On Health(y) Data. Q&A with Jan Vekemans
Fragmented data is a real problem, a potential killer without want to be too dramatic. What I mean is that when a patient and by proxy a doctor don’t have a full picture of the patient, the diagnosis might be incorrect, incomplete or just outright wrong.
Q1. What are the consequences of fragmented data in healthcare?
Fragmented data is a real problem, a potential killer without want to be too dramatic. What I mean is that when a patient and by proxy a doctor don’t have a full picture of the patient, the diagnosis might be incorrect, incomplete or just outright wrong.
It might be a populistic view, but this is what the series on TV, “DR HOUSE” is all about. When doctors must base their opinion or diagnosis they need an as complete picture as possible because even a seemingly innocent detail might be important.
I have a example from last week when my son was complaining from pain in the abdomen and my assumption was that he had a stomach ache instead of Appendicitis due to the pain not being located on the right lower left side of the belly…
Having spent a lot of time in finding key pieces of my own data I saw how scattered and unavailable data is when a patient needs it. Fragmenting data also means necessity of duplicating tests and interpretations. So cost goes up and the chance of older or incomplete information heightens the risk of mistake. So clearly it is important to have a maximum of information that is available, in a timelined, complete and trusted way.
Q2. What are your suggestions on how to avoid fragmented data in healthcare?
The first requirement is a timeline tool which has all the treatments and incidences of a patient in a simple overview. This allows to have a view of what a patient has been treated for, when and where. Once we have this view we can expand beyond that information and do, for example, predictions. But before we go down that route there are some important issues to address. If you look at the current data available in the current systems, there is agreement that that data is not fully reliable! It’s not complete and sometimes misplaced. Combining data is the moment where this misaligned data will amount to serious issues while if att the same time the patient is involved in the process some of these errors will be caught by the patient/doctor/healthcare worker. Due to the nature of medical data a lot of it can be standardized (via norms such as ICD 10/11 or Snomed coding) but due to constraints like time pressure and urgency, the data might ends up in the “wrong” place. Those are issues that need to be addressed before we move into adding data from several sources.
Q3. Are there ways to share data quickly and safely?
Taking the previous point into consideration where we need to clean up the data first, it is clear that to allow a fast reliable integration of the data we need to start by finding and compiling all data on the patient. Once we have a perfect view on all the patient’s medical records it becomes imperative to pool it (either physically bringing it together or doing it in a metaphysical way) and the best solution is via a UNIFIED CARE RECORD. This is a system that allows an overview of all data concerning 1 patient. It is a globalized, structured collection of everything that has ever happened to this person. It gives a view of what GP, Hospital and any other healthcare professional have ever done for/to the patient but it could also easily be enhanced with lifestyle information, if and when required.
Q4. What are the benefits of a centralized access to all patient medical data? Is it feasible in practice? Is GDPR limiting this?
In an ideal world this is what we want. Think about it, wherever you go, whomever you go and see, physio, GP, Hospital, … Whatever information they need to help you is available at a fingertip. Perfect, not? Well this is what would happen in a perfect world, in our imperfect world it comes with a number of caveats and let’s address those so we can get to what works and what not. Building a system that has omnipresent access to ALL data means we need to secure the system, and to take privacy and sensitivity into consideration. Making sure data can’t be misused/abused. Let’s take the example of an insurance company whio all of the sudden tripled (or worse cancelled) your insurance because a ‘lump” is found during your latest examination…
So let’s get back to the basics and take it from there. For the patient, remember this is who we do it for, the Unified Care Record can be a major advantage as now all information can be seen/consulted/used from one place. BUT now we need to make sure the patient has access to this data, so we are looking at authentication. It has to be strong for data privacy but also easy to use as it need to be accessed even by technologically challenged people. This also means Authorization, meaning people need to be able to allow someone else to do things in their stead or/and allow healthcare and perhaps even family members to have access to (some of) the data. GDPR gives the patient the right to know what data is available about him/her but no provision is made to make this data automatically available. A centralized access would take away that hurdle because the information would be available… and if we look at the hubs in Belgium than a good start has been taken to do this BUT it is too little and too limited because even through these systems not all information is available, and it is extremely difficult to make the data available.
And this is where the next step comes in… making it easy to search and sort through the data. Data Mining for dummies as you wish because having the information is nice, being able to have people action upon it is much better. Otherwise, a data lake will become a data swamp. Let’s also not forget that once data is available in a trusted, structured and over seeable way it will also be easier to provide the data for secondary use by pharma and Life Sciences. Can it be done? YES! Will it be easy, NO! The best way to start would be to provide a network where every patient has access to 1 “mailbox” and a timeline. For every document/result/diagnosis there would be 1 entry. This would not be yet the data residing in 1 location but it would link to every item so the data would continue to reside where it currently resides and it could easily accessible from the platform. Metadata concerning the linked information would be added as that is critical. The SOLE question that needs to be addressed to make such a project possible is WHO WILL PROVIDE THE SERVICE? Government? Private company … ???
Q5. You work for InterSystems. What is the contribution that InterSystems is offering in this domain?
IRIS (for Health) is our Interoperability platform that has several tools that allow access to data in either structured or unstructured form. The platform also allows people to reach beyond the boundary of the own organization so building structures on local, regional, national, and even international level is possible with them. Built onto this set of tools from InterSystems we created Unified Care Record, the HealthShare answer to connect the data from multiple Electronic Medical Record Systems (EMR) providing the doctor/patient a unified view on the data available in different systems. As this is already a way to view what is happening in multiple EMRs it is a good basis for a patient-oriented system allowing an overview of all patient’s data.
Q6. Specifically to the regions you cover, can you give us some practical examples?
Belgium has voted recently a law that forces hospitals to group together in networks. These networks are made up of hospitals that voluntarily work together. BUT these hospitals have each, many years ago, chosen a EMR software to work with. Today they are forced to cooperate but their software products are incapable to provide transparent access to the data in each of these EMR softwares. Let’s take a concrete example, in Wallonia, one of the networks consists in 8 hospitals and has 5 different EMR softwares. Without a UCR the hospitals have found no better solution that to unify and standardize on 1 EMR software BUT this imposes serious limitations.
1. Retraining personnel to be able to work with a different platform that is also configured differently (risk of faulty entry/data loss/ … ). This can lead to dangerous situation
2. Financial repercussions, current contracts are not yet at the end and this means that they will have to honor indemnity payments to the vendors
3. Even if you bring them all on the same platform, this ENORMOUS change will take many years to make it happen.
4. When we look at EMR software and how it is used/configured… in every hospital this is done differently. And now we are talking about doing exactly that PLUS doing it cross platforms as we not just want those with the same EMR to converge but also the other vendor’s EMRs…
Q7. Putting patients at the heart of healthcare is important. What are the main challenges when putting this into practice?
When we put the patient at the heart of the system we honor the patient as what he/she is; THE PATIENT IS THE DATA. Naturally no patient will know or even use all the information that is available about his/her person. Getting all that information unified and presented in a conform, trusted and safe way is asking involvement from the Healthcare organisations, government, the IT industry, doctors and other healthcare workers but also of the patient. Provisions need to be made to make it possible to cater for incapacitated patients, people who are illiterate or otherwise incapable of managing things themselves. Privacy, such as governed by GDPR and local law, are often cryptic and incomprehensible so we should start by making it easy, simple and reliable. Once access is insured we need to provide federation capability and data segmentation as to not have everything available to everybody. One needs to consider also offline capabilities as not everywhere is a data connection available or even wanted. So the real challenge is to put confidential and valuable data in the hands of someone we don’t know but we can rely on to keep it safe and do this in a safe, secure and user-friendly way.
Q8. You are one of the founder of the 1patient1record4Belgium movement. What is it?
#1patient1record4Belgium started as a selfish project as when I had cancer I could hardly get access to my data for a second opinion. Instead of giving me the data or being able to share it with others I had to find my way through red tape and issues all over. And I know what and where to ask as I had been in this business for years… So I asked myself, is this just me or do also other patients have the same issues and frustrations that I have…?
I soon found out I wasn’t alone and so I wanted to do something about it.
My mentor always told me not to come with a problem but come with solutions and that is what I did. I started talking to people to understand how ”big” the problem was and how we could solve it. That’s how 1patient1record4Belgium was born. By discussing with all actors involved a plan began to ripen and when others joined in strong ideas came to be. Strong ideas that became plans for tools and software to achieve that what had been missing. Instead of a yellow lunchbox in the fridge we would provide access to all data, digitally, safe and smart. What we want to achieve is to give access to all data belonging to a patient. One access to Medical Data, Digital Wellbeing data and all surrounding other data through one simple interface and controlled by the patient. How to make sure that data is working for the patient and not rot in silos, however good meaning the reason for having them there was!
This is where I started to think Garbage IN garbage OUT but then that doesn’t cover it all! It would be garbage in bin, bin in truck and truck to the garbage dump! What needs to happen before one should consider bringing things together is cleaning the own data and here the patient can help bringing me to the next topic…
Once a problem understood one can map it and start thinking about how to cure it. A hackathon, three intense round table conferences and help from many patient and other alike. We are turning ideas and insights into tangeable products and solutions for the world to test and to help the patient. What we all want it for this to evolve to a return to the old days. Long ago a doctor was only paid when a patient was healthy and that is the incentive we need to give back to medicine because then we will actually have a healthy and/or healing patient. Data is power, let’s wield it!
………………………………………….
Jan Vekemans, InterSystems Country Sales Manager Belgium & Luxemburg
Sponsored by InterSystems.