Treat the World

What?

A retrospective cohort study on extremely preterm infants to compare the rates of retinopathy of prematurity (ROP) and its treatment in high-income countries participating in the International Network for Evaluation Outcomes (iNeo) of neonates (1).

Who?

Preterm infants (weighing <1,500 g at 240 to 276 weeks gestation) admitted to neonatal units between 2007 and 2013 in Australia, New Zealand, Canada, Finland, Israel, Japan, Spain, Sweden, Switzerland, Italy (Tuscany), and the UK.

How?

A total of 48,087 infants were included in the study. All data obtained from eye examinations was recorded using the International Classification of ROP, and the highest stage of ROP and treatment for advanced stages with laser or intravitreal anti-VEGF were analyzed. Information on survival past 32 weeks post-menstrual age was also collected.

Findings

• Mean survival to at least 32 weeks was 81.8 percent; the highest survival rates were reported in Japan (91.8 percent) and the lowest in Spain (67.8 percent). 
• Across all countries, screening data was available for 95 percent of surviving infants, with mean rates of ROP varying between 25.2 and 91.0 percent.
• Of the screened infants, 24.9 percent received retinopathy treatment. The highest treatment rates among screened infants were reported from Japan (30.4 percent) and the lowest from Switzerland (4.3 percent).
• Non-receipt of antenatal steroids, male sex, lower gestational age, lower birth weight for gestation, and delivery by caesarian section were risk factors associated with ROP treatment in surviving infants, but multiple births was not.
• Japan was found to have a higher relative risk for ROP and retinopathy treatment compared with all other iNeo countries.

Conclusions

The authors write that their findings show considerable variation in the rates of ROP and its treatment and that, to understand the differences in outcomes, “networks should adopt common definitions to describe pathology and improve consistency in interpretation, minimize the amount of missing data, and perhaps record an expanded dataset.”