meta-analysis of proportions: subgroup analysis with missing data

Viewed 8

How can I simply arrange a subgroup a analysis of my meta-analysis of proportions ? In the example below I don't have the mean age in the study 5. If I perform a subgroup analysis, it failed due to this missing value. I know how to build a new dataframe without the study 5 and re run everything but, does someone has a trick to do this more simply ? I tried inserting na.rm = TRUE or na.omit = TRUE but it doesn't work. I have the same problem with the metareg function of the package meta.

Please find below a minimal working example:

require (meta)
study <- c("study1", "study2", "study3", "study4", "study5", "study6", "study7", "study8", "study9", "study10")
n_patients <- c(432, 255, 214, 295, 440, 509, 599, 289, 476, 269)
n_infections <- c(40, 18, 18, 40, 40, 25, 7, 15, 15, 5)
age <- c(58, 59.1, 60.50, 63.25, NA, 60, 64.15, 55.25, 62.5, 70)

# build the dataframe
my_data <- data.frame (study, n_patients, n_infections, age)

##### Compute the meta-analysis
m1 <- metaprop (event = n_infections, 
                n = n_patients, 
                studlab = study, 
                data = my_data, 
                method = "GLMM",  
                sm = "PLOGIT",    
                random = TRUE,
                hakn = TRUE,
                prediction = TRUE)

# subgroup analysis by age
meta_age <- update (m1,
                    subgroup = my_data $ age,
                    tau.common = TRUE)
Error: Missing values in argument 'subgroup'.

Thank you in advance for your help.

Charles.

0 Answers
Related