Title Multifactorial risk environment for retinopathy of prematurity /
Translation of Title Daugiafaktorinė aplinkos rizika neišnešiotų naujagimių retinopatijai.
Authors Drazdienė, Nijolė ; Bagdonienė, Rasa ; Sirtautienė, Rasa ; Vezbergienė, Nijolė ; Sliesoraitienė, Viktorija ; Sliesoraitytė, Ieva
Full Text Download
Is Part of Acta medica Lituanica. 2006, vol. 13, no. 3, p. 141-146.. ISSN 1392-0138
Keywords [eng] Retinopathy of prematurity ; Risk factors ; Artificial neural networks
Abstract [eng] Background: Retinopathy of prematurity (ROP) might be prevented by a timely diagnosis and appropriate treatment applied according to the risk factors. The goal of our research was to combine the optimal clinical and epidemiological indicators for ROP risk detection. Materials and methods: A retrospective observational research was carried out at Clinic of Neonatology of Vilnius University Children’s Hospital. The research combined examination of epidemiological and clinical characteristics of premature neonates born in 2005 and analysis of ROP protocols. Multifactorial risk environment for ROP pathological process development was elucidated. Results: The infants’ age at ROP onset was four to six weeks. Their mean gestational age was 28.1 ± 0.9 weeks, the mean birth weight being 1250 ± 214 g. Infants at the greatest risk of ROP weighed 1500 g or less at birth (p < 0.001), and their gestational age was 30 weeks or less (p < 0.05). An inverse relationship was found between the incidence and severity of ROP and birth weight and gestational age (r1 = -0.8; r2 = -0.7; p < 0.01). Low Apgar scale rates were positively associated with ROP stage and accordingly with a high ROP risk level (r3 = 0.9; p < 0.001). Delivery and pregnancy failure increase the ROP pattern as following infections and bleeding during delivery are the leading pathological status elevating ROP stage. Oxygen therapy had been applied in all ROP stages. Concomitant neonate pathologies range within a moderate to high risk level for ROP development (p < 0.001). Conclusions: The tool based on artificial neural networks has the potentiality of identifying the risk level for ROP development with high sensitivity considering positive and negative predictive values.
Type Journal article
Language English
Publication date 2006