Prediction of three year outcomes using the Bayley-III for surgical, cardiac and healthy Australian infants at one year of age.

BACKGROUND: Previous predictive research has predominantly focussed on infants who are preterm, low birth weight, who have a specific condition or who have undergone a specific procedure. AIM: This study investigated the ability of outcomes at one year of age to predict outcomes at three years using the Bayley-III for infants who have undergone early major cardiac surgery (CS) or non-cardiac (NC) surgery and their healthy peers. STUDY DESIGN: Participants who were part of the Development After Infant Surgery (DAISy) study who had complete Bayley-III assessments at one and three years of age were included in the analyses. This included 103 infants who had undergone CS, 158 who had NC surgery and 160 controls. RESULTS: Bayley-III outcomes at one, although statistically significantly associated with three year outcomes in all domains were weak predictors of those outcomes for CS, NC surgical and healthy infants. Specificity for three year outcomes was good for cognitive, receptive language and fine motor domains for infants who had undergone CS and NC surgery. Sensitivity for <-1 SD at three years was poor for cognitive, expressive and receptive language, and fine motor outcomes for CS and NC surgical participants. CONCLUSION: It remains difficult to predict how performance at one year on the Bayley-IIII predicts performance at three on the Bayley-III for infants who have undergone early major CS or NC surgery and for healthy Australian infants.