Abstract
Purpose: :
To report on the statistical methodology used for the analysis of longitudinal functional and morphological outcomes recorded in both eyes of participants enrolled in the CARMA clinical trial.
Methods: :
Carotenoids and Co-antioxidants in Age-Related Maculopathy is a randomised double masked prospective trial of L and Z with co-antioxidants; active (A) versus placebo (P). A total of 433 participants were randomised to A or P and followed for up to 3 years (7 visits). The primary outcome was change in best corrected distance visual acuity and secondary outcomes included measures of contrast sensitivity, raman spectroscopy, photopic interferometry, fundus photograding and serum antioxidants. Eligible patients could have one or both eyes entered as study eyes. Correlation coefficients between recorded BCVA for both eyes were generated and weights derived to account for this. The general linear model was used to examine continuous outcomes in a longitudinal time-frame. A 14 point joint nominal and ordinal categorical scale from photograding was re-ordered and condensed to yield a 6 point ordinal scale.
Results: :
The correlation between eyes for change in DVA at 12 months was 0.27 showing that when outcomes are obtained from both eyes, the information from the second of the pair is less than that from the first. Weights of 0.81 and 1.27 were computed for participants who contributed one or both eyes to the analysis. Of 1665 visits with morphological data, 8.8% of eyes progressed from less severe to more severe AMD status from one visit to the next. Less than 0.2% of eyes showed a change in the opposite direction.
Conclusions: :
The method described allows outcomes recorded from both eyes of participants to be used effectively
Clinical Trial: :
www.ISRCTN.org ISRCTN94557601
Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • age-related macular degeneration • antioxidants