Estimating cardiac intensive care patients’ responsiveness to late conservative fluid management using systems analysis

Abstract

Fluid overload is a common problem in critically ill patients who underwent surgery and is accompanied by a range of risks for severe organ damage. Therefore, in stable patients excess water should be removed using fluid-restricting treatment strategies in order to reach a neutral fluid balance as early as possible. Within this work, second order discrete-time transfer function models were identified for 10 selected cardiac patients staying at the intensive care unit for at least 5 days. For model estimation, the cumulative fluid intake and the cumulative fluid balance time series were used as model input and output parameters. Each individual model is capable of describing the respective patient’s cumulative fluid balance trajectory as response to the actually applied fluid therapy. In the context of decision support, systems analysis can therefore offer valuable tools not only for describing an individual patient’s reaction to a certain fluid therapy, but also for identifying patients with rebalancing issues at an early stage of recovery. In daily clinical practice, patient-specific transfer function models can be used for predicting the responsiveness to planned medical interventions, improving the fluid intake regime by focusing on the avoidance of a further increasing fluid overload and returning patients’ fluid balances to an acceptable level before release from the intensive care unit.