Statistical Analysis
Baseline characteristics and clinical data have been summarized using
the appropriate summary statistics. Variables on a continuous scale were
described as mean, standard deviation, median, and interquartile range.
Variables on a categorical scale were presented as count and percentage.
All percentages have been presented on the number of data points
recorded. The complications were reported as rates amongst the cohort.
Continuous variables were analyzed using the 2-sample t -test or
Mann-Whitney test for parametric and nonparametric variables,
respectively. The Chi-square or Fisher exact test was used for
categorical variables. To allow an unbiased comparison between SDD and
ONS, a propensity score analysis was employed The derived propensity
scores with separate multivariable models (including age, sex,
CHA2DS2-VASc
score, hypertension, and BMI) were calculated when p <0.10 was
achieved in the univariate analysis. Propensity scores were then matched
with a caliper range of ± 0.1 to obtain matched pairs of patients. The
propensity score-based greedy matching algorithm successfully matched
all patients who were sent home on the same day with patients who stayed
overnight. XLSTAT 19.4 (Addinsoft, New York, NY) was used to perform
statistical analyses. Statistical tests were based on a 2-sided
significance level of 0.05.