Can provision of near vision glasses as an early intervention improve visual outcomes in infants at risk of perinatal brain insult? the Babies in Glasses (BiG) randomised feasibility trial

Objectives: We conducted a feasibility study to evaluate the feasibility of recruiting patients to examine the effect of near vision glasses in young infants at risk of cerebral visual impairment.

Design: A three-arm, parallel-group, open-label randomised feasibility trial.

Setting: Tertiary neonatal intensive care in London, UK.

Participants: We included babies born before 29 weeks of gestation or at full term with hypoxic ischaemic encephalopathy. Babies who needed ongoing inpatient care, with established eye anomalies or with very high refractive errors at baseline (<-6.00 dioptres (D) or >±8.00D) were not included. Infants with retinopathy of prematurity were not excluded.

Interventions: At 8 weeks corrected age, we allocated 18 infants to wear glasses (+3.00D over full cycloplegic refraction) immediately (intervention 1), 18 to wear the same glasses at 16 weeks (intervention 2) and 19 infants were allocated to standard treatment (no glasses).

Outcomes: Recruitment and retention of study participants (primary), compliance wearing glasses, preferential-looking visual acuity (with glasses) and visual function as determined using A Test Battery of Child Development for Examining Functional Vision at 3-month and 6-month age post-term.

Results: Of 70 eligible families, 55 consented and 34 attended baseline assessments, and 28 completed the study. Non-attendance was due mainly to prolonged inpatient stay, infant health and scheduling conflicts. Glasses were worn for similar periods in each group (Intervention 1: median 2 hours/day (95% CI 1 hour to 4 hours); Intervention 2: median 2 hours/day (95% CI 1.5 hours to 3 hours)). Visual acuity improved from baseline to 6 months. Mean (SE) LogMAR (Minimum Angle of Resolution) improvements were standard care: 0.47 (0.45); intervention 1: 0.66 (0.44); intervention 2: 0.37 (0.36). Among the 29 very preterm infants, there were similar findings: standard care: 0.35 (0.35); Intervention 1: 0.67 (0.47); Intervention 2: 0.34 (0.40). As a functional measure, object permanence was present at the following rates by randomised arm: standard care: 29%; whereas intervention 1: 56%; and intervention 2: 44% (OR intervention 1 vs standard care: 3.13 (95% CI 0.38 to 25.57), ie, not statistically significant).

Conclusions: We demonstrate feasibility for a definitive RCT (randomized controlled trial) with good recruitment and retention and observed potential benefits for vision and development following the dispensing of glasses at 8 weeks post-term age compared with untreated controls. We identified methodological modifications to further improve recruitment processes for a future larger study.