所用r包：meta

待更新

单组连续型统计量

单组率

# 导入excel为列表
sheets <- getSheetNames("7d数据分析提取表.xlsx")
data_frame <- lapply(sheets, read.xlsx, xlsxFile="7d数据分析提取表.xlsx")

# assigning names to data frame
names(data_frame) <- sheets



#选择效应量的合并方法
rate <- transform(data_frame[[4]],
                  p = x/n,
                  log=log(x/n),
                  logit=log((x/n)/(1-x/n)),
                  arcsin.size=asin(sqrt(x/(n+1))),
                  darcsin=0.5*(asin(sqrt(x/(n+1)))+asin((sqrt(x+1)/(n+1)))))



shapiro.test(exa1$p)         # sm="PRAW"    没有转换的原始率
shapiro.test(exa1$log)       # sm="PLN"     对数转换
shapiro.test(exa1$logit)     # sm="PLOGIT"  logit变换
shapiro.test(exa1$arcsin)    # SM="PAS"     反正弦
shapiro.test(exa1$darcsin)   # SM="PFT"     双重反正弦







metaresults1 <- list()
#批量求meta分析结果
for(sheet_name in names(data_frame)){
  metaresults1[[sheet_name]] <- metaprop(x, n, studlab=paste(作者, year), data = data_frame[[sheet_name]], sm="PAS", incr = 0.5, allincr = FALSE, addinc = FALSE)
}



#在桌面创建对应的文件夹
desktop_path <- file.path(Sys.getenv("USERPROFILE"), "Desktop")
filename <- "7dmeta"
dir.create(paste0(desktop_path,"/",filename))
foldernames <- c("森林图_单组率","异质性图_单组率","漏斗图_单组率")
for (i in 1:3){
  dir.create(paste0(desktop_path,"/",filename,"/",foldernames[i]))
}


#批量绘图并导出为jpg格式到对应文件夹
# 森林图
for (sheet_name in names(data_frame)) {
  jpeg(paste0(desktop_path,"/",filename,"/",foldernames[1],"/",sheet_name, ".jpg"), width = 800, height = 600, units = "px", quality = 100)
  forest(metaresults1[[sheet_name]], digits = 2)
  dev.off()
}
# 异质性图
for (sheet_name in names(data_frame)) {
  jpeg(paste0(desktop_path,"/",filename,"/",foldernames[2],"/",sheet_name, ".jpg"), width = 800, height = 600, units = "px", quality = 100)
  radial(metaresults1[[sheet_name]], level = 0.95, text =data_frame[[sheet_name]]$zuozhe)
  dev.off()
}
# 漏斗图
for (sheet_name in names(data_frame)) {
  jpeg(paste0(desktop_path,"/",filename,"/",foldernames[3],"/",sheet_name, ".jpg"), width = 800, height = 600, units = "px", quality = 100)
  funnel(metaresults1[[sheet_name]])
  dev.off()
}


# 发表偏倚检验
metabias(meta1, method = "linreg", plotit = TRUE, k.min = 5)  # Egger检验
metabias(meta1, method = "peters", plotit = TRUE, k.min = 5)
# Method 参数包括linreg代表egger检验；peters代表peter检验
# plotit 为画图函数；
# k．min 为进行检验时所需最小的单个研究的数量，默认为 10，如果研究在3～10个之间可通过此函数进行调整.
#批量化
egger_results <- lapply(metaresults2, function(x) {
  metabias(x, method = "linreg", plotit = FALSE,k.min=1)  # 不画图，减少计算量
})
# 显示Egger检验的结果
egger_results

#合并结果
egger_df <- do.call(rbind, lapply(egger_results, function(res) {
  if (!is.null(res$p.value) && !is.null(res$statistic)) {
    data.frame(
      p_value = res$p.value, 
      test_statistic = res$statistic, 
      method = "Egger"
    )
  } else {
    NULL
  }
}))

write.xlsx(egger_df,"14天egger检验.xlsx")

#敏感性分析
metainf(meta1)        # 行敏感性分析，计算分别剔除每个入选研究后合并的 OR 值及 95％可信区间。
#批量化
sensitivity_results <- lapply(metaresults2, function(x) {
  metainf(x)  # 逐一排除分析
})

# 显示逐一排除敏感性分析的结果
sensitivity_results
#导出结果
wb <- createWorkbook()
for (i in seq_along(sensitivity_results)){
  addWorksheet(wb, sheetName = names(sensitivity_results)[i])
  writeData(wb, sheet = names(sensitivity_results)[i], x = as_tibble(sensitivity_results[[i]]))}
saveWorkbook(wb, "14天敏感性分析表格.xlsx", overwrite = TRUE)


#剔除最大异质性点的敏感性分析结果
summary_sensitivity <- tibble()
for (i in seq_along(metaresults2)){
  meta_obj <- metaresults2[[i]]
  std_residuals <- (meta_obj$TE - meta_obj$TE.random) / sqrt(meta_obj$seTE^2 + meta_obj$tau^2)
  high_residual_indices <- which.max(abs(std_residuals)) 
  sensitivity_result <- as_tibble(sensitivity_results[[i]])[high_residual_indices,]
  sensitivity_result$指标 <- names(metaresults2)[i]
  summary_sensitivity <- bind_rows(summary_sensitivity,sensitivity_result)
}
write.xlsx(summary_sensitivity,"14天去除异质性最大文章结果.xlsx")

OR/RR