Abstract
Purpose :
To describe the correlation between media haze on fundus photographs and the development of treatment-warranted retinopathy of prematurity (TW-ROP).
Methods :
A retrospective case control study. Fundus photographs of premature newborns enrolled in a telemedicine screening protocol found to have TW-ROP and those of untreated controls within the same cohort at six neonatal intensive care units from December 1, 2005 to 2015 were reviewed to grade degree of haze based on modified NIH grading system at different time points. Statistical analysis was performed using McNemar's chi-squared statistic for matched data to assess the relationship between media haze in newborns with TW-ROP compared with controls with non-TW-ROP.
Results :
Images of 62 eyes of 31 newborns were analyzed for the TW-ROP group, and images of 62 eyes of 31 newborn were analyzed for the the non-TW-ROP groups. The TW-ROP and non-TW-ROP group were similar with respect to baseline characteristics including gender and multiplicity although there was a significant difference at baseline with respect to gestational age and birth weight. While there was a trend towards more haze in the TW-ROP group at baseline exam, this was not statistically significant (p=0.1206). When comparing degree of haze at all time points, there was no significant difference between the two groups (p=0.1947). There was strong inter-eye correlation in level of haze. In photographs from subsequent exams after the baseline exam, degree of haze improved in both groups.
Conclusions :
In this exploratory analysis, presence of media haze on fundus photography of newborns undergoing ROP screening was not strongly correlated with development of TW-ROP. Media haze may preclude adequate fundus examination in infants at risk for TW-ROP, especially at baseline exam when it is most prevalent. Fortunately, the degree of haze decreases with time, thereby allowing better visualization of the fundus for adequate screening.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.