Related Experiment Video
Updated: Mar 25, 2026
Monitoring Dynamic Growth of Retinal Vessels in Oxygen-Induced Retinopathy Mouse Model
Published on: April 2, 2021
The Colorado-retinopathy of prematurity model (CO-ROP): postnatal weight gain screening algorithm
Jennifer H Cao1, Brandie D Wagner2, Emily A McCourt1
1Department of Ophthalmology, University of Colorado School of Medicine, Aurora.
Insights
A new retinopathy of prematurity (ROP) screening model, CO-ROP, uses birth weight, gestational age, and weight gain to maintain high sensitivity while reducing unnecessary screenings. This model improves specificity for detecting ROP in premature infants.
Area of Science:
- Neonatology
- Ophthalmology
- Perinatal Medicine
Background:
- Retinopathy of prematurity (ROP) is a leading cause of blindness in premature infants.
- Current screening guidelines (2013) for ROP have limitations in specificity, leading to over-screening.
- A more accurate and efficient screening model is needed to optimize ROP detection and resource allocation.
Purpose of the Study:
- To introduce and evaluate a novel screening model for ROP, termed CO-ROP.
- The CO-ROP model incorporates birth weight, gestational age, and postnatal weight gain.
- To assess if CO-ROP improves specificity in detecting all grades of ROP compared to existing 2013 guidelines while maintaining sensitivity.
Main Methods:
- Retrospective review of medical records for 499 neonates meeting 2013 ROP screening criteria.
- Analysis of weekly weights using logistic regression to establish a postnatal weight gain criterion.
- Development of the CO-ROP model by combining gestational age, birth weight, and the new weight gain criterion.
Main Results:
- The CO-ROP model identified infants with gestational age ≤30 weeks, birth weight ≤1500 g, and net weight gain ≤650 g by 1 month.
- CO-ROP demonstrated 100% sensitivity for high-grade ROP and 96.4% for all ROP grades.
- Implementing CO-ROP could reduce the number of infants screened by 23.7%, with a potential 45.9% reduction if calibrated for high-grade ROP detection only.
Conclusions:
- The CO-ROP model offers a simple yet effective approach to ROP screening.
- It maintains high sensitivity for ROP detection while significantly decreasing the number of infants requiring screening.
- Despite a small sample size, CO-ROP shows promise for improving ROP screening efficiency in clinical practice and future research.
Purpose:
To describe a novel retinopathy of prematurity (ROP) screening model incorporating birth weight, gestational age, and postnatal weight gain that maintains sensitivity but improves specificity in detecting all grades of ROP compared to current 2013 screening guidelines.
Methods:
The medical records of 499 neonates from a single tertiary referral center who met the 2013 screening guidelines for ROP were retrospectively reviewed. Weekly weights were analyzed using standard logistic regression to determine the age at which the weekly net weight gain best predicted the development of ROP, which was designated as the postnatal weight gain criterion. The 2013 birth weight and gestational age criteria were included in an "and" fashion to form the CO-ROP model. Sensitivities and specificities in detecting high grade (type 1 and 2) and all grades of ROP were calculated.
Results:
The CO-ROP model screens infants with a gestational age at birth of ≤30 weeks and birth weight of ≤1500 g and net weight gain of ≤650 g between birth and 1 month of age. In our cohort, CO-ROP had a sensitivity of 100% (95% CI, 92.1%-100.0%) for high-grade (type 1 and 2) ROP and 96.4% (95% CI, 92.3%-98.7%) for all grades of ROP. It would reduce the number of infants screened by 23.7% compared to 2013 guidelines. Calibrating the model to detect only high-grade ROP would result in a 45.9% reduction in the total number of infants screened.
Conclusions:
CO-ROP is a simple model that maintains a statistically similar sensitivity in detecting all grades of ROP while significantly reducing the total number of required ROP screenings compared to 2013 guidelines. The study had a small sample size but shows promise for future research and clinical efforts.
