Laboratory epidemiology of respiratory viruses in a large children's hospital: A STROBE-compliant article

Ye, Sheng; Wang, Tianlin

All samples were divided into 2 groups by the results of virus test (positive vs negative). Numeric variables were expressed with mean and standard deviations, and were compared between groups by using t test. Categorical variables were expressed as the number and proportion, and were compared by using Chi-square test or Fisher's exact test as appropriate. [13] The R package CBCgrps was employed for statistical description and bivariate inference. [14] Within each calendar month, the number of positive and negative samples was displayed in a barplot. The percentages of positive samples were reported in the barplot for each respiratory virus. To investigate variables associated with positive results, multivariable Logistic regression model including all potential factors (age, gender, season, source of sample, type of sample) was employed. [15, 16] Due to limited number of covariates and the large number of observations, the full model was not reduced (low risk of over fitting). Multivariable Logistic regression model was built upon complete case analysis that observations with missing values on covariates were excluded. [17] Two-tailed P value <.05 was considered as statistical significant. All statistical analyses were performed using R software (version 3.3.2).