Skip to contents

Manually fit and pool Cox proportional hazards model results from multiple imputed datasets

Source: R/cox_pooling.R
cox_pooling.Rd

This function manually fits and pools results from Cox proportional hazards models using a list of imputed datasets. It leverages the mice::as.mira and mice::pool functions to ensure proper pooling of results across multiple imputations using Rubin's rules.

Usage

cox_pooling(
  x,
  surv_formula = stats::as.formula(survival::Surv(fu_itt_months, death_itt) ~ treat)
)

Arguments

x

A list of imputed datasets with weights or raking weights (e.g., from raking_weights) and optional cluster information (for matched datasets).

surv_formula

A formula for the Cox proportional hazards model (default is Surv(fu_itt_months, death_itt) ~ treat).

Value

A data frame containing the pooled results, including hazard ratios, confidence intervals, and p-values.

Details

The function requires a list of imputed data frames with weights and optional cluster (matching) information, as well as a formula for the Cox proportional hazards model. The data frames must include a column named weights, and optionally a column named subclass (for matched datasets to indicate cluster membership).

This function is particularly useful when working with imputed datasets that are not in the form of mimids or wimids objects, such as when intermediate steps like raking weights (via raking_weights) are applied. It provides a flexible way to fit and pool Cox models while ensuring compatibility with Rubin's rules for multiple imputation.

The function follows these steps:

  1. Fit a Cox proportional hazards model to each imputed dataset. If a subclass column is present, it is used as a cluster variable for matched pairs.

  2. Convert the list of fitted models into a mira object using mice::as.mira.

  3. Pool the results using mice::pool, which applies Rubin's rules for combining estimates and variances across imputations.

  4. Format the pooled results, including exponentiating the hazard ratios and calculating confidence intervals.

See also

coxph, pool, as.mira, matchthem, weightthem

Examples

library(encore.analytics)
library(mice)
#> 
#> Attaching package: ‘mice’
#> The following object is masked from ‘package:stats’:
#> 
#>     filter
#> The following objects are masked from ‘package:base’:
#> 
#>     cbind, rbind
library(MatchThem)
#> 
#> Attaching package: ‘MatchThem’
#> The following objects are masked from ‘package:mice’:
#> 
#>     cbind, pool
#> The following object is masked from ‘package:base’:
#> 
#>     cbind

# Simulate a cohort with 500 patients and 20% missing data
data <- simulate_data(
  n = 500,
  imposeNA = TRUE,
  propNA = 0.2
)

# Impute the data
set.seed(42)
mids <- mice(data, m = 5, print = FALSE)
#> Warning: Number of logged events: 765

# Fit a propensity score model
fit <- as.formula(treat ~ dem_age_index_cont + dem_sex_cont + c_smoking_history)

# Weight patients within each imputed dataset
wimids <- weightthem(
  formula = fit,
  datasets = mids,
  approach = "within",
  method = "glm",
  estimand = "ATO"
)
#> Estimating weights     | dataset: #1
#>  #2
#>  #3
#>  #4
#>  #5
#> 

# Create a list of imputed and weighted datasets
wimids_list <- MatchThem::complete(wimids, action = "all", all = FALSE, include = FALSE)

# Define a survival model formula
cox_fit <- as.formula(survival::Surv(fu_itt_months, death_itt) ~ treat)

# Fit and pool Cox proportional hazards model results
cox_pooling(wimids_list, surv_formula = cox_fit)
#>    term  estimate  std.error statistic     p.value  conf.low conf.high
#> 1 treat 0.7297697 0.08792742 -3.582799 0.000374082 0.6139819 0.8673933
#>              b       df dfcom       fmi      lambda m         riv        ubar
#> 1 4.570604e-05 487.4987   496 0.0111428 0.007094245 5 0.007144933 0.007676384