This article is compiled based on the speech delivered by the chief medical information officer of Kaiser Medical Group at the "Medical Big Data & Smart Healthcare" forum of Songhe Capital Boutique Forum. Minor cuts.
First of all, I am honored to be here today to communicate with you. I am also very interested in the interactive Q & A just now. The reason is that we used to have the exact same debate, and this debate is now happening in various countries around the world. Today, I want to mention some of our understanding and solutions to these problems. We have spent many years improving these solutions. We cannot say that we have solved these problems, but we hope to share our ideas with everyone.
I come from a medical group that integrates insurance and medical services. In this group, all our data and people in all roles come together to work. We have 20,000 doctors, 50,000 nurses, and more than 10 million patients, as well as 40 hospitals and more than 600 outpatient centers. Here we put all the data. So many hospitals and outpatient centers use the same system, and integrate the data of various departments and dimensions. We do data analysis and data mining based on this data set. The purpose is to improve our clinical care as quickly as possible.
I also very much hope that all the experts and leaders present here can ask questions, continue the debate just now, and ask questions at any time.
I am a marine biologist and evolutionary biologist, and as a doctor, I answer the questions that Vice Chairman Wang started with. I have practiced internal medicine, emergency department, preventive medicine, trauma medicine, and high-pressure medicine. Developed the very early electronic medical record system in 1984. In 1996, he was the initiator and founder of the international standard for Clinical Document Architecture (CDA) and a member of the Board of Directors of SNOMED International. In 2008, he led the implementation of the largest non-governmental, non-military electronic medical record system in the United States. Since 1984, after so many years of hard work, there is still a certain distance from our goal and there is still a lot of work to be done. It is very important to have a trustworthy partnership in integrated medical care.
Mr. Caesar founded Caesar Insurance Company, and Dr. Garfield founded the Doctors Group. In the process of cooperation between insurance companies and doctors groups, just like marriage, everyone will work hard for a goal, but everyone will be different, so it is normal that there will be conflicts and differences from time to time. This mutual trust between them is very, very important. This is one of my favorite pictures. This is a diagram drawn by Dr. Garfield 50 years ago. What he thinks of our medical services. It's been 50 years since we just put the systems in this diagram together, so we've just realized Dr. Garfield's vision 50 years ago. In any cooperative relationship, it is very important for us to clearly distinguish the role and responsibilities of each party.
The medical service team is responsible for the continuous improvement of our clinical quality and clinical operations, while our insurance team is responsible for managing financial risks and financial support for operations. Some things can be done by the two teams separately, but some things must be done by us together. Together, our medical and insurance teams are responsible for improving the overall medical experience and financial health. Just like I see patients in the emergency department, if I were a nephrologist, I couldn't tell the cardiologist how to see the patient's heart disease, and similarly, the cardiologist couldn't tell me how to see the patient's kidney disease. Mutual trust and communication are the key to success.
In the process of continuous improvement, there are many aspects we need to improve. On the one hand, clinical results, patient satisfaction, cost results and functional results, that is, how quickly I can go back to work, and doctor satisfaction., our own reputation, brand effect. We are currently transitioning from evidence-based medicine to the use of medical-based evidence. This work started at Stanford Medical School, and their platform allows doctors to ask this question: Can I quickly find in my database the 5-10 patients who are most consistent with my patient's condition? If it is a very rare combination of conditions, I may not find any medical literature that can tell me how to best treat this patient. Then I have to look in my database to find out which treatment plan works best for these 5 - 10 patients. I may use this to guide my treatment of patients.
Dr. Garfield had it 50 years ago and I wanted to put all the data together.
This is what we did about five years ago, putting all the systems together for our 20,000 doctors and every patient they treat. Under this system, all information and data are interconnected and put together. With such a system, we can see how each patient makes his outpatient appointment, whether he is hospitalized or transferred, what fees he needs from insurance, and what services he pays for under insurance. Our doctors are in this system when they see outpatients, and they are also in this system when they manage inpatients. Our insurance personnel also see the same system when managing insurance, and even our staff in charge of company finance see this kind of data, as well as regular financial reports generated from this data set.
Here, all our above data will be collected into the following data warehouse. This data warehouse is divided into two parts, one part supports real-time queries and reports. The other part is to do offline. We will do data mining and generate new knowledge. On top of this system, we have a portal on the Internet that connects both doctors and patients. If a doctor writes any doctor's order or leaves any information for a patient, this information will be immediately stored in the electronic medical record, and the patient will be able to immediately see the data when he goes to the portal website to check it. What often happened before we had this system was that a doctor or clinical staff felt that he had entered guidance or results into it, but the actual work was done, because the system was not integrated, and the work was handed over to another person. He didn't know that the problem had indeed been solved. Now that we integrate all systems, we can clearly see the entire process of a problem from the beginning to the end. The following data warehouse we use, a data warehouse is used to provide guidance to our clinical staff on how their clinical work should be completed every day. The second part of the data warehouse is where we do data mining and data analysis. For example, if I give this patient a drug that lowers blood lipids, when his cholesterol drops by a certain percentage, how many times can he go to the outpatient clinic or how much time he spends in hospital?
This is the question our second data warehouse can answer. What this big data and artificial intelligence can bring us is that we can now let data show us these assumptions. Before we had big data and artificial intelligence, we might need to have an assumption ourselves first, and then Go to the data to find answers. There is a danger in the hypothesis-generation process brought about by big data and artificial intelligence, that is, the correlation between this data does not indicate the true causal relationship. So now there is a new field called data review, which is equivalent to having a data review process to help you verify whether the associations you find are true or for some reason are actually wrong.
We now have an average of 30,000 pieces of safety information every day that doctors and patients communicate before. At the same time, the portal also provides patients with millions of queries on test data and other data on their own medical records. If a patient needs a mammogram or a colonoscopy, the system will remind him in time. Thanks to our reminder system, the rate of early screening has been greatly improved, and our mortality rates from breast cancer and bowel cancer have been greatly reduced. We had an actual patient who came to our optical center to get glasses, and her ophthalmologist saw that this patient should have a breast examination. Because of this reminder, the patient had a breast examination the next day and found a lump. Biopsy was done that day and surgery was performed within a week. He had a good prognosis for breast cancer.
There will be such reports and analysis functions in our electronic medical record system, and more importantly, decision support. This is also what Dr. Gao and our team are helping to do. Putting this knowledge and reminders into the system can reduce the burden on clinicians, so that they don't have to worry about what I, a patient, should do. Although most of this system is developed by manufacturers, as part of our team, we need to make tens of thousands of decisions. These decisions will determine how we set up these systems and how to make these systems correspond to our own workflow, including doctors 'and nurses', to better serve us. After this process begins, we need to continue to make thousands of decisions to keep this system evolving and evolving to help us do better.
As our medical care improves more and more, we will think of a question: Where should we provide the best medical services? We should wait for the diabetic patient to come and treat the diabetic patient, and if the patient has a heart attack, we should give him first aid. Should we treat patients this way or should we need to move forward and help patients change their diet and their lifestyle? This way we may be able to prevent these patients from coming to our emergency care centers or already having various complications. This is very expensive for both society and the patients themselves. How we usually eat, how we are in our usual mood, how much activity we have and how much exercise we have. All of this has a very, very big effect on our own health and on our medical care.
Our five principles for happiness are: first, first, you must feel useful to this society, and your skills are needed by this society; second, you must feel that you can help this society; third, you will feel that you are grateful and grateful by this society; fourth, you must be grateful for everything that life brings you; fifth, the whole environment should be a process of continuous learning and constant innovation.
Vint Cerf is one of the inventors who invented Intel, and I have the honor to work with him on several global projects. One of the projects is equivalent to investing in helping children have healthy living habits. Economists at the University of Chicago have long proven that if you invest in children's health habits, the rate of return is much higher than that if you help this person or even treat his disease when he is an adult. Because we have comprehensive data, we can see what is beneficial to society as a whole, including aspects such as life, health or medical care. We will not be like most institutions, where I may just do kidney treatment, and I can only do heart treatment. Thanks to this kind of data and a comprehensive way of looking at problems, we can make more effective investments in this community and society, and more importantly, we can also evaluate whether our investments are effective.
I don't know if you know about the "blue zone", which means that people in an area can live a long life, be very happy, and be very happy. Five such regions were studied around the world. If we look at the data from these five blue zones, this data tells us about their diet, exercise, sleep and social interactions. These are crucial to their health and happiness. So in our organization, we invest US$1 billion every year in the community to help farmers bring fresh agricultural and sideline products to our community and provide them to community residents, as well as carry out various activities in the community to help children exercise to help them eat healthier.
Those of us in charge of medical information systems here may have the same experience. If there is a problem with our information system, we may need to find the root cause layer by layer. If we find the root cause, we can solve the problem, but if we don't do enough in-depth investigation and just go through one or two layers and just solve the current problem, then the problem may reappear. We did this kind of survey and saw how many diabetic patients, how many myocardial infarction patients could actually be avoided, and how many patients could avoid this clinical outcome through diet or healthy lifestyle. At the same time, research has also proved that our Social networks, before we had various Internet tools such as Weibo, our Social networks were already very important to our health. When we look at communities, this data tells us which communities we might need to help quit smoking and which communities we need to help reduce alcohol consumption. However, existing and some social applications under development are actually very beneficial tools in changing the behavior of friends and family.
Nowadays, some people are also developing smart assistants. It may be that the image of an old man's grandson appears on his mobile phone to remind him to quit smoking. This is good for your health and for you to live a longer life for a few years. It could also be the image of the patient's daughter to remind the mother to do more exercise, which would be better for him. Personally, the most exciting or far-reaching technology in the future is virtual reality technology. There's a laboratory at Stanford that does what they call remote physical transmission. Now with such equipment, when you are doing something in one place, there is an image of yourself in another distant place doing the same thing you are doing in this place, but you are observing it as a third party. It's very difficult to change people's behavior. It's difficult to see what will happen if you change your behavior. It's equally difficult to be able to take a step back and see what it will be like when you do it, and this technology is to be able to do both. Another technology developed by this laboratory is that allows a person to speak different languages and have different looks in this system, which can help you have more sympathy for people from different cultures and different regions. Putting this kind of smart assistant includes putting the image of your loved ones in it, or even putting the image of yourself when you were young and eighteen years old, hoping that you could go back young. Integrating this technology with smart assistant technology and virtual reality technology based on remote physical transmission is a far-reaching tool.
These words may sound crazy coming from a data scientist like me, but if you focus on the root causes of our medical health as I do, it all becomes natural. Our previous speaker mentioned that we need to integrate all kinds of data, including our medical data, wearable data, and out-of-hospital data. This technology will make this data useful.
Finally, I would like to talk about overall health. Now more and more evidence leads us to focus on some medical treatment methods that have been available since ancient times. After I return to the United States, I will attend a conference, which is a conference on overall health. The topic I'm going to talk about is actually about thinking, doing exercise, and making decisions that not only change our health, but also change our DNA.
I talk about this very interesting topic, and these topics are now scientifically based. This change to DNA does not change its coding order, but changes the methylation in this coding order. These methylation determines whether some of our genes will increase or decrease expression. It may be that good genes will increase expression, or it may be that bad genes will increase expression. And these are what we actually observed in our research. When we look at a person over the course of a day, the expression level of genes changes very much over time during the day. The expression level of good genes and the expression level of bad genes actually have a very big impact on our health. The expression of these genes has a very profound impact on whether we are doing cancer treatment or some other medical treatments. This study of genomics and gene expression levels actually tells us the same fact, that if we have a healthy diet, exercise and living habits, it will help us live better and longer lives.
The following is question and answer time. I hope everyone can raise the most difficult questions. Whether it is the areas we discussed earlier that everyone felt were insufficient or even if it is about the United States or Europe, you can raise any questions. Thank you!
Question: I want to ask two questions. The first question is that patient-centered medical homes, abbreviated as PCMH, are being built in the United States. How did this model develop at Caesar? The second question is John Maddison's visit to China. I hope to also give us some suggestions on how to carry on as a doctor group in China.
John Maddison: Patient-centered medical families are consistent with the modern and responsible medical institutions advocated by Obamacare. In fact, Caesar has been this model for decades since its establishment, so Caesar does not need to change anything. Caesar has always had a patient-centered medical and family mechanism. For Caesar, it is enough to continue this work. There are two key points. The first point is that we must be financially stable and sustainable. There is an example in the United States. There are many responsible medical institutions ACOs. They have great ideals, but they have financial problems and fail halfway through. After there are no financial problems, another very critical thing is to always put patients first. Whether it is business, clinical and other aspects, patients should always be put first in all decisions. For example, when setting up our information system, our doctors may have an opinion, our pharmacists may have an opinion, our nurses may have an opinion, and our cardiologists and kidney experts may have different opinions. What can we have a mechanism to integrate so many of our different opinions and make it executable? It is for everyone to recognize this principle. We must put patients first. With this principle, Our differences can be better resolved.
Question: The data warehouse John Maddison just talked about, how often does the data warehouse update? Are there anything else in it besides this?
John Maddison: I said two data warehouses, but I actually missed one. We have three. The first one is a real-time data warehouse, and all information goes through this real-time data warehouse. This data warehouse is where we support daily medical services, all services. This data is constantly updated every second, as we operate. Each database will be optimized for a certain application or certain applications. We have a second data warehouse that is updated every 24 hours.
You asked a very good question. In fact, the functions provided by these three databases are different. Usually people use all their questions on the database they are familiar with. But in fact, your problem should be related to the structure of the database, how often the data in the database is updated, and the technology it is good at. The third data warehouse is a data warehouse that uses big data technology to do offline big data analysis.
Question: Insurance in the United States has commercial insurance and government-provided CMS insurance. What is this insurance like?
John Maddison: There are many advanced experiences in the United States that may be useful for other places to learn from and refer to, but the American medical finance model or payment model may not be one of them. But I still want to answer your question. There are three payment methods for the U.S. government's medical insurance. The first payment is the traditional payment based on service items, and you will be paid for each service provided for each patient; the second is the newly introduced model according to the U.S. medical reform just mentioned. Pay-by-risk model; the third is an early model based on value-based payment or quality-based payment. Caesar is the third model of medical insurance in the United States. For a medical institution or a service institution, you can only choose one model and do it with CMS. Ideally, our market or the entire country might contain 10 to 20 value-based, quality-based institutions, and there is enough competition between projects so that the market can feel which is the best value-based medical institution. Based on my guess, your doctor group will be one of the 20 medical institutions that can win the competition in the China medical market.
