|
|
Analysis of Correlated Two Eyes Data for Ophthalmic Clinical Research |
Guishuang Ying,Yuanbo Liang |
Center for Preventive Ophthalmology and Biostatistics, Scheie Eye Institute, University of Pennsylvania,Pennsylvania, America;Center for Clinical Ophthalmic and Optometric Research, Wenzhou Medical University, Wenzhou 325027,China. |
|
|
Abstract Objective: To describe the proper statistical approach in ophthalmic research for analyzing correlated data for contralateral eyes of the same subjects. Methods: For data from a paired design and a design for two eyes that are commonly used in ophthalmic research, we described the appropriate statistical approaches for analyzing the correlated data for both eyes of the same subjects. As an example, we compared the intraocular pressure between patients < 60 vs. ≥ 60 years of age, using the inappropriate and appropriate statistical analysis approaches. Results: The inappropriate analysis of data for one eye leaded only to a biased or inefficient estimate of the difference between two groups. The analysis of data for both eyes of the same subjects without accounting for inter-eye correlation had a smaller p-value (P=0.02) than the analysis that accounted for inter-eye correlation (P=0.098). Conclusions: Appropriate statistical analytical approaches should be used to account for the inter-eye correlation. The statistical analysis using data for only one eye or ignoring the inter-eye correlation can lead to an incorrect conclusion.
|
Received: 17 May 2017
|
Corresponding Authors:
Guishuang Ying, Center for Preventive Ophthalmology and Biostatistics, Scheie Eye Institute, University of Pennsylvania, Pennsylvania, USA; Center for Clinical Ophthalmic and Optometric Research, Wenzhou Medical University, Wenzhou 325027, China (Email: gsying@mail.med.upenn.edu)
|
|
|
|
[1] |
Age-Related Eye Disease Study 2 Research Group. Lutein+zeaxanthin and omega-3 fatty acids for age-related macular degeneration: the Age-Related Eye Disease Study 2(AREDS2)randomized clinical trial. JAMA, 2013, 309(19): 2005-2015.DOI: 10.1001/ jama.2013.4997.
|
[2] |
Ray WA, O'Day DM. Statistical analysis of multi-eye data in ophthalmic research. Invest Ophthalmol Vis Sci, 1985, 26(8):1186-1188.
|
[3] |
Rosner B, Glynn RJ, Lee ML. Incorporation of clustering effects for the Wilcoxon rank sum test: a large-sample approach.Biometrics, 2003, 59(4): 1089-1098.
|
[4] |
Glynn RJ, Rosner B. Accounting for the correlation between fellow eyes in regression analysis. Arch Ophthalmol, 1992, 110(3): 381-387.
|
[5] |
Glynn RJ, Rosner B. Regression methods when the eye is the unit of analysis. Ophthalmic Epidemiol, 2012, 19(3): 159-165.DOI: 10.3109/09286586.2012.674614.
|
[6] |
Liang KY, Zeger SL. Regression analysis for correlated data.Annu Rev. Pub. Health, 1993, 14: 43-68.
|
[7] |
Wei LJ, Glidden DV. An overview of statistical methods for multiple failure time data in clinical trials. Stat Med, 1997, 16(8):833-839; discussion 841-851.
|
[8] |
Ying GS, Liu CC. Analysis of the clustered data using SAS®systems. NESUG Conference Proceedings, 2006.
|
[9] |
Katz J. Two eyes or one? The data analyst's dilemma.Ophthalmic Surg, 1988, 19(8): 585-589.
|
|
|
|