Smartphone application links severity of retinopathy of prematurity to early motor behavior in a cohort of high-risk preterm infants

Abstract

Purpose: To evaluate the General Movement Assessment (GMA) with the Motor Optimality Score-Revised (MOS-R) as a neurodevelopmental marker in infants with retinopathy of prematurity (ROP).

Methods: Infants screened prospectively for ROP were evaluated at 3 months' post-term age using a smartphone application to complete the GMA and MOS-R. Results were analyzed by ROP severity.

Results: Of 105 enrolled infants, 83 completed the study. Of these, 54 (65%) had any ROP, 32 (39%) had severe ROP, and 13 (16%) had type 1 ROP. The proportion with aberrant GMA was significantly higher in infants with severe ROP (14/32 [44%]) compared with infants who had milder ROP (8/51 [16%]; P = 0.006). Of those with severe ROP, there was no significant difference comparing infants with type 1 ROP treated with bevacizumab (7/13 [54%]) to infants with type 2 ROP without treatment (7/19 [37%]; P = 0.47). Although the presence of any ROP, stage of ROP, and severe ROP each predicted lower MOS-R scores on univariate analyses, only severe bronchopulmonary dysplasia and markers of brain injury remained significant in the multivariate analysis.

Conclusions: The GMA was a convenient, short-term method of data collection with low attrition. Although severe ROP initially appeared linked to poor early motor scores, this association is likely confounded by neurological and respiratory complications, which frequently accompany severe ROP.

Fig 1.

Stage of retinopathy (ROP) and general movement assessment. Boxplot showing median MOS-R scores by ROP stage, with interquartile ranges. Infants with ROP stages 0–1 had significantly higher MOS-R scores than infants with ROP stages 2 (P = 0.01) or 3 (P = 0.004); there were no significant differences between infants with ROP stages 2 and 3 (P = 0.81; overall 2-df P = 0.005).