day at Health 2.0 is a whole new generation of data-driven applications and devices that are going to make the health care user experience much more like the one my daughter has.

When we get there, the real improvements in both productivity and safety, as well as in quality and even cost, will emerge and we’ll wonder why we ever were having this fight.

Doctors Look For A Way Off The Medical Hamster Wheel


“Working at Starbucks would be better” – Dr. Ben Crocker

Internist John Henning Schumann wrote a short NPR bulletin on the new white paper In Search of Joy in Practice A Site-visit Analysis of Twenty-three Highly Functional Primary Care Practices. If I were a front-line physician I would be keenly interested in the findings of authors Sinksy et al. As Dr. Schumann writes:

As the hassles have gotten worse, I’ve seen many colleagues jump ship. Most have sold practices to larger enterprises like hospitals, leaving the risks of business management to the buyers.

I’ve thought of bailing myself and entering the so-called direct practice model. I wouldn’t take insurance then, and patients would have to pay me directly via monthly subscription.

But there might be another way. Dr. Christine Sinsky, an internist in Dubuque, Iowa, is looking for one. She has made it her mission to find ways to mitigate the drudgery of modern doctoring so that we can find joy in our work.

With funding from the American Board of Internal Medicine Foundation, she and four colleagues (including her physician husband) travelled the U.S. in search of practices that provide top-notch, effective primary care, while making the work satisfying for the doctors and other health professionals.

I’m still reading the 29-page white paper. While there are many reported cases of successful practice re-design, one common theme that continues to astonish is the loss of productivity caused by the introduction of electronic medical records (EMR). Computers have been a primary driver of the dramatic compounding productivity improvements seen in every other sector that I can think of. Medicine seems to be the unique failure case. I have seen so many physician complaints that they feel they have been turned into data entry clerks. Surely there must be successful EMR implementations – but those stories do not attract media coverage. On the failure side, I noted one example of how not to do EMR back in May – at Beth Israel Deaconess Medical Center. To me, such maladroit computer implementations are shocking. For example, if innovations like the Google EMR effort had not been killed by vested interests, then today physicians would have low-cost access to modern, efficient, easy to use systems.

(…snip…) The volume of work associated with record keeping has increased over the past decade, with the introduction of electronic health records, the emergence of quality monitoring initiatives and the increasing complexity of billing regulations. The burden on primary care practices cannot be overstated. Tasks that took a few seconds in the pre-EHR world can take several minutes in the electronic world. Visit notes that previously were a few paragraphs of text, are now lengthy disjointed documents, formatted on a billing template, extending over many pages,  and complicating, rather than facilitating, the cognitive work of finding key information.

I’m keen to hear what physicians think of this Sinksy et al white paper. It looks to me that her team has done a real service by systematically researching practice remodeling cases that have made doctoring worth doing again. E.g., here is one happy case that caught my attention:

Quincy Family Practice: In the “Office of the Future,” a pilot at one of the Southern Illinois University School of Medicine family practice residency clinics, there are no physician desks in the exam room. Their absence reflects the innovation: the doctor does not sit at a desk and attend to the computer. He sits on a rolling stool and attends to the patient.

The MA reviews the pre-appointment questionnaire in which the patient is asked “What three questions would you like answered today?” She completes medication reconciliation and begins recording the HPI, working through a structured history for each complaint or condition, checking off boxes and adding free text as needed. Depending on the complexity of the visit, this may take 8-15 minutes per patient. The process generates components for billing and frees the physician from keyboarding standard elements of the history. When finished, the MA returns to the nursing station and briefs the doctor.

The MA and physician return to the room together. Unlike the first minutes in most physician office visits, Dr. Joseph Kim does not move directly to the computer, but to the patient and any family members. There is no signing in and orienting the visit around the computer. The physician’s full focus is on the patient. The MA listens to the physician and supplements the history she recorded earlier.

The MA is able to move quickly between templates for recording the history and the exam. Once or twice in each visit, when information comes too quickly to keep up, the MA jots a few notes on paper. Because the need for scribing ebbs and flows during the visit, we observed that the MA was always able to catch up within 15-30 seconds.

Finally, Dr. Kim verbalizes his assessments and plan, which the MA also records. She writes the diagnoses and plan, e.g., “High cholesterol, simvastatin 40 mg daily, appointment 1 month with lipids” on the pre-visit questionnaire to give to patient; queues up prescriptions, which the physician signs between patients; and checks off lab and next appointment information for the patient to take up front to the scheduler.

In addition to the two MAs, Dr. Kim works with an LPN, who uploads into the EHR new patient questionnaire data regarding past, family and social histories, fields calls for prescription renewal, provides phone advice and manages coumadin by protocol. When the MAs need extra help she also rooms and scribes.

The clinic is calm and runs on time. There is no sense of chaos and running behind as in many clinics. At noon all lights are out. No one is eating lunch at the desk while answering phone calls or responding to email. The physician and staff leave the building for lunch and refreshment, ready to return for the afternoon appointments, beginning at 1:30 pm.

The pilot, begun in 2010, has expanded to two faculty physicians. Family medicine residents rotate through the clinic for an experience of how practices can be designed. The “2 + 1” staffing (two MAs plus 1 LPN) results in staffing ratios similar to the 2.5-3.0 clinical assistants per physician at the University of Utah and Newport News. Quality, efficiency and patient experience have improved.

Other clinics in our study have developed similar models that increase the collaborative nature of the visit and decrease the keyboarding work for physicians. At Martin’s Point much of the history and exam is recorded by the nurse, who also updates the chronic care plan, and does the computerized order entry. At the APF the MA rejoins the physician toward the end of the appointment, assisting with computerized order entry, prescription processing and scheduling next steps.

What about the economics of the team-based care that is common to all of the 23 successful cases? This study was not designed to do in-depth cost/benefit analysis. Like Quincy quoted above, many of the cases were implemented using support from organizational subsidies. Some sites covered the higher staffing costs by higher productivity:

How can a practice afford to hire an additional staff person without fundamental payment reform? Six of our sites: Mayo Red Cedar, Newport News, Allina, Medical Associates Clinic, Martins Point and the Cleveland Cliinc pilot, work exclusively in a fee-for-service environment, without organizational subsidy. Most determined that an additional two patients per day covers the cost of the additional MA/nurse.

I would like to see a lot more systematic work like this Sinksy et al study – investigating process innovations from a full cost-benefit perspective. Like the patient safety innovators (such as the Cleveland Clinic), learning best-practices from the innovators should lead to the widespread adoption of team models that work. I’m sure I don’t appreciate all the impediments to change (like government regulations), but isn’t it obvious that burned physicians are not the path to quality, affordable care?

Why Electronic Medical Records are hated by doctors and nurses

Shirie Leng, MD is a practicing anesthesiologist at Beth Israel Deaconess Medical Center in Boston. She blogs regularly at Medicine for Real – one of our top ten healthcare blogs. This post was so good I don’t see how to summarize it – please read the original here A Top Ten List for EMR and subscribe to Medicine for Real – Dr. Leng will give you an insider’s perspective on healthcare – as practiced at the top level. You don’t want to think about what goes on at the “average hospital”.

February 14, 2013
Today I’m doing anesthesia for colonoscopies and upper GI scopes. Nowadays we have three board-certified anesthesiologists doing anesthesia for GI procedures every single day at my institution. I’ll probably do 8 cases today. I will sign into a computer or electronically sign something 32 times. I have to type my user name and password into 3 different systems 24 times. I’m doing essentially the same thing with each case, but each case has to have the same information entered separately. I have to do these things, but my department also pays four full-time masters-level trained nurses to enter patient information and medical histories into the computer system, sometimes transcribed from a different computer system. Ironically, I will also generate about 50 pages of paper, since the computer record has to be printed out. Twice.

No wonder everyone hates electronic medical records (EMR)! I don’t know anything about computers, and I don’t know what systems other hospitals have. I may be dreaming of a world that doesn’t exist or that world is here and I haven’t heard about it. Nevertheless, here’s my wish list for a system that doctors would actually want to use:

1. ELIMINATE THE USER NAMES AND PASSWORDS. You can’t tell me that in this day of retinal scans and hand-held computers that there isn’t a better way to secure data. What if each person had their own iPad that you only have to sign into ONCE a day that automatically signs your charts. If you’re worried about people leaving them sitting around use a retinal scan or fingerprint instant recognition system.

2. ELIMINATE THE PAPER. If you’re going to have full-time people entering data for you, why print it out? It’s on the computer for anyone to access.

3. ALL DATA SYSTEMS MUST BE COMPATIBLE. You can’t have patient data entered in one place that doesn’t automatically import into another place. If my anesthesia record can’t talk to the hospital OMR, I have to RE-TYPE everything in, which is completely ridiculous.

4. EVERYBODY HAS TO USE THE SAME SYSTEM. Everybody, state-wide. Right now, electronic records from a nearby hospital are not available at my hospital, even though the two hospitals are right across the street from each other.

5. DON’T MAKE ME TURN THE PAGE. All the important information about a patient should be on the first page you open when you look up a patient. I shouldn’t have to click six different tabs. Specific to anesthesia, all the relevant data about the patient including what medications they have received during the case should be automatically displayed on the screen when you start a case. Specific to primary care, all the latest labs and data, recent appointments with specialists, current med list and anything else the doctor wants to see commonly should be right on the first screen.

6. DON’T MAKE ME HAVE TO REPEAT MYSELF. If I do eight cases the same way, with the same documentation required for each case, I still have to enter that documentation each time. If I’m seeing 20 patients in primary care clinic and the rules require the same documentation for each person, I shouldn’t have to enter that documentation each time.

7. INVEST IN DEVELOPMENT OF REALLY GOOD VOICE-RECOGNITION SOFTWARE. If I’m sitting across from a patient, I want to look at them and talk to them, not talk to them and look at a computer screen. If my mother didn’t make me take typing in high school, I don’t want to have to spend 3 hours after-hours every evening pecking my conversations with my patients into a computer, or worse, checking boxes electronically.

8. GET RID OF THE WIRES. In this day of wireless, why am I still tripping over monitoring wires and untangling cords? My spin bike at the gym can pick up my heart rate without a wire. Why can’t my anesthesia monitor?

9. IF YOU NEED A TYPIST, HIRE A TYPIST. Every time a new rule or documentation requirement pops up, which in my institution is daily, it is always laid on the nurses to add that to their computer records. Nurses used to be nurses. Now they are data-entry specialists. Their checklist and pre-operative paperwork is longer than mine. And they aren’t doing any diagnosis or treatment.

10. TRIPLE BACK-UP THE SYSTEM. Computers crash. Paper doesn’t. There’s got to be a way to make the system rock-solid reliable.

There are a mind-boggling number of rent-seeking interests in the US medical system. I would really like to know the true story of how Google’s effort to offer free EMR was defeated by these interests. Google knows how to build, operate and refine such systems at scale – and FAST. What we have now is a complete mess – which will take decades to rationalize.