[MUSIC] Hi, I'm Nicholas Genes, and the focus of this lecture is going to be on how electronic medical records actually work. We'll be revisiting some material from earlier in the course to explain some of the idiosyncrasies of electronic health records. Here's a screenshot of a common electronic health record. You might say, it's very busy, very crowded, that there is a lot going on. And this is just one screen for one basic patient care function that's kind of an overview for a doctor to look at a patient record. And you'll see there's a lot of navigation areas on the left, there's some action items on the right, and in the middle a lot of details about the patient. The patient's name and photo is up near the top. The doctor's name is up near the top and the vendor. But other than that, it’s a lot of crowded material, it’s difficult to quickly navigate through. Yeah, it’s very dense. How did it get this way? Is this the best way it can be displayed? Really electronic health records, electronic charts, I think if we were starting today from scratch we wouldn't have this kind of system. But EHRs are built upon a long historical legacy and a technical legacy. They are the product of thousands of decisions over many decades. Paper charts began, in earnest, the conventional form that we use today, back in the 1960s when Professor Larry Weed developed what he called the SOAP note, S-O-A-P. It stands for Subjective, Objective, Assessment, and Plan. Subjective being what the patient describes, their symptoms, their history, their understanding of disease. Objective being exam findings, lab results, radiology reports. Assessment being the doctor's summation of the subjective and objective. There would medical decision making about how the patient is presenting and what has been uncovered. And then finally, P is for plan, it's what's going to come next. And this worked great on paper, and it's how I learned it, and how many of my colleagues learned how to write notes. And I have fond memories of these cylindrical carousels where patient charts would be held in folders, and we would write our notes in this SOAP format. And we'd also execute our plans, we would put little flags in to signal to the clerk our orders. The clerk would take the folder, enter the orders and then later, the results would come back and be inserted into these folders. And we really lived them and died by them. And we really couldn't afford to lose these folders as we were working on patients in the hospital. EHR over many years has assumed the role of these paper charts and the order entry and the lab results. And it has become a massive software system that we saw a screenshot of just a moment ago. But the EHR is different for different people. It has many different roles. Just among doctors, there are very different views if you're seeing a patient in the clinic versus in the emergency department versus in the operating room or an inpatient ward. And finally, labor and delivery has unique demands of the EHR because you're going from one patient to two patients. Also, you can imagine that if you're a primary care doctor, you still want to be able to log in and view the record of a patient that has been hospitalized, the patient of yours that has been hospitalized. But the view that you get and the tools that you have available to you are very different depending on whether you're in charge of the patient's care or just following as a consultant or following from afar. And that's just the doctors, there is the nurse role, clerks, case managers, all these folks are writing notes, are executing commands, are using the EHR in very different ways. And so their view of the EHR is very different. And finally, there are the roles of the administrator, the researcher and the student. These roles have less executable actions, less things that they can do on a patient or to a patient, but they have very advanced reporting capabilities. So administrators can look at a lot of patient data, one on one, one patient at a time, or in aggregate, a whole floor of patients, a whole hospital full of patients, or a whole population of patients. And finally, speaking of patients, the patient's role in the EHR is perhaps the most important. The patient has to be able to log in and see some form of their records and be able to act on that. To accomplish what it does, the EHR interfaces with a number of other systems throughout the hospital, throughout the country. Every time I log into an EHR, the system in the background quickly checks and makes sure that I have my appropriate credentials, and that my preferences are properly set. So that I always log in and see the emergency department, and see it the way I like it to be laid out. Every time a patient appears, their medical record number is checked against the master patient index. And they're also registered and logged in the scheduling software. Every time a patient moves around the emergency department of the hospital, the bed management system is interfaced. Every time labs are ordered, radiology tests are ordered, those signals, those messages sent out to those systems. And finally, when I order medications in the emergency department, the signals are sent to the pharmacy system in our hospital. When I e-prescribe, messenges are sent out to local pharmacies. Finally, depending on the test results that are received, depending on the commands that are executed, depending on patient consent, the data that is collected during an encounter is shared with the Health Information Exchange, and in some cases shared with federal repositories of data. So, let's look back a little bit to the past decade. Electronic health records had been building up their functions, their capabilities, but adoption was still very slow. And as I touched upon, there were a lot reasons for that, historical reasons, cultural, financial and legal reasons. For these reasons, hospitals and clinics stayed with paper longer than a lot of other industries. And when you think all of different roles that are involved and all the training that's necessary to get doctors, nurses, case managers, administrators, clerks all up to speed on how to use the electronic health record. I guess it's not terribly surprising that these systems, these doctors and healthcare institutions were really reluctant to upgrade and to adopt the electronic health record and to move away from paper or to adopt more comprehensive instances of electronic health record. There is these risks of it's going to be difficult to train people, it's going to make it be harder to move patients in and out of the clinic or in and out of the emergency department or hospital. And it would dissatisfy both providers and patients. And yet the government recognized that the population is getting older, care is getting more fragmented, more divided among multiple specialists and costs were increasing without a clear benefit. And the government really viewed the electronic health record as a way to at least get a handle on that situation. So this has been a bipartisan effort over decades. George W Bush in 2004 announced the creation of the Office of the National Coordinator of Health Information and Technology. And this was a new role at the time that aimed to standardize a lot of different practices, a lot of different systems, and really worked with a common framework to get electronic health records, more broadly implemented across the country. When President Obama took office, as part of the American Reinvestment and Recovery Act, a big stimulus package, there were provisions for what was called the HITECH act, this was what was called the most important piece of healthcare legislation in 20 or 30 years. And this is before the Affordable Care Act. So, it gives you some instance of the scope of this. This was a big incentive program to really encourage doctors and hospitals to adopt the electronic health records. And the program that was decided on was called Meaningful use and the code use was going to be executed in three stages. And it wasn't enough for doctors or hospitals to simply buy an electronic health record and show their receipt to the government and again reap the get the benefits, the incentives. They had to show that they were meaningfully using the software, and the software had to meet certain criteria. And to have a lot of functions and some of the functions are displayed on this slide. It worked, in that meaningful use for every bit spur electronic health record adoption, but the vendors really rushed to capitalize on these incentives. They added a lot of new features to their systems, and software that had been stagnant for many years or had been slowly evolving over many years, very quickly acquired a lot of new capabilities and features. And sometimes these features were not deployed in the most responsible fashion. The systems became even more unwieldy and difficult to use. One quote from a lobbyist said, meaningful use is simply enshrined, mediocre software across the United States. I show in this slide, there's a lot of different screens, the top half of the screen shows a really famous electronic health system. In Indiana and how little it had changed between 1984 and 2010. In the background a lot of things were happening and the systems's actually quite powerful and it's been well reviewed over the years. But from physician's perspective it really looked the same for decades, contrast that with desktop software or personal computer operating systems that adopted the graphical user interface and really gained a lot of user friendly features over that same interval. I really think it's instructive to kind of compare electronic health records to the operating systems on personal computers. Let's consider some core functions of EHR and EHR's were originally designed as billing software, designed to capture just a few details about a patient's encounter or hospital stay. And capture some details about things that were done for the patient during that stay. And then send back to billing company for reimbursement. Over time the EHR accumulated more functions order entry, lab and radiology result reporting, and finally documentation. Physicians could write their notes, nurses could write their notes all into a patient record. This evolution happened in fits and starts over many years. And as you look through the different aspects of an EHR you can sometimes see relics of different implementations, different modules in the HR and they all follow slightly different rules. The chart is called different things in different places, some of the buttons look different, the navigation looks different. It's kind of, it's not a seamless experience as it often is with modern consumer software. Let's compare that to personal computer OSs which developed really just to store and find files, and then the OSs took on basic file editing capabilities, and web browsing, and finally e-mail and calendar functions. The modern computer that you buy at a store would have music and photo editing capabilities, and released viewing capabilities, and even has built in social media functions and integrations. Today's electronic health records similarly supports a lot of messaging between providers and sometimes between patients and providers. And has some more sophisticated reporting technology. But in general, like I said, the electronic health record remains a more convoluted system, but that's not as seamlessly integrated as personal alliances are. Similarly there are a lot of plug in modules, and a lot of customization that is possible on your computer, and in electronic health records. An EHR, you can get custom scheduling software, clinical decision support tools, discharge instructions. Even entire patient portals can be customised. This is analogous roughly to downloading an application, a program for your personal computer. Originally those were for helping with programming, and spreadsheets and statistics, word processing, slide presentations and now photo editing and more customized applications. However, as you navigate your personal computer, I think it's pretty clear you know when you're using the operating system versus using an app. Yeah, there are some standard conventions, but in general, I think, PC users can tell the difference. And yet, if you ask a doctor whether they're using a custom module or a core function of their electronic health record, they often don't know, because the experience is so disjointed, so unseamless, so seamed, I guess I could say, that the boundaries are a lot less defined. Let's look at a few specific examples of how electronic health records work. Take the example of drawing blood, collecting labs, getting lab results. As a doctor in the emergency department, when I see a patient and I decide the patient might need blood testing, I'll sit down at my electronic health record, log into the patient's chart, and type in my order or select a check box with my order. I might see some clinical decision support, some CDS related to that order. In the background the electronic health record is checking my patient's age, past medical history, circumstances of their visit and might get a warning but maybe the test is not the most appropriate test and I should order something else. And I may respond to that CDS and change my order. Let's say, I go ahead, I place my order, a lab is then notified to expect the test. And in fact, they'll do a quick check and see if a similar test has been ordered in a small duration of time. They actually may cancel my order if they see that a colleague has ordered the same test in a very short time period. A label is printing close to the patient so that nurse can take that label and affix it to a tube. And of course the nurse will receive the doctor's orders as well, and a mark that order as acknowledged. Nurse will then go to the patient, draw blood or in-patient you can imagine a phlebotomist accomplishing this role. Visual scan and affix the label and send that tube up to the lab system. And nurse will then document that an IV has been placed and the patient tolerated the procedure. When I go back to the chart later I can see that my lab order has been marked as collected, and when it gets sent to the lab it's marked as in process. And the lab will perform its testing upstairs. The results will appear in the electronic health record I can then mark that I acknowledged the results, I saw them. And depending on what the results are, a lot of other things can happen. There could be clinical decision support about a positive test result. I might get a pop up warning me that my patient is at risk of some condition. Critical values, values that are so far beyond normal that they demand immediate attention. Those are often conveyed by phone call. A lab will actually pick up the phone, see who ordered the test, and then call that provider to warn them that something very amiss is going on. Other providers can look at the test results, and this is not just providers that are in the hospital or in the emergency department. But even the primary care provider can see it. And ultimately patients can view their results. Usually there's a step where someone releases that data to the patient health record. So someone should review that data before it's made public because it might be spurious or it might require a conversation. Every event that is a part of this sequence is timestamped so later we can go back and look at the turn around times for each step. Did it take too long for the lab to get drawn or from the lab to get sent to the lab system, or maybe the lab system took too long in processing the results or maybe the results just didn't show up fast enough in the electronic health record. All of this can be reviewed and analyzed and we can fight bottlenecks. Also lawyers can discover this data if there's a bad outcome, a malpractice suit. Every item is scrutinized and every path in this mesh work is reviewed. And I like to think that we are continuously making effort to reduce steps, to reduce redundancies. To simplify the workflow so that it happens faster and more reliably. But we can also think of some special circumstances here, such as in the emergency department when we don't know the name of the patient. It doesn't happen that often, but we will have some John Doe patients that come in, and they're unconscious. And we need to send blood test on them. How, how do we do that? It's a special circumstance where a special medical record number is created for that patient and all those test results are aggregated and yet it has, the system has to be flexible enough so that when we do discover the patient's name when they regain consciousness we can then assign that name to all that data that has been collected. There's also situations where sample QNS, quantity not sufficient. The lab has to communicate to us that the blood test that we sent will not be able to be processed. And this is a time sensitive thing as well. It's not as important or not as critical as communicating a critical lab value but I certainly don't want to be waiting in the emergency department for hours for a test result that's never going to come. So we have special ways of signaling to doctors that they have to reorder their test. And finally the lab will sometimes comment on the interpretation. We're not simply giving a number back, a potassium level, a sodium level, but we're getting some text as well as the number. To kind of explain why some readings might be out of whack. Let’s talk briefly about radiology because it is even more complicated than the laboratory system we discussed. Really when I order a CT scan in the emergency department and truly the start of a consultation. Imagine a case where I suspect a young person has appendicitis and I order an abdominal CT scan. I will then place that order into the patient's chart and a radiologists will review that order. They'll see my name, my contact information. And some patient details including why I suspect they have appendicitis. The radiologist may then pick up the phone and ask me some questions because they need to protocol the study. They need to decide how much radiation to dose, whether IV contrast should be given alongside the radiation to better visualize the structures. And they have a lot of fine details that they have to base the study protocol on. Not just the diagnosis but patient characteristics, how well their kidneys are functioning for instance, how heavy the patient might be, and machine characteristics too that go into their decision making. So these orders are not always straightforward. And then carrying out these orders involves scheduling the study with the radiology scheduling software. Transporting the patient to radiology, performing the imaging, uploading images to a PACS, picture archiving computer system. Then the radiologist has to view those images and interpret the results, dictate that interpretation into a system and have that associated report make its way to the PACs and also back to the electronic health record where I can read it. And as a doctor in the emergency department I want to be informed of each step along this process. I'd like to know when my patient has gone to radiology, when the images have been procured, and when the radiologist has dictated their report. So, a lot of signalling back and forth between the electronic health record and the radiology systems. It is a complex process. Finally I want to talk about doctors notes, and how these are recorded in the electronic health record. I chose this image because I feel this is the patient perspective, the number of times in the doctor's clinic. Doctor has a big monitor that's obscuring, that's in between the patient and the doctor, and the doctor sometimes spending more time on the computer than they are at the bed side. This is a regrettable state of affairs. But notes can be written during or after the encounter. Doctors are compelled to document a history of present illness, past medical history, review of systems, exam findings and medical decision making. And, they may use templates to help guide note creation. But, in general they have to click a lot of check boxes or write a lot of free text to capture all the details of the encounter that they are expected to capture. Doctors can, you'll sometimes hear, going up to E&M level 5, this is an evaluation in management criteria, that asks for so many check boxes, or so many cardinal features of a presentation to be recorded. In addition, if a procedure is conducted in a clinic or the emergency department, the doctor will be documenting on different procedures, counseling that takes place during an encounter, so lots, lots to do, a lot of check boxes in some cases or paragraphs to write, certainly the medical decision making component tends to be a free text paragraph or several paragraphs if the case is really complicated. As the doctor writes their note and then signs the note, the documentation is then available to other providers, to the nurse. And if the note is released to the patient, then it's accessible via patient health records as well. Documentation is again often driven by billing, the reason that doctors are clicking all these check boxes is because this is what we have been trained to do to get reimbursed for the different visits that we see. Often before the chart goes out for billing, it is reviewed by coders on site and they will sometimes capture things or identify things that are missing and then send it back to the doctor through the EHR to finish the documentation. These are deficiencies that are identified and highlighted for the doctor to review. Then at a scheduled pre-arranged time, the note goes out to coders who review whether the note fits the diagnosis for the visit. And then reimbursement happens as the coders and the billing company interfaces with the insurance. I read a stat many years ago that feels true and I think has survived through electronic health records as they've become more wide spread, that 90% of notes are never read again after they are coded which is kind of regrettable that so much time is spent on these notes and they are never seen again. But in those 10% of cases where they are seen again they can be very valuable. [SOUND]
