In Denmark, health data is used to reach a national consensus, however there is a lot to improve. What can be expected in the near future and in longer term perspectives? These are the topics covered in the second part of the interview (the first part here) by Leo Petersen-Khmelnitski, www.cifs.health Editor, with Jens Winther Jensen, CEO of the Danish Clinical Data Quality Programme, an agency that manages more than 85 databases with health related, mostly clinical data.
Leo: You have referred to having so many databases with such a rich and long history, what about their future?
Jens: I foresee that in the near future we will setup specific databases in specific clinical areas. Data here is a tool, it’s not the product. The product is better quality and better research, and that starts with the clinicians and patients. The core basis of what we do here in the Danish Clinical Quality Programme is that we bring a group of clinicians and patients together, help them to build a national consensus on what they regard as good quality. That consensus building is essential, because in modern medicine there is so much evidence around, and you need to have a direction for the country to go. Again, size helps Denmark. We bring them together, and they have this discussion. And afterwards, they tell us, we would like to look into this area or that area, show us the data in those areas. And then at times it is necessary to collect different thoughts on new data. So that engine of building clinical development in terms of research and quality, we still need to build. There will be new pandemics, new developments, clinical development is something that is ever changing. And we will never be in a situation, where we will have enough data to describe even the near future.
Leo: I see. What’s on the way?
Jens: There is a tension there with IT systems because the IT systems, at least the way I see them, they do not like this agile thing, that you have to change fast because clinical developments are fast. It only takes a cardiologist going to a conference in San Diego – at the moment they don’t do it, but hopefully they will soon – and then they change guidelines to treating acute myocardial infarction, because there is new evidence, and they import it very fast in their evidence base. IT systems have to support that. That engine needs to be more effective.
When it comes to the clinicians, we are still not in the situation where data is easily available to clinicians. Data around the quality of your practice or works of your department is not readily available to clinicians elsewhere
Leo: Why not? Is it due to security and privacy concerns?
Jens: No, it’s due to the systems that support them. And it actually goes both ways. The BI systems that are supposed to support clinicians, are not developed in an easily usable way. It is also a question of whether or not it is requested by clinicians. And there is also a need for development of stimulus, when clinicians are asking for this data, in order for data to be used in a better way. One thing that I hope we will see more of, is decision support. There are still some barriers for decision support, there is some law issues in this area.
Leo: Do you mean that regulations are missing?
Jens: Yes. And important questions are yet to be answered, for example, are electronic health records really supporting this development in a good way? There is something to be done there as well. And another thing that I would really like to see in the near future, and this has been a task for many years, is the ability to build systems on a large scale. From the clinical point of view and especially from the patient’s point of view, the fact that we have this old, priorly collected data may give both patents and clinicians an idea, if you choose this treatment, how will that effect you in your treatment journey.
Leo: But is this data structured enough?
Jens: Until we start building on a really large scale, I would argue that we have enough structured data to advance in quite a few areas.
Leo: The issue of unstructured data, is it a non-issue in Denmark?
Jens: I would argue that compared to other countries it is not as big an issue. Of course, we can still find a number of areas, where we would want it to be more structured than it is today, from a registry point of view.
Leo: in which areas, for example?
Jens: Some areas are more structured than others. There is difference between an Epic journal in east Denmark and a Columna journal in west Denmark, as Columna in general uses less structurization then the Epic journal. On the national scale, if you go into surgery, cardiology, obstetrics and gynecology, some of the cancer areas, these are areas with structured data, and you will find less in psychiatry, geriatrics, and these areas. We should not structure data because we just like to structure. We need to implement inside-out building. What we have learnt with acute coronary syndrome databases? We have learnt in that short journey, that when we integrated them into Columna and into Epic, clinicians started asking us to structure data. What we have given them is an opportunity to structure data inside the electronic health record, if they do it, we can then take the data out and put it into a database. In this way, the structure of the data becomes a decision support tool as well, because when they input this data, it helps them to make right decisions about a patient. Now physicians are asking for it, which is a new thing. In the past when you came with all this data to a clinician, in many cases he or she did not like it very much. However, when you build it from that inside out point, I believe that in many instances you will have a solid support by clinicians.
Leo: Do you think that future health data infrastructure will be based on this inside out approach?
Jens: At least that is the way we start to work now in Denmark. I do not foresee a huge national structuralisation project with electronic health records. Some players will have to get together, we work with two electronic healthcare records to build the level of structuralisation needed to support the level of clinical questions that are asked, in databases and other areas. But it is not only for clinical databases, it can be also for clinical research purposes. You need data for clinical research. At the moment we build it mostly in Red Cap environments and alike. It bears a huge registration burden. Things can be a lot more effective, if electronic health records will also serve to support clinical development, whereas now they are built to support documentation in clinical setting.
Leo: Some predict that future digital infrastructure in healthcare will be decentralised, built on decentralised ledgers, on blockchain, where clinicians will be able to verify data themselves, without any central agency. What is your view to this?
Jens: I am not an IT expert, and perhaps there is a lot for me to learn here, but from my current standpoint, if an IT system at one point measures blood pressure, and another IT system at another place measures blood pressure, if I am to use these two datapoints because it is of clinical value, I need to have some kind of verified information around that it is actually the same blood pressure. That it is measured at the same arm, same time, at the same patient. At the moment, I cannot see how AI or blockchain can help me give this information on an easier level compared to us coming together and giving each other this information via our IT systems. But maybe my imagination is limited.
Leo: Future uses of health data open up for some exciting opportunities, and that is why in conclusion I would like to ask you, what kind of usage do you envisage for health data, that do not exist today? I do not expect an answer from the point of IT, I expect the answer from the point of functionality. What will clinicians need, and what is it they will be able to do?
Jens: I think that in many areas clinicians and patients will ask for much more granulated data. Personally, I am encouraged by health data that I get from my portable devices, and I think that in terms of developing yourself as a clinician, in terms of ‘how do I become a better doctor’, ‘how do I become a better nurse’, I think we will move into those directions as well, and I think the patients will ask for the same.
What is the quality of the healthcare that I am giving? How can I make better use of data that are stored around me in these big systems, to develop my navigation capacity, if I am a patient? What I say here is not a revolution. But it brings us to a very concrete vision for a system that is capable to support a patient in a meaningful way. Also, let’s pose a general question: why can I not have a copy of my own health data? I would like to have this copy. I will use it together with another data to make my living better. We are unable to support that at the moment, but I think the demand for it will grow.
Leo: When you think this functionality will be implemented in Denmark? In what kind of time span? Are we talking ten years or are we talking two years?
Jens: Probably more ten years than two years. Some of these issues require technological development, but as I said there is a huge amount of data that is stored under my CPR (Danish national ID) in the current databases. Giving me access to them is not a technical issue. That can be done. It comes down to political will.