Digital innovations such as, for example, apps, smartwatches, AI applied to medical images and home tests are desperately needed to keep healthcare affordable and humanised. But not every innovation contributes to this, which is why it is essential to properly assess their value. Erik Koffijberg, professor of technology assessment of digital health innovations, says so in his Inaugural lecture on Friday 15 December.
Many innovations aim to improve the quality of healthcare while making it more expensive. Only 20% of innovations make care both better and cheaper. According to Koffijberg, this is not necessarily a problem. He says: "As long as we know which innovation contributes to the sustainability of care and which does not. We currently have too little insight into which digital innovations have added value. In terms of evaluation, we do a lot better with regular drugs."
Not a drug
But digital innovations are not the same as a drug. A drug is developed over five to ten years, studied in clinical trials, and then evaluated once for added value. Digital innovations, on the other hand, are developed super fast and continuously advanced. "They are 'moving targets' that cannot be evaluated with traditional methods and clinical studies," says Koffijberg. To estimate the added value of digital innovation, Koffijberg works with simulation models in his research. "In them, we simulate the care pathways that patients go through in their lives, with and without the use of the innovation, and then compare the corresponding consequences for quality of life and life expectancy, healthcare costs and deployment of healthcare staff. We already use those models for one-off evaluations, but we can easily use them much more often, and that is essential with digital innovations."
Continues evaluation
To get a better grip on the value of digital innovations, Koffijberg argues in his oration for a more continuous evaluation. "In the case of digital innovations, today's evaluation is already outdated tomorrow." Simulation models in particular can be used for continuous evaluation, as they can immediately provide a new estimate of added value when new data and insights become available. Koffijberg: "Continuous development must go hand in hand with continuous data collection and continuous re-evaluation. This will only succeed if developers, simulation experts and end-users work together on this."
Context-dependent
The value of many digital innovations is also linked to the context: who, when, where and how the innovation is used. This increases the need for efficient evaluation with simulation models, and for sharing those models and their outcomes. For example, artificial intelligence is increasingly used in hospital radiology departments to assess medical images. Hospitals evaluate its value mainly locally and sometimes reach different conclusions because of differences in context. "A small hospital with only a few experienced radiologists will benefit much more from the help of artificial intelligence than a large hospital with several experienced radiologists around 24/7," Koffijberg says. "Fine, but we don't need to reinvent the wheel for all these evaluations. By including these context factors in our simulation models, we can perform evaluations more efficiently, more consistently and repeatedly."
More information
Prof Erik Koffijberg holds an MSc in Technical Computer Science from Delft University of Technology and a PhD in Decision-Analytic Modelling from Utrecht University in the Netherlands. He is a professor of Technology Assessment of Digital Health Innovations and chair of the Health Technology & Services Research section (Faculty of BMS / TechMed Centre). On Friday 15 December, he will deliver his oration entitled: 'Impact of Digital Health Innovations: optimisation through simulation'. This oration will be given in English.
