Intensive care medicine is one of the most technology-laden areas of healthcare. It consumes ~1-2% of GDP (or ~10% of health budgets) in modern countries, while treating 0.1% of patients or less.
Intensive care units (ICU) patients are very complex and highly variable, making management difficult.
ICU medicine involves a wide range of sensors and devices to measure patient state and further devices to deliver care.
These existing technologies, such as infusions pumps and mechanical ventilators, are relatively simple devices mechanically and electronically. They thus provide an excellent foundation upon which to develop digital twins (DT).
We will focus on two important areas of intensive care where DT can revolutionize care: mechanical ventilation and nutrition.
Mechanical ventilation (MV) is the core therapy for patients suffering from life-threatening respiratory failure or unable to breathe due to lung injury or diseases, such as pneumonia, acute respiratory distress syndrome (ARDS), or Covid-19. MV supports the work of breathing, ensures adequate gas exchange, and recruits pulmonary volume. However, suboptimal MV settings lead to over-distension and ventilator induced lung injury (VILI), all of which increase length of MV, length of stay, morbidity, mortality, and cost, where MV doubles the cost per ICU day. The main challenge is to deliver MV safely and optimally, without insight into pulmonary mechanics necessary, something which our models can now do.
Nutrition support for ICU patients is a major, controversial debate. The main challenge is to optimise nutrition based on the patient-specific transition from the acute phase to the recovery phase.
However, no biomarker or signal exists to accurately identify this patient-specific transition. Recent reviews highlight the critical need for this information, where it would significantly improve and personalise nutrition delivery, reducing complications and related costs of care.