IC ERASMUS MC ACCELERATED TOWARDS DATA-DRIVEN CARE BY THE CORONA CRISIS

Klik hier voor de Nederlandse versie.

‘We are one of the first IC’s to implement real-time data analytics in practice’

The corona crisis resulted in new insights on all kinds of fronts. Erasmus MC’s intensive care unit experienced that real-time data is essential for fast learning. And that this data could potentially save lives. That’s why the innovative hospital is now focussing on data-driven healthcare even more. ‘We can use data to provide better quality care with fewer staff and organise this care more efficiently,’ according to internist-intensivist Michel van Genderen.

Erasmus MC’s intensive care unit, just like in the rest of the Netherlands, was filled with corona patients for the first time in the spring of 2020. The IC team led by prof.dr. Diederik Gommers was under immense pressure. ‘We didn’t understand the clinical picture very well to start with,’ internist-intensivist Dr. Michel van Genderen from the intensive care unit for adults informs us. ‘We started seeing more and more patients. We struggled with staff shortages too. But one thing we did have was data. Our ICU patients are monitored in all kinds of different ways. We were looking for a way to gain an efficient insight and overview with that data. On the ICU as a whole and where the individual patients were concerned. And subsequently the clinical picture too.’

DASHBOARD WITHIN SIX WEEKS 

The crisis situation meant we needed to act quickly. The regional ethics committee for medical research waived the need for written informed consent to patients, allowing Erasmus MC to quickly get started. A dashboard was built in just six weeks, in which all kinds of data sources came together. Patient demographics (such as gender, age and body mass index) were combined with data regarding treatment (such as prone position and steroids (yes/no)), complications and outcome (such as ventilator-free days and ICU death). ‘That gave us the opportunity to check the state of affairs at literally any time of day or night,’ according to Van Genderen. ‘This streaming data allowed us to make faster decisions about treatment regimens, without having to wait for findings from national and international database studies.’

A detailed description of setting up the data dashboard during the corona crisis was published in the renowned BMJ Health & Care Informatics in August, a peer-reviewed open-access medical journal on health informatics. 

DATA SAVES LIVES

Data saves lives, literally. ‘We were one of the first hospitals to realise that blood clots in the lungs were causing respiratory problems,’ says Van Genderen. ‘We were able to respond to that. The data therefore enabled us to offer better quality care. We were also able to better predict which patients could leave the ICU within a few days at a later stage during the crisis. This allowed us to estimate the care pressure in the ICU more effectively. You can undoubtedly imagine that we will also be able to plan our staff more efficiently in the future with this data, something which will become very important, given the extreme work pressure. It will therefore work in more ways than one.’

LESS STAFF, BETTER CARE 

‘The corona crisis has enabled us to gain an even better insight into clinical pictures and processes,’ Van Genderen continues. ‘The crisis meant we were able to accelerate the building of a streaming data foundation. Now we want to push ahead.’ Future healthcare is largely based on data, that’s Erasmus MC’s conviction. Van Genderen: ‘Demand for healthcare is increasing, while the staff shortage is by no means easily solved. We can use the data to provide better quality healthcare with fewer staff and organise this care more efficiently too. We can monitor the human dimension more effectively, as we’re providing the right care at the right time and can make more efficient predictions thanks to algorithms. This means, for example, that we can discharge patients from hospital at an earlier stage, as we can monitor them remotely. This makes a huge difference where capacity is concerned, while it’s often also much nicer for the patient to be able to go home earlier. Plus this will allow us to spend more time with those patients who really need it. The final call is always up to a doctor, but data can help with his decision-making.’

SEARCHING FOR THE HOLY GRAIL

Erasmus MC is therefore working on a method for structural and standard use of streaming data. First in the intensive care unit, but later also in other departments. ‘That’s the holy grail, but it will take a number of years,’ Van Genderen says. ‘Because there are so many aspects to it. A great deal is possible from a technical perspective, but how do we view the use of data from an ethical and legal point of view? What if a patient in a coma enters the ICU and can’t give permission? The human aspect is very important to us. The data must also be presented in such a way that doctors and nurses can effectively interpret this. Doctors and nurses must be able to trust data just like a pilot can trust all data in his cockpit. How far do we want to, and can we, go as a hospital in the field of data? These are all things we’ll include in our innovation project, in which we’re also working together with the computer science programme at TU Delft.’

FROM BYTES TO BED

Van Genderen published a review about the use of artificial intelligence and data in intensive care units last summer, in collaboration with Erasmus MC colleagues Prof.dr. Diederik Gommers, dr. Jasper van Bommel and dr. Eva Klijn. ‘This showed that 95 percent of those initiatives don’t make it to the clinic,’ according to the internist-intensivist. ‘That clearly shows how difficult it is to get, as we call it, from bytes to bed. That’s why we take the time to conduct thorough research into the use of data and to work in an extremely careful manner. We’re dealing with human lives. And to improve the quality of life for people who survive the IC. A large proportion of these people will ultimately suffer from long-term complaints. That number is sadly only increasing, as better medical care increases the chances of people surviving the ICU. We certainly also have to take responsibility for that as a hospital.’

GROUNDBREAKING 

Erasmus MC’s IC has emphasised its status with the research into data-driven care. This is also apparent from the fact that Erasmus MC started using the DataHub on 1st January 2022. This is a place where innovative ideas come together, while looking at how clinical data’s streaming data analytics can be implemented in the healthcare process. ‘We are an innovative IC,’ Van Genderen states. ‘We have been working with virtual reality for some time now. To train our people, but also for the follow-up care of patients who have spent time in the ICU. We are really taking the next step with this data initiative, whereby we’re truly distinguishing ourselves as a result of the real-time aspect. We are one of the first intensive care units to achieve this. I’m not exaggerating when I say that we are truly committed to future healthcare. In that sense, we are doing groundbreaking work.’

Would you like to know more about the use of data within Erasmus MC’s intensive care department? Contact Michel van Genderen via m.vangenderen@erasmusmc.nl.

You can also read the (English) scientific publications regarding data by Van Genderen and his colleagues:

• Moving from bytes to bedside: a systematic review on the use of artificial intelligence in the intensive care unit (June 2021)
• Generating insights in uncharted territories: real-time learning from data in critically ill patients–an implementer report (August 2021)
• Progressive respiratory failure in COVID-19: a hypothesis (December 2020)
• Predicting thromboembolic complications in COVID-19 ICU patients using machine learning (October 2020)

Date: 11 March 2022