Comparisons of Demographic Factors and Treatment. Table 2 gives results for mean keratometric astigmatism values when compared among levels of the demographic factors. There is almost no difference in mean keratometric astigmatism between treatment groups at baseline (CL: 1.98 D; IOL: 2.00 D; Diff: -0.02 D, p = 0.94); this should be expected, though, since patients were randomly assigned to treatment groups. However, the difference between treatment groups at 1 year of age is -0.47 D (CL: 1.62 D; IOL: 2.09 D), a statistically significant difference (p = 0.023). Despite the significant difference at 1 year of age, the difference in treatment groups in the mean change from baseline to 1 year is -0.30 D (CL: -0.29 D; IOL: 0.01 D), not a significant difference (p = 0.343). Figure 3(a) shows boxplots of keratometric astigmatism at baseline, 1 year, and the change, according to treatment group. Figure 3(b) displays boxplots of keratometric astigmatism at the two time points, and the change between them, according to age group. Fifty-two of the 114 patients had surgery at less than 49 days from birth, while 62 patients had surgery at 49 days or greater. Keratometric astigmatism levels at baseline in these two groups differed by -0.40 D, not a statistically significant difference (< 49 Days: 1.77 D; > 49 Days: 2.17 D; p = 0.104). The difference between the two groups remained approximately the same at 1 year of age (Diff: -0.30 D; p = 0.110); change from baseline to 1 year of age was nearly identical in the two age groups (Diff: 0.005; p = 0.988); these results can be seen in Table 2. Boxplots showing keratometric astigmatism versus gender can be found in Figure 3(c). Males, comprising 54 of the 114 patients, had a mean 2.19 diopters of astigmatism at baseline, while the 60 females had a mean of 1.81 diopters. This was not a statistically significant difference (0.38 D; p = 0.118). At 1 year of age, males and females had a decreased difference of only 0.06 D (Males: 1.89 D; Females: 1.83 D; p = 0.761). The mean change in keratometric astigmatism between baseline and 1 year in males and females was not significant (Diff: -0.22 D; p = 0.482); however, the variances in the two groups were significantly different. Thus, the Xxxxxxxxxxxxx approximation for the 95% confidence interval shown in Table 2 was used.
