使用gtsummary实现Poisson回归风险人日及事件数汇总
完全可以通过gtsummary自带的tbl_custom_summary()函数实现你的需求,这个函数专门用于自定义分组自定义统计规则,刚好适配你要按变量分组求和的场景,完整实现代码如下:
library(tidyverse) library(gtsummary) # 示例数据集 df = tibble( study_arm = c("control", "intervention", "control", "intervention", "control", "intervention", "control", "intervention"), events = c(3,4,12,6,0,3,11,9), strata_group = c("A", "A", "A", "A", "B", "B", "B", "B"), days_at_risk = c(100,100,200,200,300,300,100,100) ) # 1. 生成分组汇总表:包含事件总数、总风险人日、发生率三列 tbl_summary_part <- tbl_custom_summary( data = df, include = c(study_arm, strata_group), # 自定义三个统计指标的计算规则 stats = list( n_events = ~sum(.x$events), total_days_at_risk = ~sum(.x$days_at_risk), rate = ~sum(.x$events)/sum(.x$days_at_risk) ), type = all_categorical() ~ "categorical" ) # 2. 生成Poisson回归结果表 tbl_reg_part <- glm( events ~ study_arm + strata_group, offset = log(days_at_risk), family=poisson(link = "log"), data = df ) %>% tbl_regression(exponentiate = T) # 3. 合并两张表得到最终结果 tbl_final <- tbl_merge( tbls = list(tbl_summary_part, tbl_reg_part), tab_spanner = c("分组统计", "Poisson回归结果") ) # 预览最终表格 tbl_final
上述代码生成的汇总表和你用dplyr手动计算的结果完全一致,可直接用于后续输出或导出。
内容的提问来源于stack exchange,提问作者ophelia
相关产品推荐
相关产品推荐

