Premium
Measurement properties of reading tests in subjects with maculopathy
Author(s) -
Brussee Tamara,
Alagahgi Basel,
Nispen Ruth M. A.,
Rens Ger H. M. B.
Publication year - 2021
Publication title -
acta ophthalmologica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.534
H-Index - 87
eISSN - 1755-3768
pISSN - 1755-375X
DOI - 10.1111/aos.14574
Subject(s) - repeatability , reading (process) , test (biology) , reliability (semiconductor) , reproducibility , vision test , optometry , computer science , visual acuity , psychology , medicine , ophthalmology , mathematics , statistics , linguistics , paleontology , philosophy , power (physics) , physics , quantum mechanics , biology
Purpose A reliable reading test provides a standardized measure of the visual component of reading performance. This study evaluated reproducibility, agreement and feasibility of five Dutch language continuous text reading tests used in clinical practice and research in visually impaired participants. Methods In 42 participants with macular pathologies (mean age 77 years), the Colenbrander Reading Card (Colenbrander), International Reading Speed Texts (IReST), Laboratory of Experimental Ophthalmology (LEO) charts, ‘de Nederlanders’ (NED) and the Radner Reading Charts (Radner) were evaluated. The coefficient of repeatability was calculated for different reading parameters, and agreement between the reading tests was determined. Results Between the reading tests, the differences found in repeatability for reading performance were mainly within the limit of one line (0.1 logMAR). Exceptions were the inter‐session repeatability for critical print size: Colenbrander (0.35 logMAR), LEO (0.34), Radner (0.23). The highest agreement was found between the LEO and Radner; Reading acuity bias 0.03 logMAR (SD 0.10), CPS 0.03 (0.12). Conclusion This study shows that reading performance results obtained with reading tests are not always reliable and reading parameters could not always be properly assessed in participants with maculopathies. Therefore, choices regarding which reading test to use especially for research purposes should be based on both the feasibility and reliability of the reading test. The NED (a historical test) was the least feasible, and it is recommend that this test is no longer used. To allow standardized and comparable analysis of reading performance a highly standardized reading test, like the Radner is recommended.