A smartphone application to measure neurodevelopmental outcomes among infants with retinopathy of prematurity

Abstract

Introduction: To describe a smartphone application (BabyMoves) for remote neurodevelopmental evaluation of infants with ROP using the General Movement Assessment (GMA). The GMA reliably predicts cerebral palsy (CP) at 3-4 months corrected age, is the most sensitive and specific early marker of neurocognitive impairment, and correlates with white matter abnormalities on MRI as well as Bayley (BSID-III) scores. Methods: Prospective study including patients screened for ROP from 10/2019 to present. The primary outcome measure was the absence of fidgety movements on the GMA, which predicts CP. Results: Seventy families have consented, of which 38 videos were completed-a majority black (80%) and female (61%). Overall, 22/38 had any ROP, 11/38 had severe ROP, and 4/38 received primary bevacizumab for type 1 ROP. Twelve (32%) had videos suspicious for Cerebral Palsy (CP). Compared to infants with mild or no ROP, infants with severe ROP had 4x increased odds of CP (P = 0.06), which was not significant after controlling for severe intraventricular hemorrhage (IVH). At this time, bevacizumab does not appear associated with CP (P = 0.41), while shunted hydrocephalus is strongly linked to CP (P = 0.01). Conclusion/Relevance: While enrollment is ongoing, preliminary results are reassuring that remote assessment via a smartphone application can expedite neurodevelopmental outcomes among patients with ROP. While BSID scores are needed, the GMA can help establish the developmental trajectory. Particularly in post-COVID times, remote assessment has signficant benefits. BabyMoves offers a convenient, economical way to evaluate the systemic implications of ROP treatment.