The diagnostic accuracy and feasibility of paediatric visual field tests: A systematic review.

Abstract

Purpose: Alternative non-standard paediatric visual field (VF) tests have been developed to address the challenges associated with standard approaches. However, diagnostic accuracy of these new VF tests has not yet been rigorously evaluated. This systematic review aims to explore diagnostic accuracy and feasibility of non-standard VF tests in paediatric patients.

Methods: The following databases were searched for English language studies comparing a non-standard paediatric VF test to standard methods, such as standard automated perimetry (SAP), manual kinetic perimetry (MKP) and confrontation testing (CT): EMBASE, PubMed, Ovid MEDLINE, Web of Science, Scopus, VisionCite, Cochrane Library, ClinicalTrials.gov, African Index Medicus, LILACS, Trip and grey literature databases. Studies included were of children ≤18 years old with suspected or known VF defects (n > 3). Case reports, case series, editorials and letters were excluded. This review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool.

Results: Thirty studies (n = 2074 children, age range: 2 months to 18 years) published between 1990 and 2023 met the inclusion/exclusion criteria. Twenty index tests were reported, sorted into four categories of methods: behavioural/observational (5), electrophysiological (2), modifications of standard perimetry (11) and eye tracking (2). Risk of bias, based on the QUADAS-2 results, was unclear or high for most studies. Sensitivities of 10 studies (behavioural/observational [2]; electrophysiological [2]; modifications of standard perimetry [2] and eye tracking [4]) were 60%-100%, 75%, 80%-94% and 70%-100%, and specificities were 98%-100%, 86%-88%, 68%-100% and 50%-100%, respectively.

Conclusions: Alternative non-standard paediatric VF tests, and behavioural/observational methods in particular, show potential for adequate diagnostic accuracy and feasibility, although limited to certain conditions and ages. Studies with complete reporting and low risk of bias are needed to determine the diagnostic accuracy and feasibility of non-standard paediatric VF tests.