An artificial intelligence tool to predict fluid requirement in the intensive care unit: a proof-of-concept study

Table 1 Patient variables evaluated as possible predictors of maintenance fluid requirement.

General demographics	Age
	Sex
	Weight
Physiological variables	Blood pressure
	Heart rate
Variables that affect and/or represent total body water	Total fluid input during the first 24 hours
	Total fluid output during the first 24 hours
	Serum creatinine, as a surrogate marker of kidney function
	Serum sodium
Variables that affect insensible fluid loss	Body surface area
	Temperature
Variables that reflect severity of illness	Serum albumin
	Serum lactate
	Maximum number of vasopressors and inotropes
	Maximum number of sedatives and narcotic agents
	Serum bilirubin
	Haemoglobin
	Platelet count
	PaO₂:FiO₂ ratio

FiO₂ = fraction of inspired oxygen; PaO₂ = partial pressure of oxygen in arterial blood.

ISSN: 1364-8535