**1. Introduction**

Throughout the history of extracorporeal membrane oxygenation (ECMO) from the first reported successful use in an adult patient in 1972 [1] and subsequently in neonates with respiratory failure [2], and then followed by the exponential increased use in adults following the CESAR trial [3], there has been interest in identifying patients who would benefit most from this high-risk resource-intensive therapy. Early efforts focused on using predictive scores of severe neonatal respiratory failure [4]; Newborn Pulmonary Insufficiency Index (a score developed by plotting serial inspired oxygen values with serial pH measurements in the first 24 hours of life) [5] and serial alveolar-arterial oxygen gradients (A-a DO2) [6], were deployed with mixed results. Since then, there have been extensive efforts at developing tools to aid in early identification of patients who would benefit most from timely institution of ECMO support and those with a high risk of mortality while being supported by ECMO. In this chapter, we provide a review of the currently available ECMO

prediction tools, their development, validation and limitations and an outline of potential future directions of ECMO decision support tools.
