\u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\u3068\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u3092\u4e26\u3079\u3066\u6bd4\u8f03<\/li>\n<\/ol>\n\n\n\n# \u6b20\u640d\u5024\u3092\u542b\u3080\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u6e96\u5099\u3059\u308b\n# ph.ecog01, meal.cal, wt.loss\uff1a\u6b20\u640d\u5024\u542b\u3080\ndf0 <- df[,c(\"time\", \"status01\", \"ph.ecog01\", \"age\", \"sex\",\n \"meal.cal\", \"wt.loss\")]\n# mice \u30d1\u30c3\u30b1\u30fc\u30b8\u306e mice \u95a2\u6570\u3067\u591a\u91cd\u4ee3\u5165\u30c7\u30fc\u30bf\u3092\u4f5c\u6210\u3059\u308b\n# (\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u6570 m=5)\uff1a\u672c\u5f53\u306f\u5c11\u306a\u304f\u3068\u3082 m=100 \u304c\u3088\u3044\nlibrary(mice)\ndf_imputed <- mice(df0, seed=0)\ndf_imputed\n# MatchThem \u30d1\u30c3\u30b1\u30fc\u30b8\u306e weightthem \u95a2\u6570\u3067 IPTW \u3092\u4f5c\u6210\u3059\u308b\nlibrary(MatchThem)\nweighted_imputed <- weightthem(ph.ecog01 ~ age + factor(sex) +\n meal.cal + wt.loss, data=df_imputed)\nsummary(weighted_imputed)\n# \u591a\u91cd\u4ee3\u5165\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u4f7f\u3063\u305f\u66ab\u5b9a\u7684\u306a IPTW Cox \u56de\u5e30\u3092\u884c\u3046\n#\uff08\u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\uff09\nlibrary(survival)\nwi_iptw_cox <- with(weighted_imputed,\n coxph(Surv(time, status01) ~\n ph.ecog01 + age + factor(sex) +\n meal.cal + wt.loss,\n weights=weights,\n robust=TRUE\n ))\n## \u7d71\u5408\npooled_wi_iptw_cox <- pool(wi_iptw_cox)\nsummary(pooled_wi_iptw_cox, conf.int=TRUE, exponentiate = TRUE)\n# \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u8a08\u7b97\u306e\u305f\u3081\u306b\u3001\n# \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3092\u7528\u3044\u305f Cox \u56de\u5e30\u306e\u95a2\u6570\u3092\u5b9a\u7fa9\u3059\u308b\n# cox_reg \u95a2\u6570: \u7279\u5b9a\u306e\u5b8c\u5168\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306b\u5bfe\u3057\u3066\n# Cox\u6bd4\u4f8b\u30cf\u30b6\u30fc\u30c9\u30e2\u30c7\u30eb\u3092\u9069\u7528\u3057\u3001\u56de\u5e30\u4fc2\u6570\u3092\u8fd4\u3059\u95a2\u6570\ncox_reg <- function(data, indices)\n # \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u6a19\u672c\u306e\u62bd\u51fa\n data <- data[indices,]\n# Cox\u6bd4\u4f8b\u30cf\u30b6\u30fc\u30c9\u30e2\u30c7\u30eb\u306e\u9069\u7528\nfit <- with(data,\n coxph(Surv(time, status01) ~\n ph.ecog01 + age + factor(sex) +\n meal.cal + wt.loss, weights=weights,\n robust=TRUE\n ))\n# \u30cf\u30b6\u30fc\u30c9\u6bd4\u306e\u63a8\u5b9a\u5024\u3092\u8fd4\u3059\nreturn(coef(fit))\n}\n# \u591a\u91cd\u4ee3\u5165\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306e\u30aa\u30d6\u30b8\u30a7\u30af\u30c8\u540d\u3092\u6e96\u5099\u3059\u308b\n# imps \u30d9\u30af\u30c8\u30eb: \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u306e\u5404\u8a66\u884c\u3067\u4f7f\u7528\u3059\u308b\n# \u5b8c\u5168\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306e\u8b58\u5225\u5b50(m=1~5 \u307e\u3067)\nimps <- paste0(\"imp\", 1:weighted_imputed$object$m)\n# \u591a\u91cd\u4ee3\u5165\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u4e00\u3064\u4e00\u3064\u306b\u5bfe\u3057\u3066\u3001\n# boot \u30d1\u30c3\u30b1\u30fc\u30b8\u306e boot \u95a2\u6570\u306b\u3088\u3063\u3066\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u6cd5\u3092\u5b9f\u884c\u3059\u308b\nlibrary(boot)\nfor (imp_name in imps){\n # \u5b8c\u5168\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306e\u4f5c\u6210\n imp <- complete(weighted_imputed,\n action = as.numeric(substr(imp_name, 4, 6)))\n result <- cox_reg(data=imp, indices=1:nrow(df0))\n set.seed(0)\n boot_res <- boot(\n data = imp,\n statistic = cox_reg,\n # \u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u56de\u6570\uff1a10000 \u306e\u307b\u3046\u304c\u3088\u3044\u304c\u53d6\u308a\u6025\u304e 1000\n R = 1000\n )\n print(boot_res)\n # \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u7d50\u679c\u3092\u4fdd\u5b58\n assign(paste(\"boot_res_\", imp_name, sep=\"\"), boot_res)\n}\n# \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u306e\u7d50\u679c\u3092\u683c\u7d0d\u3059\u308b\u30ea\u30b9\u30c8\u3092\u4f5c\u6210\u3059\u308b\nboot_res_lists <- lapply(paste0(\"boot_res_imp\",\n 1:weighted_imputed$object$m), get)\n# \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u7d50\u679c\u3092\u53d6\u308a\u51fa\u3059\n## \u63a8\u5b9a\u5024\u3068\u6a19\u6e96\u8aa4\u5dee\u306e\u30d9\u30af\u30c8\u30eb\u3092\u521d\u671f\u5316\nestimates <- c()\nvariances <- c()\n## for\u30eb\u30fc\u30d7\u3067\u5404\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u306e\u7d50\u679c\u3092\u51e6\u7406\nfor (i in 1:5) {\n m <- mean(boot_res_lists[[i]]$t[, 1])\n se <- sd(boot_res_lists[[i]]$t[, 1])\n estimates <- c(estimates, m)\n variances <- c(variances, se^2)\n}\n# \u63a8\u5b9a\u5024\u3068\u6a19\u6e96\u8aa4\u5dee\u3092\u5143\u306b\u7d71\u5408\u63a8\u5b9a\u5024\u3068\u4fe1\u983c\u533a\u9593\u306e\u8a08\u7b97\npool <- pool.scalar(Q = estimates, U = variances, n = nrow(df0))\n## \u30d7\u30fc\u30eb\u6a19\u6e96\u8aa4\u5dee\u3092\u8a08\u7b97\n## t1* (ph.ecog01) \u306e SE \u306e\u7d71\u5408\npooled_se <- sqrt(pool$ubar)\n## \u7d71\u5408\u63a8\u5b9a\u5024\u306e\u53d6\u308a\u51fa\u3057\npooled_m <- pool$qbar\n## 95 \uff05\u4fe1\u983c\u533a\u9593\u306e\u8a08\u7b97\nLL <- pooled_m - qnorm(0.975)*pooled_se\nUL <- pooled_m + qnorm(0.975)*pooled_se\n# \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u3092\u51fa\u529b\noutput <- c(pooled_m, LL, UL)\nnames(output) <- c(\"Estimate\", \"LowerLimit\", \"UpperLimit\")\n## \u5bfe\u6570\u30cf\u30b6\u30fc\u30c9\u6bd4\u3068\u4fe1\u983c\u533a\u9593\noutput\n## \u771f\u6570\u306b\u3057\u305f\u30cf\u30b6\u30fc\u30c9\u6bd4\u3068\u4fe1\u983c\u533a\u9593\nexp(output)\n# \u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\u3068\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u3092\u4e26\u3079\u3066\u6bd4\u8f03\n## \u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\nsummary(pooled_wi_iptw_cox, conf.int=TRUE,\n exponentiate=TRUE)[1,c(2,7,8)]\n## \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\nexp(output)<\/code><\/pre>\n\n\n\n\u7d50\u679c\u306f\u3001\u4ee5\u4e0b\u306e\u3088\u3046\u306b\u51fa\u529b\u3055\u308c\u308b<\/p>\n\n\n\n
\u66ab\u5b9a\u7684\u306a\u7d50\u679c\uff08\u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\uff09\u306f\u4ee5\u4e0b\u306e\u3068\u304a\u308a<\/p>\n\n\n\n![\"\"](\"https:\/\/best-biostatistics.com\/toukei-er\/wp-content\/uploads\/2024\/06\/20240609212649.png\")
<\/figure>\n\n\n\n<\/span><\/p>\n\n\n\n\u3053\u3053\u3067\u306f\u3001\u7740\u76ee\u3057\u3066\u3044\u308b ph.ecog01 \u306e\u4ea4\u7d61\u56e0\u5b50\u306e\u63a8\u5b9a\u5024\u3082\u8868\u793a\u3055\u308c\u3066\u3044\u308b<\/p>\n\n\n\n
IPTW \u3092\u4f5c\u6210\u3059\u308b\u305f\u3081\u306b\u4f7f\u7528\u3057\u305f\u4ea4\u7d61\u56e0\u5b50\u3092\u3001\u518d\u5ea6\u5171\u5909\u91cf\u3068\u3057\u3066\u6295\u5165\u3057\u3066\u3044\u308b<\/p>\n\n\n\n
\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u306e\u51fa\u529b\u306f\u4ee5\u4e0b\u306e\u3068\u304a\u308a<\/p>\n\n\n\n![\"\"](\"https:\/\/best-biostatistics.com\/toukei-er\/wp-content\/uploads\/2024\/06\/20240609213314.png\")
<\/figure>\n\n\n\n<\/span><\/p>\n\n\n\n\u30ed\u30d0\u30b9\u30c8\u4fe1\u983c\u533a\u9593\u3068\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u3092\u4e26\u3079\u3066\u307f\u308b\u3068\u3001\u5c11\u3057\u7570\u306a\u308b\u306e\u304c\u308f\u304b\u308b<\/p>\n\n\n\n![\"\"](\"https:\/\/best-biostatistics.com\/toukei-er\/wp-content\/uploads\/2024\/06\/20240609213335.png\")
<\/figure>\n\n\n\n<\/span><\/p>\n\n\n\n\u307e\u3068\u3081<\/h2>\n\n\n\n
\u3053\u306e\u8a18\u4e8b\u3067\u306f\u3001\u591a\u91cd\u4ee3\u5165 IPTW Cox \u56de\u5e30 \u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593 \u306e\u4e00\u9023\u306e\u8a08\u7b97\u3092 R \u3067\u306f\u3069\u306e\u3088\u3046\u306a\u30b9\u30af\u30ea\u30d7\u30c8\u3092\u66f8\u3051\u3070\u5b9f\u73fe\u3067\u304d\u308b\u304b\u306e\u4f8b\u3092\u6319\u3052\u305f<\/p>\n\n\n\n
\u4e00\u3064\u3067\u5b8c\u7d50\u3059\u308b\u95a2\u6570\u306f\u306a\u3044\u306e\u3067\u3001\u3044\u304f\u3064\u304b\u306e\u95a2\u6570\u3092\u7d44\u307f\u5408\u308f\u305b\u3066\u3001\u5b9f\u884c\u3059\u308b\u5fc5\u8981\u304c\u3042\u308b<\/p>\n\n\n\n
\u4f55\u3089\u304b\u53c2\u8003\u306b\u306a\u308c\u3070<\/p>\n\n\n\n
\u53c2\u8003\u6587\u732e<\/h3>\n\n\n\n
Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"\u8aac\u660e\u5909\u6570\u304c\u3044\u304f\u3064\u304b\u3042\u308b\u751f\u5b58\u6642\u9593\u30c7\u30fc\u30bf\u3067\u3001\u3044\u304f\u3064\u304b\u306e\u8aac\u660e\u5909\u6570\u306b\u6b20\u640d\u5024\u304c\u3042\u308a\u3001\u591a\u91cd\u4ee3\u5165\u3057\u3066 IPTW Cox \u56de\u5e30\u5206\u6790\u304c\u3057\u305f\u3044\u5834\u5408\u3001R \u3067\u3069\u306e\u3088\u3046\u306b\u3059\u308c\u3070\u3088\u3044\u304b IPTW Cox \u56de\u5e30\u306e\u5834\u5408\u3001\u30d6\u30fc\u30c8\u30b9\u30c8\u30e9\u30c3\u30d7\u4fe1\u983c\u533a\u9593\u304c\u3088\u308a\u9069 […]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"swell_btn_cv_data":"","_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[7,5,21,22,20],"tags":[],"class_list":["post-27","post","type-post","status-publish","format-standard","hentry","category-iptw-","category-r","category-21","category-22","category-20"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/27","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/comments?post=27"}],"version-history":[{"count":4,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/27\/revisions"}],"predecessor-version":[{"id":3539,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/27\/revisions\/3539"}],"wp:attachment":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/media?parent=27"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/categories?post=27"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/tags?post=27"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\"https:\/\/best-biostatistics.com\/toukei-er\/wp-content\/uploads\/2024\/08\/1920x1080-video-Excel-300x169.jpg\")
<\/figure><\/div>\t\t\t\t\t![\"\"](\"https:\/\/best-biostatistics.com\/wp\/wp-content\/uploads\/2023\/11\/bn_r_03.png\")
<\/a>\r\n