Abstract
Background
Ophthalmology is an image-oriented speciality where disease diagnosis is often aided by photographic information. Digital retinal imaging (DRS) technology can provide support with ophthalmic disease management, particularly in rural areas with limited access to ophthalmologists, and diagnostic medical screening. Retinopathy of prematurity (ROP) is a paediatric retinal disease that affects premature infants. The primary purpose of this study is to determine the diagnostic accuracy of wide field digital retinal imaging (WFDRI) in screening for retinopathy of prematurity, in comparison with the reference standard of Binocular Indirect Ophthalmoscopy (BIO).
Methods
Seven databases were searched from January 2010 to December 2013. Reference lists from included studies were reviewed, and hand searching of selected journals was carried out to identify additional studies. Information about baseline characteristics of participants in the study, the diagnostic screening protocols adhered to, the size of the infant population selected, eligibility criteria, and the healthcare setting where the tests were carried out, was extracted. A methodological quality assessment of the included studies was conducted, and related organisations, as well as authors, were contacted, where possible, for relevant publications on the research subject or clarifications in methodological design.
Results
In this review, 2 studies out of a total of 119 were included. Both studies measured the diagnostic efficacy of WFDRI as compared to the reference standard of BIO across all stages of ROP. Outcomes included sensitivity and specificity percentages, as well as positive predictive and negative predictive values. One study reported greater sensitivity and specificity outcomes for WFDRI in detecting moderate forms of ROP (89.5% and 97.1% for physician no.1, and 81.0 % and 98.5% for physician no.2). The second study found improved sensitivity and specificity outcomes for detecting treatment-requiring ROP using WFDRI (100% for both measures), and comparable diagnostic outcomes between WFDRI and BIO in detecting plus disease (100% for both modalities). Diagnosis of mild stage ROP by WFDRI was found to be poorer in comparison to the reference standard of BIO (a sensitivity of 58.6% only for WFDRI), and screening closure for acute ROP screening programmes, using WFDRI, was also inferior in comparison to BIO.
Conclusion
Screening for retinopathy of prematurity using wide field digital retinal imaging is highly accurate for moderate and severe stages of the disease. Mild forms of the disease are better detected using BIO. In addition, WFDRI provides inadequate closure for acute ROP screening protocols if implemented as part of a diagnostic screening initiative.