When the first citizen of Switzerland meets Calyps

Artificial Intelligence

The use of health data has become a major challenge that should empower the emergence of innovative services to provide better quality of care, while reducing costs. Nowadays, most of the data remains in silos, which jeopardizes interoperability and restricts the ability to innovate.

On August 4, CALYPS — represented by its CEO Tony Germini — welcomed a federal delegation composed of Ms Isabelle MORET, President of the Swiss Assembly and of H+, the umbrella association for Swiss hospitals, and Mr Hans STÖCKI, President of the Council of States. They shared with CALYPS their views on the use of Artificial Intelligence (AI) in the analysis of patient flows, in particular those related to COVID-19, as well as on current and future developments in data health.

Predicting hospital emergencies in Switzerland?

It is noticeable that very few AI projects have gone beyond the prototype stage in Switzerland, despite several successes in France.

While CALYPS provided the Swiss government with a detailed understanding of the COVID-19 situation and an anticipation of its evolution, Mr. GERMINI had to admit that a few health care providers were reluctant to introduce intra- or inter-hospital coordination solutions, despite the innovation they could bring, partly because these solutions may affect their organizations or, in some cases, threaten the balance of power. In addition, the nature of our Confederation does not facilitate communication between health actors, considering the different cantons, languages, data sources and structures of our little country, not to mention the obsolescence of means allowing data interoperability.

Over the past two years, CALYPS has learned several lessons from the deployment of AI solutions in hospitals of Switzerland or France. It is by constantly exchanging with the ground players — wards managers, doctors, nursing staff — and by relying on real data that innovation becomes possible. The key factor for success lies in working in succeeding stages, exploiting data from the outset (especially when using learning solutions such as the AI developed by CALYPS), then improving the solution by iteration. However, you have to be careful and set limits: when nothing relevant happens after sixty days, it is very likely that the direction you are heading in is not the right one.

There is no doubt that the post-COVID era will boost the digital transformation of hospitals. Caution should remain though: a few months ago, a board member of a Swiss hospital compared us to predatory colonizers whose sole ambition would be to exploit sources of raw commodities. While the anecdote may make smile, it is nevertheless representative of a mindset that tends to lock itself in an ivory tower. Data has little value when kept in silos: only data interoperability can bring innovation and improve the way healthcare is delivered (and managed).

Data interoperability is key

With better collaboration and a standardized exchange of information between several health actors, it would be possible to anticipate, among others, the care and rehabilitation of a patient even before he or she arrives in hospital, to ultimately make his or her journey smoother. Such coordination can only have a positive impact on the care chain, while reducing costs and the stress perceived by caregivers or patients. This must be done without adding platforms or screens, but simply by relying on the data already produced and available, without neglecting the vast issue of data privacy.