Stata ic version 14

Data were analyzed using STATA/IC (version 14.2). Descriptive measures were presented in percentages. Differences in distribution between groups were compared using logistic regression with an estimate of 95% confidence intervals (CIs). For all tests, p-values of 0.05 or less were used to determine the level of significant difference. Multivariate analysis was employed to assess the relationship between levels of knowledge about antibiotics and receiving information and demographic data such as gender, age, area of residence, education level and wealth status as shown in Table 2. The variables were selected by reviewing literatures and analysing the significant bivariate association. We used the cut off point for the outcome variable of level of knowledge as lower and equal to and higher than three correct answers (>50% of total questions).

First, we compared the patients` data across diabetes status. The normality assumption for continuous variables was checked graphically by a histogram plot and the Shapiro-Wilk statistical test. Because of the non-normal distribution of the study variables, they were reported as the median and interquartile range (IQR) and were compared using Mann-Whitney U tests. In order to compare categorical variables, Chi-square tests were used, and the data were reported as numbers and percentages. We used univariate logistic regression analysis to evaluate the association between diabetes and in-hospital outcomes. Then, we used the change-in-estimate (CIE) criterion with a cut-off of more than 10% to detect the probable confounders of this association. Multiple logistic regression analysis was used to adjust for confounders identified by the CIE criterion. Both crude and adjusted ORs were reported for the association between diabetes and in-hospital outcomes with a 95% confidence interval (CI). Data analyses were performed by the statistical package STATA/IC version 14.2 (Stata Corp LP College Station, TX, USA).

We summarized continuous variables using means with standard deviations and categorical variables with percentages and 95% Wald Confidence Intervals (CI). To test the statistical significance of the difference between the intention to vaccinate of each categorical variable, the χ2 test was applied. T-test was used when testing continuous variables. Backward stepwise logistic regression analysis was used to study the effect of independent variables on the intention to the HPV vaccine. A p-value <0.2 was required for retention of each variable in the final model. Multicollinearity of independent variables was checked by the variance inflation factor (VIF) statistic. Results are presented as crude (ORc) or adjusted odds ratio (aOR) and 95% confidence intervals (CI). The Area Under the Curve (AUC) and Hosmer-Lemeshow goodness of fit tests were considered as discrimination and calibration parameters for the model. All statistical analyses were performed using the statistical software Stata IC, version 14.2 (Stata Corp., College Station, TX, USA). Due to the complex sampling design, we used the svyset command to account for sample stratification and weighting to provide unbiased estimates of the population parameters. Statistical significance was defined as a two-sided p-value <0.05 for all tests.