Abstract
Background: Big data is a hot topic and provides an unprecedented opportunity to improve care on the neonatal intensive care unit. Personalization and predictive models can be developed as long as enough of the right data is available. But data acquisition is usually not implemented in a way that all the raw measurement data is stored. Let alone that all clinical events are registered.
Methods: Collecting all patient monitoring data together with the clinical events helps us to learn from our patients. It enables the development of more personalized care, better visualization of data and the development of dynamic predictive models. Systems like HERO (Medical Predictive Science Corporation), expediting the diagnosis and treatment of septic patients with a day, show the potential of such models. Models developed in one center need to be validated in other centers in order to be truly useful. This is only possible with (international) collaborations. Standardization of data storage and annotation are crucial.
Results: The Erasmus MC - Sophia Children's Hospital has started with logging all the real-time data from their neonatal patients. The intention is to set up a European network together with other NICU's to enable to store and share big data and develop and validate dynamic predictive models.
Conclusions: Big data and predictive models might well be the next game changer in neonatology after surfactant and the pulse oximeter. But the full potential is only unlocked when collaborations are formed such that the developed models are widely applicable.
Methods: Collecting all patient monitoring data together with the clinical events helps us to learn from our patients. It enables the development of more personalized care, better visualization of data and the development of dynamic predictive models. Systems like HERO (Medical Predictive Science Corporation), expediting the diagnosis and treatment of septic patients with a day, show the potential of such models. Models developed in one center need to be validated in other centers in order to be truly useful. This is only possible with (international) collaborations. Standardization of data storage and annotation are crucial.
Results: The Erasmus MC - Sophia Children's Hospital has started with logging all the real-time data from their neonatal patients. The intention is to set up a European network together with other NICU's to enable to store and share big data and develop and validate dynamic predictive models.
Conclusions: Big data and predictive models might well be the next game changer in neonatology after surfactant and the pulse oximeter. But the full potential is only unlocked when collaborations are formed such that the developed models are widely applicable.
Original language | English |
---|---|
Pages (from-to) | 225-226 |
Number of pages | 2 |
Journal | Journal of Neonatal-Perinatal Medicine |
Volume | 10 |
Issue number | 2 |
DOIs | |
Publication status | Published - 2017 |
Bibliographical note
Embase identification number (PUI) L624548614(entry date:) 2018-10-30
M1 - 2