# Table 3 Results of the regression analysis for strong agreement with the statement, “I do not like the location of my surgical scar” (Q12). In the left-hand column, study groups are written in bold; beneath them are all the independent predictor variables tested in the model (in order of decreasing significance). Predictor variables in gray were clearly not statistically significant (p ≥ 0.1). The further columns present the odds ratio, the 95%CI of the OR, and the p-value. [Recall the following about odds ratios: a) an OR of 1.0 would mean that the predictor variable had no effect on the outcome variable; b) an OR of 2.0 would mean that if the predictor variable was present (or for each unit of increase for ordinal or continuous variables, e.g. each year for age), then there were double the odds for the outcome variable of strongly agreeing with the statement of not liking the location of the scar; c) an OR of 0.5 would mean that if the predictor variable was present (or for each unit of increase for ordinal or continuous variables, e.g. each year for age), then there were half the odds of the outcome variable of strongly agreeing with the statement of not liking the location of the scar. So for example, among the patients who had mastectomy only, college graduates had 0.3 times lower odds of strongly agreeing with the statement, “I do not like the location of my scar.” It is important to keep in mind that the OR for a continuous variable, such as age, is for each increase of one unit (here, a year of age) and cumulative. So for example, among lumpectomy-only patients, women who were 10 years older (than another woman of whatever age) would have an OR of (0.96)10 = 0.66, and women who were 25 years older would have an OR of (0.96)25 = 0.36; among mastectomy-only patients, women who were 10 years older would have an OR of (0.97)10 = 0.72, and women who were 25 years older would have an OR of (0.97)25 = 0.45.]

OR lower
95%CI
upper
95%CI
p
Lumpectomy Only (n = 215)
Age (years) 0.96 0.94 0.99 0.010
Significant Other 2.3 0.86 6.3 0.098
Income (6 brackets) 0.9 0.7 1.2 0.5
College Graduate 0.7 0.3 1.8 0.5
Mastectomy Only (n = 140)
College Graduate 0.3 0.1 0.7 0.006
Age (years) 0.97 0.94 0.999 0.04
Income (6 brackets) 0.8 0.6 1.2 0.4
Significant Other 1.2 0.6 2.7 0.6
1. Note: The regression model fit the lumpectomy data well according to Pearson chi-Square (203, p = 0.6), likelihood ratio (10.9, p = 0.03), and Hosmer-Lemeshow (6.5, p = 0.6). The regression model fit the mastectomy data well according to Pearson chi-Square (138, p = 0.4) and likelihood ratio (14.7, p = 0.005) but was perhaps a marginally poor fit according to Hosmer-Lemeshow (15.2, p = 0.056)