Prediction of retinopathy of prematurity using the WINROP screening algorithm in a South African population
Date
2018
Authors
Kesting, Samantha Jane
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Introduction
WINROP (Weight, IGF-1, Neonatal Retinopathy of Prematurity) is an online screening algorithm with sensitivities of 84.7 – 100% for severe ROP in other countries. The aim of this study is to test the efficacy of WINROP as a screening tool in a South African population.
Methods
This retrospective record review included infants born between 1 January 2013 and 1 December 2014 who underwent ROP screening. Gestational age, date of birth, weekly weights and final ophthalmology screening results were entered into WINROP. The outcomes of ophthalmology clinical examinations were compared to alarms triggered on the system. Sensitivity and specificity, positive predictive and negative predictive values, the mean time of alarm and average weight gain per week were calculated. Results were compared between patients with complete and missing weekly weights.
Results
Two hundred and twenty infants were included with a mean gestational age of 29.1 weeks (SD 1.8 weeks), and a mean birth weight of 1115.5g (SD 201.0 grams). Infants with complete weights totalled 193 patients with a mean gestational age of 29.1 weeks (SD 1.8 weeks) and a mean birth weight of 1115.8g (SD 201.2 grams), with no statistical difference between groups. The rates for all stages/severity of ROP were 5.9% and 2.3% for severe ROP. Weekly weight gain ranged between 4.6 to 83.8g/kg per week. WINROP triggered a high risk alarm in 70.5% of infants at a mean of 30.7 weeks gestational age. The sensitivity for severe ROP was 100%, but 76.9% for ROP in all infants and 83.3% in the complete weight group. The specificity was low for both severe and all ROP at 30.2% (complete weights 26.6%) and 30.0% (complete weights 26.5%) respectively.
Conclusion
Our rates of ROP are low. Rates of severe ROP have been found to be lower in African populations. The high number of alarms with a low negative predictive value, would reduce the number of screens by 29.5%. The alarms were triggered before scheduled screening, possibly helpful in planning discharges and follow up visits. The poor growth rates postnatally may have resulted in the increased alarms. This may be due to lower levels of IGF-1 and absolute weight gain postnatally in black infants. Screening algorithms relying on growth and IGF-1 levels are race dependent, which needs to be considered in future algorithms for African populations.
Description
A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree
of
Master of Medicine in the branch of Paediatrics
Johannesburg, May 2018.
Keywords
Neonatal Retinopathy of Prematurity