Stata 11.2 se

All data are given as mean. Data were analyzed for normality and equal variance across groups. As no normality for the here presented results could be detected, differences between groups were assessed using one-way ANOVA by the Kruskal-Wallis comparison followed by the appropriate post hoc comparison test (Wilcoxon-Mann-Whitney). The overall statistical significance was set at p < 0.05. Statistics were performed using the software package Stata 11.2 SE (StataCorp LLC, College Station, TX, USA).

Statistical analysis was performed using Stata 11.2/SE (Texas, USA). Statistical significance was assumed where p<0.05. Descriptive statistics were presented separately for individuals with and without pneumonia. The Chi-squared test was used for categorical variables and the t-test for continuous variables. Stroke outcomes (inpatient mortality and long LOS) and poststroke complications (namely sepsis, respiratory failure and convulsions) were assessed using four logistic regression models. Patients without SAP were the reference category. These models were as follows: 1. unadjusted; 2. adjusted for age and sex; 3. adjusted for age, sex and stroke sub-type; 4. adjusted for age, sex, stroke sub-type and co-morbidities.
The models were repeated after stratifying by stroke subtype (hemorrhage, ischemic and ischemic stroke of unidentified cause). Patients with non-aspiration pneumonia were the reference category in the sub-analysis comparing outcomes in aspiration vs non-aspiration pneumonia. The logistic regression models used in the sub-analysis were identical to those described above. To investigate for any change in trend for mortality rate over the study period, the overall rate of death through time and the log odds of death in patients without pneumonia were plotted and illustrated for the period of 2004-2012 (the years for which full data was available).