A comparative study on the inter-session and inter-examiner reliability of corneal power measurement using various keratometry instruments

• Main Authors: Md Mustafa, Md Muziman Syah, Abdul Mutalib, Haliza, Malkeet Singh, Sharanjeet Kaur, Mohd. Kamal, Khairidzan
• Format: Article
• Language: English; English; English
• Published: The Faculty of Medicine, International Islamic University Malaysia 2016
• Subjects: RE Ophthalmology
• Online Access: http://irep.iium.edu.my/51072/1/Vol15No1JUNE-2016-069-074.pdf; http://irep.iium.edu.my/51072/4/51072_comprative%20study%20on%20the%20inter_session_Scopus.pdf; http://irep.iium.edu.my/51072/5/51072_comprative%20study%20on%20the%20inter_session_WOS.pdf; http://irep.iium.edu.my/51072/; http://iiumedic.net/imjm/v1/download/Volume%2015%20No%201/original_article/Vol15No1JUNE-2016-069-074.pdf

Introduction: The purpose of this study was to evaluate inter-session repeatability, inter-examiner reproducibility and inter-device agreement of corneal power measurements from manual keratometer, autokeratometer, topographer, Pentacam high resolution and IOLMaster. Methods: Two sets of mean corneal power measurements (n=40) were compared for inter-session repeatability and inter-examiner reproducibility in each instrument. Repeatability and reproducibility were evaluated by within-subject standard deviation (Sw), coefficient of variation (COV) and intraclass correlation coefficient (ICC). A oneway repeated measures analysis of variance was conducted to compare differences in the corneal power between each instrument pair. The Bland and Altman analysis and Pearson’s correlation were employed to assess agreement and determine strength of relationship between measurements. Results: There were no significant differences in mean corneal power measurements between 2 different visits (p > 0.05). The Sw and COV values between 2 visits were lower than 0.09 D and 0.20 % respectively. The ICCs were stronger than 0.99 in all instruments. For reproducibility of each instrument, differences of the measurements between 2 different examiners were also insignificant (p > 0.05). The Sw and COV values between 2 examiners were lower than 0.11 D and 0.23 % respectively. The ICCs were 0.99 and above in all instruments. The 95% limit of agreement between instruments ranged from -0.29 to 1.13 D and the r-values were stronger than 0.84. Conclusion: The corneal power measurements using these 5 instruments were repeatable and reproducible. These instruments can also be used interchangeably, however the topographer should be used with caution.