Romain Pirracchio, MD, PhD, Professor, Anesthesia, University of San Francisco, California, Chief, Anesthesia and Perioperative Care, ZSFG
By essence, intensive care medicine (ICU) is meant to be curative. However, ICU mortality remains high and long-term complications are frequent. Many acute conditions observed in critically ill patients, such as acute hypotension or acute rise in the intracranial pressure, are associated with higher morbidity and mortality. So far, the goal of ICU care has been to promptly detect such conditions and rapidly treat them. Indeed the longer the time spent under these conditions, the higher the chance of severe complications. However, the emergence of big data and machine learning may induce a complete shift in paradigm. If we were able to accurately predict the risk of an acute condition, we could preventively treat ICU patients to avoid such events to happen. Switching from curative to preventive ICU care may offer new opportunity to significatly reduce morbidiy and long-term complication in critically ill patients.