tab1_re takes lists of continuous and/or categorical variables and returns Median (spread) for continuous variables and N (%) for categorical variables. Produces a table with both an overall column and columns by another variable. For a binary by variable only, it produces p-values from a random effects model.

tab1_re(contvars, catvars, byvar, re, dat, col = TRUE,
  spread = "range")

Arguments

contvars

is a list of the continuous variables you want in the rows e.g. list('Age'). Can be NULL.

catvars

is a list of the categorical variables you want in the rows e.g. list('Gender','Race'). Can be NULL.

byvar

is the categorical variable you want to tabulate by across the columns (needs to be in quotes). MUST BE 0/1 since it will be used as the outcome variable in glmer.

re

the name of the random effect variable, should be supplied in quotes, e.g. re = "ptid"

dat

is the dataset to use for analysis

col

takes the value TRUE or FALSE indicating whether you want column percent (TRUE, default) or row percent (FALSE)

spread

takes the value "range" or "iqr" indicating whether you want (min, max) or (Q1, Q3) in summaries of continuous variables. Defaults to "range".

Value

Returns a dataframe. If there are warnings or errors from glmer then NA is returned in place of the p-value.