The validity of telemedicine-based screening for retinopathy of prematurity in the Premature Eye Rescue Program in Hungary

Abstract

Purpose

To assess the validity of wide-field digital imaging (WFDI) and telemedicine-based screening compared with examination by binocular indirect ophthalmoscopy (BIO) and to present some of the results from the first five years of telemedicine-based screening in the Premature Eye Rescue Program in Hungary.

Methods

We performed a retrospective analysis in two periods that aimed to assess (a) the validity of retinal digital imaging and (b) routine bedside screening. The validity was assessed in two neonatal intensive care units (NICUs), one in the First Department of Paediatrics and the other in the Second Department of Obstetrics and Gynaecology, Semmelweis University. The telemedicine-based WFDI (WFDI-TM) screening program was introduced in two phases. In the first phase (from 30 November 2009 to 8 August 2010), BIO and WFDI were performed by the same paediatric ophthalmologist (Group A). In the second phase (from 9 August 2010 to 29 March 2011), BIO was performed by the paediatric ophthalmologist, while retinal images were captured by a trained neonatal transport nurse practitioner (Group B). BIO screening was the reference method as a gold standard in both phases.

Results

During the validity assessment period 634 examinations were performed in 153 preterm infants. Overall, 76 babies were screened in Group A and 80 were screened in Group B. We found lower sensitivity and specificity in cases of any ROP (sensitivity 86%, specificity 99%) compared with those of treatment-requiring retinopathy of prematurity (TR-ROP) (both sensitivity and specificity 100%).

In the Premature Eye Rescue Program between 1April 2011 and 31 March 2016, we used WFDI in 3035 infants (4589 procedures). Over this five-year period, 100 (9.6%) infants were treated by laser, and no child who received care in any of the Semmelweis University NICUs became blind from ROP.

Conclusions

(a) WFDI-TM ROP screening is a useful and efficient approach, although it cannot completely replace BIO; (b) no ROP-related blindness developed among the screened preterm babies; and (c) WFDI-TM ROP screening can be implemented in the logistics of a neonatal emergency and ambulance team infrastructure with neonatal transport nurse practitioners as ‘photographers’.