 Good afternoon, ladies and gentlemen. Welcome to this lecture on electronic medical records. My name is Yashik Singh and I'm from the Department of Telehealth at the Nelson R. Mandela School of Medicine in South Africa. In this lecture, we're going to actually speak to you about what electronic medical records are. We're going to describe some disadvantages of the current way in which we practice medicine that is paper-based records and in fact show you the advantages of electronic medical records. We have two videos which are examples of electronic medical record systems used. And we're going to discuss very briefly how you would assess the benefit of a computer-based record system or an electronic medical record. All videos used in this presentation is freely available from YouTube. So the data that is stored when a clinician stores or writes information about an interaction between himself and a patient is called a medical chart or a medical record. And in fact this idea stems from the Greeks. The Greeks used to record this information whenever a clinician had an interaction with a patient. But when they record the information, they have to actually chisel the information onto stone slabs. So you can actually imagine how difficult it was to be a doctor in those days. Imagine carrying all your patients charts with you each time you want to see them. Thanks to the invention of paper by the Chinese, it became much, much easier to actually record patient information on paper. Of course we can imagine writing on a piece of paper is much easier than chiseling information on a stone slab. So why should these clinicians actually record this information? And basically it's to recall observations, to inform others that's communication in case you want a second opinion, to instruct students and gain knowledge that education, to monitor performance as well as to justify interventions. And this was actually discussed in the introduction to medical informatics lecture. So what is lacking in the current way in which we are doing things? Now in this new technological century, we have computers. We have technology. We have different ways in which we can communicate with people. And we find that the use of paper is becoming very problematic. And that's because of the inherent disadvantages that are in the paper itself. And these disadvantages can actually be divided into four sections. Organizational, logistical, research and passive nature of paper. Organizational basically deals with how we store the data in the chart. How do we structure the data in such a way that we can actually retrieve it to that's organizational. How do we organize the data in such a way that it can be retrieved? Logistical refers to the actual use of the data. Research is actually quite self-explanatory is how do we use the data for research. And the passive nature of paper is the fact that the paper cannot actually tell if you're making a mistake or not. So let's discuss these four aspects with an example. This example might be a little overdone but I will throw in a few things that actually happened to me. I'm from Newcastle and I once went to the doctor and I go to the doctor, I go to the receptionist, I ask the receptionist, can I please the doctor? She says yes. And she didn't ask my name and she goes and tries to find my chart. Now imagine her looking for a chart in a room where there are thousands and thousands of paper-based charts. What are the chances of those charts actually being in alphabetical order? Think about it, at the end of the day the nurses are tired, they're frustrated, they have seen many, many cranky doctors and patients and then they're actually expected to put paper charts, hundreds of them, back in the correct order. I myself can't even keep my own small filing cabinet organized. I can imagine how difficult it would be to keep a room full of paper charts organized. So she starts looking for my paper chart. What happens if she doesn't find it? If she can't find it, she has to create a new paper chart for me and all of that information in the old paper chart is now lost. My medical history is lost. All of the information that is actually required for a proper diagnosis of certain types of diseases are lost. So let's say she finds it, she finds my record, she takes it, she gives it to the doctor. Now in the actual incident that I have had when this receptionist actually found my record and she was walking to the doctor, she actually dropped my ECG. And while I was walking to his room, I saw it and I picked it up and I saw it had my name on it and it was my ECG. Now imagine if I didn't pick it up. Okay, an ECG, maybe that's not so bad, but what if it was results from an HIV test? What if I was HIV positive and now my wife or my parents or my boss looks at it? That is an invasion of my privacy. Those are security disadvantages to paper-based records. Confidentiality, privacy disadvantages to paper-based records. So eventually I made my way to the doctor's table and there were tens of charts just lying on his table, pieces of paper everywhere. He looks at this chart, he asks me what's wrong, I tell him the symptoms, he starts paging to this paper chart, trying to find maybe a piece of information. Now doctor's handwriting is notoriously terrible. We all know that. Now imagine him trying to understand his writing where he was actually rushing and writing and now he can't understand it. That's another disadvantage of paper-based records. Doctors write badly and they cannot actually read their own handwriting. So let's say now paging through this thick chart of my medical history, we can assume I have some kind of chronic disease and diabetic and therefore my chart is actually quite thick. He eventually finds this piece of information that he was looking for. Of course he could understand his handwriting, he realized what type of treatment I need, he writes it again on a piece of paper and he just puts it into the file. He doesn't put it in any chronological order. Maybe it's a small piece of paper like my ECG and while he gives it to the nurse, the nurse will drop it down, it'll get lost. Maybe the file gets misplaced in his table under all of everyone else's files. There are lots of organizational and logistical disadvantages when it comes to paper-based records. Where and where in South Africa is actually a law that states that a paper-based record needs to be kept for a certain number of years and I think it's actually 10 years. Paper, as you all know, is something that is very very fragile. We've all written on a piece of paper and have left it in the sunlight for a few hours only to find that the paper becomes brown, the ink has faded. Imagine keeping this record for 10 years. How do we ensure that we will still be able to actually read the writing on this piece of paper? That the paper doesn't have dog ears and it's torn. That the paper doesn't become brittle with age. Backups. Backups of paper-based records or paper-based systems are tremendously difficult. Clinicians don't have time. They're in a high-paced and very very frustrating field because they need to see lots of patients in a very small amount of time. You can't expect them to write in triplicate, even if you have carbon paper. So making backups of paper is very difficult. The other disadvantage associated with this is also cost. If we do make backups, there's tremendous cost in terms of buying paper, storing paper, creating the environment where paper's lifetime can actually be improved. Paper can also be used only at one place at one time. If you, in fact, want to use paper somewhere else, you have to make a copy of it, unless you are creating redundancies, research. Imagine trying to do research if you only had paper information. It is so difficult. Imagine going into a hospital, a public hospital, where there are six, seven, eight thousand paper files. And you now want to do a study that tries to find out if there's any correlation between smoking and heart disease. So you go through each and every one of these papers to find all the people that had heart disease, to find all the people that are smokers. And then you've got to find all people that had heart disease and were smokers. So retrospective studies are very, very difficult when you only use paper. The passive nature of paper is also a very big disadvantage, especially in the highly paced field that many clinicians find themselves in public hospitals. If you are making a mistake, paper cannot scream at you and tell you, listen, you are making a mistake. Paper will not tell you, you know, putting in a person's height as 16 meters might not be right. Maybe you meant 1.6 meters. Paper is passive. It can't inform you that you are making mistakes or suggest to you other alternatives in terms of therapy, or suggest what patients may be suffering from. So these are basically the four categories of disadvantages of paper-based records, organizational, logistical, research, and the passive nature of paper. So seeing that it is a technological century, we can actually create or use technology to overcome many of these problems. And that leads to the advent of electronic medical records. And this field is actually an evolving field. An electronic medical record, in fact, has no clear definition of what should constitute it. I will speak more about this later. But very basically, an electronic medical record can be defined as a systematic collection of electronic health information about an individual or a population of individuals. It is in fact a record in a digital format, in an electronic format. And it is capable of being accessed in different healthcare environments. And this is mainly because of networks, internet, the TCP-IP port calls, etc. Again, an electronic medical record makes comprehensive use of data. So the data that is stored has a very, very wide range. And depending on who you are, depending on your own definition of electronic medical records, there can be many things, many pieces of information that can be stored in these electronic medical records. You could have demographic information. You could have medical history, medications, allergies, lab tests, radiology images, vitals, billing information, administration information, etc. In our local context, we actually have a difference between what is called a computer-based patient record and an electronic medical record. Basically, a computer-based patient record is patient-orientated. So it only has an electronic version of the paper-based record. But an electronic medical record not only has the electronic paper-based version of the paper record, but it also has other things that help the patient, like a clinical decision support system, etc. Coming back to the slide, I just want to bring your attention to the different names that we can call an electronic medical record. We can call it an electronic patient record. We can call it an electronic health record, etc. I think for the purpose of this lecture, let's just assume that all of these different names actually refer to the same thing. It is a collection of electronic health information with other tools that a clinician can actually use to improve quality of care for a patient. The next two things that I actually want to show you are videos by clinicians. These videos actually describe to you what electronic medical records are. There are two examples of them being used in private practices in Hawaii. I want you to pay careful attention to how it is used, to how it is organized, to how it actually makes the lives of the clinicians, the admin staff, and everyone else who is involved in this clinical environment much, much easier. Please watch the two videos now. These videos are called video one and video two and you'll find the link to it on the left hand side of your screen. Also, these videos are actually provided with the lecture. So let's talk a little bit about these two videos that you've seen where they have taken technology and actually created an intervention. So by using this database technology, we find that we have many, many, many advantages compared to using paper-based records. And the most obvious one is the information retrieval. From these videos, we found how easy it is to actually retrieve information if it is in a database. We can actually create links between information that we won't normally see. We can query this database asking absolutely obscure pieces of information and we will actually get this information back. For instance, if we go back to our example of trying to find an association between smoking and heart disease, by using a database, we can easily select all patients who had some sort of cardiac disease and smoke. And these records, these patient names it identifies and all the information you want would be easily displayed on your screen in a matter of seconds. It's also important to realize that there are different levels of uses when it comes to electronic medical records. And again, this is because there is no clear-cut definition of what an electronic medical record is. We found that in the first video, there were two electronic medical records described. One was where the clinicians simply scanned pieces of paper into an electronic medical record. And the other where the person actually types out information into the record. And this brings us to the discussion of structured versus unstructured data. Structured data refers to data that is stored in a very meticulously in a way that can be easily processed, like typing out information. Whereas unstructured data is data like scanning a piece of paper and storing that in an electronic medical record. Or maybe saving an audio of an interaction and saving that in an electronic medical record. Structured data is much, much more easily processed. It is much more easily analyzed, which means the more structured data is stored in an electronic medical record, the more useful it becomes. You can analyze the data more. You can ask more questions. You can find out more interactions between the data. Whereas in unstructured, although it is possible to analyze using various types of complex software, it is in fact difficult and takes time. An important part that came out in these lectures is the point of scheduled backups. It is so much easier to make backups using technology. This can all be done automatically. At midnight, at particular days, the entire database can be backed up and even automatically sent to another location where it can be stored. And if we think about the amount of space that is required to store large volumes of information when it comes to technology and when it comes to paper-based files, we realize that the entire room full of paper-based files can be stored in a small, portable USB device. That's no bigger than 10 or 20 centimeters compared to entire rooms full of shelves. The other important things that came out of these two videos was the fact that technology can be designed and used in such a way that you can actually create modules. And these modules make the use and design of electronic medical records much, much simpler. You choose which modules you need to use in your system and you only add those. You reduce the complexity and you in fact save space and you actually make the electronic medical record more efficient. You just use whatever you need. If you remember the first lecture we said the simplest form of technology to solve a problem, just use what you need. There are different modules you just load on the modules that you actually require. Data could be added into the electronic medical record either automatically or semi-automatically. You could, for instance, have a mic attached to your ear and mouth. You could speak while computer software will actually write or record those words in Microsoft Office, for instance. You could have handwriting recognition. You could have other types of voice recognition. You can even just record the voice directly into the electronic medical record. You could have direct entries via keyboard or a light scribe as you've seen in some of these two examples in these two videos. Templates also play an important part when it comes to data entry. We know that clinicians actually get used to writing on a piece of paper. They no longer see what a certain place actually needs. They know exactly where to write what piece of information without looking at the question associated with that space. So when you create templates, electronic templates, you've got to make sure that these templates mimic the paper-based records exactly. And this will actually help in terms of recording data. Data communication is made so much more easily. You can have a medical record being at many places at one time where people can actually read the record, see the record, but not edit the record. Using the internet, using other types of technologies, satellite, 3G cards, etc. You can have people from all over the world looking at a certain record helping a patient. Security. We find that many clinicians have unwarranted fear when it comes to applying technology to medicine. When it comes to using electronic medical records. And this probably is due to many movies that we've seen. Where we find people with a small laptop sitting in a dingy corner somewhere hacking into the CIA or the FBI. And the truth is that this fear is actually very unwarranted. It is very difficult to break into a secure system if you take the right precautions. It will be very difficult to break into an electronic medical record and get this information if you take the right precautions. So what are the right precautions? Well, we have physical and we have software precautions that we can take. Physical precautions are things like ensuring that there's locks on the doors, tying the computer down, having timers so computers can come on at a certain time, go off at a certain time, etc. Software solutions would be things like having strong passwords. So having a password that is made up of both capital and small letters as well as numbers. We can have encryptions, digital signatures. We can have firewalls, antiviruses, etc., etc. All of these things put together will actually ensure that this information is actually very, very secure. When you are thinking about security in terms of electronic medical records, it's always good to actually think about how different security is compared to what is currently being done with paper-based records. How secure are paper-based records really? I mean, records are kept in a filing cabinet. Maybe it's kept in a secure room with a door and a lock. Filing cabinets can be easily broken and the paper record taken out. But if you compare that to actually storing it in a computer, which is also locked in a room, which can be broken into like the paper-based record, but you've got to be an expert to put the computer on to break through any timers that are actually there that have computer on at certain times. You've got to actually break the password. Once you break the password, it's the encryptions. You've got to break the encryptions. If there are digital signatures, you've got to break digital signatures. And in fact, it makes it very, very difficult. Other advantages, clinical decision support systems. These can actually be used to tell you, listen, you are making a mistake. You are recording the wrong data. Maybe there are errors. It can help in medication errors. It will tell you, you are prescribing the wrong medication for this disease. The dosage may be wrong, et cetera, et cetera. Using electronic medical records actually save a lot of money in the long run. Yes, initially, they will be the cost of buying the technology, transferring the data, training, et cetera, et cetera. But in the long run, it saves thousands, thousands of rams or dollars. In fact, in a paper published in 2005 in Health Affairs by Hillstead Richard, he actually mentioned that the USA will save 23 billion US dollars a year if clinicians actually use electronic medical records. It improves quality of care. And that's true because it will ensure that the minimum standard of care is actually done for each patient. It will also ensure that follow-ups are done and thus provide better continuity of care. It also promotes evidence-based medicine. Because there is such a large amount of real data in the electronic medical record, this data can be mined and information knowledge can be taken out from it. So basically, these two slides summarize the advantages and some of the issues with electronic medical records in the two videos that you've seen. So we spoke about the disadvantages of paper-based records. We showed how technology can actually overcome these disadvantages. But is it easy to actually implement technology in clinical environments? Is it easy to actually implement electronic medical records? And the answer is no, it isn't. And there are many, many problems that are associated with electronic medical records being implemented, especially in developing countries. The technology itself poses a problem because many developing countries don't have infrastructure. And this reminds me of a story which I read in a journal. It was an article that actually spoke about the experience of implementing an electronic medical record in a province in South Africa. The hospital managers wanted to implement an electronic medical record in their hospital. So they brought experts from many developed countries to South Africa to the province to actually do this. These experts are actually well respected. They come from a company whose major contribution is actually building innovative technology. So when they came, they brought down big servers with lots of backups with a system that was actually very, very technologically advanced. They installed the system in a small rural hospital in a province of South Africa. Everyone had meetings about workflows. They thought everything was correct. And they were about to turn it on and let the system run. So they had a big ceremony. They turned on the electronic medical record and they waited. It worked the first day, worked the second day, the third day, the fourth day, maybe yes. But a week later, they found that the system suddenly went down. The power suddenly got cut. So these experts thought and they said, you know, we need more technology to actually prevent this. They found the systems were overheating because of the heat in South Africa. And they spent even more money actually creating an environment which will remain cool. At the end of this, they probably spent a couple billion millions in creating the system. It started working again. It worked for a few days and again on Monday at a certain time it went off at 6 o'clock. And this continued happening and eventually they didn't know why. They couldn't understand the reason why the power suddenly just cuts off at 6 o'clock on particular days. Until a nurse who was speaking to one of these experts actually gave them the reason why it's been cut off. And she said that this system is on a grid that a baker is actually using. So the baker down the road puts on his ovens at 6 o'clock in the morning to actually bake that overloads the grid and the power gets cut. Now, if you were not aware of the problems in a developing country, you wouldn't have realized that this could have been one of the reasons why the system is actually failing. Training and expertise. In developing countries there are very few expertise when it comes to this type of technologies. People need to be trained. Many people are in fact computer illiterate that work in technical environments. So not only do you have to train them to actually use a computer, you've got to train them to understand what electronic medical records are, to use electronic medical records and then get benefit from the electronic medical records. The perception we spoke about, the security perception, the unwarranted security perception that many people have in developing countries. But one of the biggest challenges to implementing electronic medical records is in fact the human factor. We find that in a developing country, implementing electronic medical records can actually cause discord among clinical staff. And let me give you an example why. Many senior nurses are actually quite old in developing countries in here in South Africa. These senior nurses are really computer illiterate. They are 50 or 60 years old and they really don't have any information towards using an electronic medical record. But the newer nurses that come out of the universities are actually very computer savvy. So when you introduce an electronic medical record in this environment, you'd find that the younger ones, the younger nurses will take to this so much easier compared to the older, more mature nurses that are in senior positions. So the senior nurses will feel threatened by the fact that they can't use this electronic medical record as easily as the junior nurses. And this causes conflict. That is why change management is such an important part in implementing electronic medical records. Whether it's in a small practice, whether it is in a big hospital. Another important problem that one faces when implementing an electronic medical record is the fact that many people feel that this causes a dehumanization of the relationship between a clinician and a patient. And of course they may be justified in this but my own personal opinion is that this dehumanization does not have to happen. If the right technology is used, it will just be used as a tool, a tool for the doctor to use. It will not be an obstacle in the relationship between doctors and clinicians. For instance, a doctor can use a light scribe pen and write down his interaction with the patient. He can use a small period and type out quickly as he is interacting with the patient. If he can't type quickly, he can recall his audio voice and say that. Or he could have voice recognition software built into his small portable network book or his notebook. So there are technologies that can actually prevent this whole dehumanization of the doctor-patient relationship. So all electronic medical records are beneficial and I am sure by now you know the answer to this. So how do we assess it? There are various ways to actually assess the benefit of an electronic medical record. And we actually have an entire lecture on that. But briefly speaking, there are certain things that you should keep in the back of your mind. The first one is comprehensiveness of the information. Does the electronic medical record store all the information that you require to treat your patient, to ensure standard of care for your patient, to improve continuity of care for your patient? So does the electronic medical record store the information that you require? How long does it store this information? That is the duration and retention of the data. As I said before, some governments have a minimum amount of time the data needs to be stored. Does the electronic medical record do this? What is the degree of the structure of data? The more structured the data, the better it is to analyze, the more value you will get out of the electronic medical record. Can you access the information when you need the information? Does the electronic medical record change the workflow for the sake of changing the workflow? Or does it change it just because the computer scientist or the designer couldn't design anything else? Yes, again, if workflow has to change so that there's more advantages to the patient, then by all means change the workflow. But never have a system that changes your workflow just because it's built that way. This is an important thing to actually analyze. Components of an electronic medical record. What are the components that make up this record? And again, because there is no clear-cut definition of what an electronic medical record is, many people will give many components, their own components, or their own opinion that what should be in an electronic medical record. This five components actually comes from Edward Chocolate. He says that they must be an integrated view of the patient data. So you should be able to see the patient data from different aspects, all the different aspects must be integrated together. There must be clinical decision support systems. There has to be a clinical audit entry system. You must have access to knowledge bases, to knowledge resources, and there must be integrated communication. There are of course other issues with computer-based patient records or electronic medical records. And we'll speak about one of them now and that's data entry. Data entry is an important problem with issues that you must keep in mind when you are implementing your electronic medical record. How are you going to capture the data? Are you going to capture the data in real time or batch? Are you going to have the doctor type out the interaction while he's seeing the patient? Or is the doctor going to write in a piece of paper and have the data capturer capture the data? And of course there is no right answer. It all depends on the environment that you are in. And this again reminds me of another problem that we faced when we tried to implement an electronic medical record here in South Africa. Data captures. There has to be some way in which you can encourage data captures to actually capture the data correctly. When we first install an electronic medical record because of time and because of computer literacy, we thought it would be best for doctors to actually write in a piece of paper that's given to a data capture and the data capture captures the data. So this happened for a few while. We started trying to analyze the data and we found that there were lots of mistakes. And we realized that those mistakes are mistakes from the data captures point of view. So we thought of ways in which we can try and encourage data captures to actually capture the data better. Just being out of university and arrogant as we were, we thought we could actually scold them. So we called a meeting. We threatened them about telling them that we know exactly who's entering the wrong data and that they shouldn't do it, we can trace back all mistakes to a certain person and we sort of tried to throw our weight around. This helped but it helped for like a few weeks. And we found that the quality of data got worse. Then we thought threatening them didn't help. Maybe we could actually bribe them. So we told them, listen, from our own pockets, of course our boss didn't know this, we will give them airtime if they could actually record the data without any mistakes. And of course this worked for one week, it worked for two weeks and again we found that the quality of the data recording was actually getting worse and worse. And then we thought, no we can't threaten them, that didn't work. We can't bribe them, that wasn't working. So how can we encourage them to actually record the data correctly? And then we realized something. Every single human being wants to feel important. They want to know that they're making a difference that is fundamental to actually being a human. And that was the approach that we took. We sat them down, we spoke to them, we told them how important their job is. We told them, without them, the system would just not work. That if they record wrong data, they in fact can actually kill patients. And this in fact made the difference. I feel that the few months that the system was actually being piloted, all of the data was excellent quality. There were very, very, very few mistakes and those were actually mistakes that anyone could have actually made. So this is a very interesting story to actually tell you guys. If you think about data capturing or capturing data for electronic medical, record how to prevent errors and data input errors. Another issue that we should actually think about is the display. A display like the one shown on the left is very, very bad. Displays should be assessed based on how busy they are, the colors that they use, based on how intuitive it is. I would assess the display on the right much, much better because it's less cluttered. It's more intuitive. It's doing charting instead of having raw data, et cetera, et cetera. So those were in fact how you would assess or what you should think about when you're trying to assess the benefit of electronic medical records. Thank you very much for your time.