Combining several bivariable multilevel model results into one display table

Question

I am conducting an analysis in R markdown where I want to look at the relationship between haemoglobin (outcome) and several other variables that may influence it (exposures e.g. season and ses).

I want to start by creating several models with haemoglobin as the outcome and looking at each exposure variable individually.

I have multilevel data - I have more than one observation per person, so I am using multilevel modelling using the glmmTMB package.

Here is an example of some of my data:

data = data.frame(id = c(1,1,1,2,2,2,3,3,3), 
                 ses = c(0.13, 1.23, -0.78, 1.32, 0.56, -0.04, -1.43, 1.45, 2.01),
                 season = c("good", "good", "bad", "bad", "bad", "good", "good", "good", "good" ),
                 haemoglobin = c(15, 14, 16, 9, 10, 11, 12, 10, 11))

# id   ses season haemoglobin
# 1  1  0.13   good          15
# 2  1  1.23   good          14
# 3  1 -0.78    bad          16
# 4  2  1.32    bad           9
# 5  2  0.56    bad          10
# 6  2 -0.04   good          11
# 7  3 -1.43   good          12
# 8  3  1.45   good          10
# 9  3  2.01   good          11

Here is an example of two models I will use:

library(glmmTMB)
mod1 <- glmmTMB(haemoglobin ~ season + (1 | id), data = data)
summary(mod1)
confint(mod1)
mod2 <- glmmTMB(haemoglobin ~ ses + (1 | id), data = data)
summary(mod2)
confint(mod2)

My question is how to present my data in a table without having to enter it all manually (ie in case my estimates change, so I don't have to enter it all again!) Perhaps I would be able to do it using kable but I'm not sure how.

I want a table that looks something like this (below). It doesn't have to be the exact same, just that allows me to display the results of several bivariable models together.

table = data.frame(Variable = c("seasonbad","seasongood","ses"), 
                  Estimate = c("ref", 0.05188, -0.650),
                  Confint = c("-", "-1.62, 1.72", "-1.01, 0.29"),
                  Pval = c("-", 0.951, 0.000442))
# Variable Estimate     Confint     Pval
# 1  seasonbad      ref           -        -
# 2 seasongood  0.05188 -1.62, 1.72    0.951
# 3        ses    -0.65 -1.01, 0.29 0.000442

Any help/package suggestions would be incredibly appreciated. Many thanks!

Answer 1

This is more a data manipulation question rather than a question about html/markdown. You need to get the data you want from the mod objects, once you have that the rest is easy. Here's an example of what I mean:

# define a function to extract the info and returns it in the shape you need
get_summary <- function(mod){
  mod_summary <- summary(mod)
  mod_conf <- confint(mod)
  
  var <- dimnames(mod_summary$coefficients$cond)[[1]][-1]
  res_mod <- mod_conf[paste0("cond.", var), 1:3] %>% t() %>% data.frame()
  res_mod["Pval"] <- mod_summary$coefficients$cond[var, 4]
  res_mod["Variable"] <- var
  return(res_mod)
}

# get the data from each model and bind the rows of the results together
rbind(get_summary(mod1), get_summary(mod2)) %>%
  # show the result as an HTML table
  tableHTML(rownames = FALSE, round = 2,widths = rep(100, 5)) %>% 
  add_theme("rshiny-blue")

you should get something like this: