Although the data are ordinal in nature and the use of inferential statistics is not optimal in this situation, then they were used for several reasons. First, this was considered preferable to a large volume of chi-square tests. A comparison of medians was also considered but while groups often had similar median values, subtle differences emerged when means were used. Finally, the sample size for the majority of the comparisons was sufficiently substantial to allow the use of inferential statistics in this situation [6]. However, the more conservative nonparametric tests were used to assess all associations. The associations of age and body mass index with the seven questions were assessed by means of the nonparametric Spearman’s correlation.

The association of gender and presence of a previous surgical scar (abdominal or nonabdominal) with the seven questions was assessed by means of the Mann-Whitney U test, while the association for the three levels of age and BMI were assessed by means of the Kruskal-Wallis test. In order to provide an adequate sample to allow for subgroup analysis, enrolment was aimed at approximately 300 patients. For all analyses, the significance level was set at P < 0.05 (two-sided), although results that fell short of statistical significance were noted if they were deemed to be of clinical interest. 3. Results Three hundred thirty-five patients completed the survey. Demographic and physical characteristics are summarized in Table 1. Nine percent were ��29 years of age, 26% were 30�C49 years, and 64% were ��50 years; for BMI, 29.

9% were at a healthy weight, 34.9% were overweight, and 29.6% were obese (6% were missing height and/or weight). As this was a voluntary, anonymous survey, there were very few missing data (see Table 2). For the few items that were missing, analyses were completed on the subset without missing data, as the type of detailed information typically required for imputation was not collected. Table 1 Patient demographics. Table 2 Missing data (n = 335). 3.1. Attitudes towards Scars Younger respondents (<50 years of age), females, and those of a healthy weight indicated that cosmetic issues such as scars were more important, as compared to older, male, and heavier respondents (P �� 0.001 for all three comparisons) (Table 3). Amongst all surveyed, 87% of respondents had some type of scar. Of these, 58% indicated that it did not bother them at all, but 9.9% indicated that they were bothered quite a bit or extremely by their scar(s). Women placed significantly greater importance on abdominal scars than men and were more greatly impacted Anacetrapib by them; fifty-six percent of women were bothered by some degree by their current scars as compared with 23% of men (P < 0.001).