The purpose of this study is to determine if intensive training of Non-Nutritive Suck (NNS) using contingent positive reinforcement with mother's voice can improve the feeding outcomes of infants with congenital heart defect (CHD) at high risk for oral motor problems.

Randomized controlled trial to Improve Oral Feeding Outcomes in Infants with Complex Congenital heart disease using Pacifier Activated Device-Training with Mother's Voice Specific Aims Infants with complex congenital heart disease (CCHD) have high rates of delayed or impaired oral feeding skills, associated with other neurodevelopmental problems such as worse motor and language outcomes in early childhood. The acquisition of efficient suck strength and patterns is essential to promote later optimal oral feeding. Therefore, ineffective non-nutritive suck (NNS), one of the earliest feeding milestones, can result in later maladaptive oral motor skills or even oro-sensory aversions. Evidence-based interventions are lacking for CCHD infants hospitalized in the Cardiothoracic Intensive Care Unit (CTICU) to improve early oral feeding patterns and promote acquisition of later developmental milestones. NNS training (as opposed to simply offering a pacifier) safely improves oral feeding strength and rate in other high-risk populations, such as preterm infants, even when they have suffered severe neural insults. This type of training uses operant conditioning with positive stimuli to reinforce learning and establishment of functional pathways between lower and higher order neural networks. Furthermore, NNS training is associated with acquisition of essential developmental milestones in the first year. The investigators propose the following hypothesis: Intensive NNS-contingent training using reinforcement with mother's voice can improve the oral feeding outcomes of infants with CCHD. They further hypothesize that establishment of optimal NNS in the first 6 months can improve feeding and developmental outcomes by one year of age.

To test this hypothesis, the investigators propose a randomized controlled trial (RCT) in CCHD infants hospitalized for surgical repair. Because the proposed intervention has been demonstrated to only have positive effects in other high-risk infants, the study will use a wait-list design: CCHD infants who do not receive their intervention before surgery will receive it after, so that the entire cohort will eventually benefit from the intervention. The intervention will utilize a Pacifier Activated Music Player (PAM) with a sensor detecting changes in suck pressures and patterns, and a speaker contingently delivering mother's voice, in infants 0-6 months in the CTICU. Quantitative oral feeding metrics before and after the intervention, as well as, immediately after discharge will be used to measure efficacy. Secondly, to assess the impact of NNS training on developmental milestone acquisition, a standardized feeding and neurodevelopmental assessments of all trial infants at 12 months of age in the Follow-Up clinic will also be conducted. Randomization will account for differences in CCHD diagnoses known to impact outcomes. The study design will control for the type of surgical procedure as a proxy for severity of illness and exposure to anesthesia and for variables associated with neurodevelopment such as maternal education and prematurity.

Aim 1 A: Compare oral feeding skills as measured by suck strength, rate and burst patterns in CCHD infants with vs. without NNS training using PAM/mother's voice prior to surgery, in an RCT design. Aim 1 B: Correlate improvements in oral feeding skills after NNS training with PAM/mother's voice in the entire study cohort with higher scores on standardized feeding questionnaires at 1 month after discharge. Aim 2: Demonstrate that improvements in oral feeding skills in the CTICU will correlate with better neurodevelopmental outcomes as measured by standardized scores on the Bayley Scales of Infant and Toddler Development (BSID III) at 12 months. This study holds significant public health relevance as it validates an inexpensive, parent-centered rehabilitative strategy for infants at high-risk for impaired oral feeding skills, with a device and approach easily adapted to many levels of CCHD care and with the potential to improve longer-term neurodevelopment.

Inclusion Criteria:

• Infants that have a diagnosis of a complex congenital heart defect and scheduled for a surgical intervention; meet CTICU feeding guidelines criteria. Infants with acquired brain injury will be included, as well as those whose mothers speak languages other than English.

Exclusion Criteria:

• Infants that are on assisted ventilation, continuous positive airway pressure, general anesthesia within 24 hours; lethal congenital abnormalities or congenital brain malformation.

The assessors and the investigator are masked to the group assignment