{"id":430,"date":"2019-08-15T22:13:43","date_gmt":"2019-08-15T13:13:43","guid":{"rendered":"https:\/\/best-biostatistics.com\/toukei-er\/entry\/competing-risk-regression-in-r\/"},"modified":"2024-10-07T12:28:00","modified_gmt":"2024-10-07T03:28:00","slug":"competing-risk-regression-in-r","status":"publish","type":"post","link":"https:\/\/best-biostatistics.com\/toukei-er\/entry\/competing-risk-regression-in-r\/","title":{"rendered":"R \u3067\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u3092\u5b9f\u884c\u3059\u308b\u65b9\u6cd5"},"content":{"rendered":"\n

\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u3068\u306f\u3001\u5171\u5909\u91cf\u8abf\u6574\u3092\u3057\u305f\u7af6\u5408\u30ea\u30b9\u30af\u5206\u6790\u306e\u65b9\u6cd5\u3002<\/p>\n\n\n\n\n\n\n\n

\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u306e\u524d\u306b\u7af6\u5408\u30ea\u30b9\u30af\u3068\u306f\uff1f<\/h2>\n\n\n\n

\u7af6\u5408\u30ea\u30b9\u30af\u306b\u3064\u3044\u3066\u306f\u3001\u4ee5\u4e0b\u3092\u53c2\u7167\u3002<\/p>\n\n\n

\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\u3042\u308f\u305b\u3066\u8aad\u307f\u305f\u3044<\/span>\n\t\t\t\t\t
\"\"<\/figure><\/div>\t\t\t\t\t
\n\t\t\t\t\t\tR \u3067\u7af6\u5408\u30ea\u30b9\u30af\u5206\u6790 Gray \u691c\u5b9a\u3092\u884c\u3046\u65b9\u6cd5<\/a>\n\t\t\t\t\t\t\u7af6\u5408\u30ea\u30b9\u30af\u3068\u306f\u4f55\u304b\uff1f Gray \u691c\u5b9a\u306e\u5b9f\u884c\u65b9\u6cd5 \u7af6\u5408\u30ea\u30b9\u30af\u3068\u306f\uff1f \u518d\u767a\u304c\u30a8\u30f3\u30c9\u30dd\u30a4\u30f3\u30c8\u3067\u3042\u3063\u305f\u304c\u3001\u518d\u767a\u3059\u308b\u524d\u306b\u6b7b\u4ea1\u3057\u3066\u3057\u307e\u3063\u305f\u306e\u3067\u3001\u89b3\u5bdf\u3067\u304d\u306a\u304b\u3063\u305f\u3002 \u8133\u6897\u585e\u306e\u767a\u73fe\u304c…<\/span>\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\n\n

\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u306e\u7a2e\u985e<\/h2>\n\n\n\n

\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u30e2\u30c7\u30eb\u306b\u306f\u56db\u3064\u8003\u3048\u3089\u308c\u308b\u3002<\/p>\n\n\n\n

    \n
  1. \u7d76\u5bfe\u30ea\u30b9\u30af\u56de\u5e30 Absolute Risk Regression<\/li>\n\n\n\n
  2. \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u30ea\u30b9\u30af\u56de\u5e30 Logistic Risk Regression<\/li>\n\n\n\n
  3. Fine-Gray \u56de\u5e30 Fine and Gray Regression \uff1d\u72ed\u7fa9\u306e\uff08\u3044\u308f\u3086\u308b\uff09\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30 Competing Risk Regression<\/li>\n\n\n\n
  4. \u539f\u56e0\u5225 Cox \u56de\u5e30 Cause-Specific Cox Regression<\/li>\n<\/ol>\n\n\n\n

    \u304a\u3059\u3059\u3081\u306f\uff11\u756a\u306e\u7d76\u5bfe\u30ea\u30b9\u30af\u56de\u5e30\u3002<\/p>\n\n\n\n

    \u5272\u5408\u306e\u56de\u5e30\u5206\u6790\u306a\u306e\u3067\u3001\u7d50\u679c\u306e\u610f\u5473\u3042\u3044\u304c\u308f\u304b\u308a\u3084\u3059\u3044\u3002<\/p>\n\n\n\n

    Link function\u306f\u30ed\u30b0 <\/p>\n\n\n\n

    $$ \\log p $$<\/p>\n\n\n\n

    \uff12\u756a\u76ee\u306e\u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u30ea\u30b9\u30af\u56de\u5e30\u306f\u3001\u30aa\u30c3\u30ba\u6bd4\u306e\u56de\u5e30\u5206\u6790\u3002<\/p>\n\n\n\n

    Link function\u306f\u30ed\u30b8\u30c3\u30c8 <\/p>\n\n\n\n

    $$ \\log \\frac{p}{1 – p} $$<\/p>\n\n\n\n

    \u4e00\u756a\u4e16\u9593\u306b\u77e5\u3089\u308c\u3066\u3044\u308b\u306e\u306f\uff13\u756a\u306e\u3044\u308f\u3086\u308b\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u306e Fine-Gray \u56de\u5e30\u3060\u304c\u3001\u56de\u5e30\u4fc2\u6570\u306e\u89e3\u91c8\u304c\u96e3\u3057\u3044\u306e\u304c\u6b20\u70b9\u3002<\/p>\n\n\n\n

    Link function\u306f complementary log-log<\/p>\n\n\n\n

    $$ \\log{(- \\log{p})} $$<\/p>\n\n\n\n

    \uff14\u756a\u76ee\u306e\u539f\u56e0\u5225 Cox \u56de\u5e30\u306f\u3001\u30a8\u30f3\u30c9\u30dd\u30a4\u30f3\u30c8\u3054\u3068\u306b Cox \u56de\u5e30\u3092\u884c\u3046\u3082\u306e\u3067\u3001\u3084\u3063\u3066\u3044\u308b\u3053\u3068\u306f\u308f\u304b\u308a\u3084\u3059\u3044\u304c\u3001\u7af6\u5408\u30ea\u30b9\u30af\u3092\u540c\u6642\u306b\u6271\u3044\u305f\u3044\u3068\u3044\u3046\u5e0c\u671b\u3092\u304b\u306a\u3048\u3066\u306f\u3044\u306a\u3044\u3002<\/p>\n\n\n\n

    \u53c2\u8003\u6587\u732e\u306f\u3053\u3061\u3089\u3002<\/p>\n\n\n\n

    https:\/\/www.ncbi.nlm.nih.gov\/pmc\/articles\/PMC4547456<\/a><\/p>\n\n\n\n\n\n\n\n

    \uff1e\uff1e\u3082\u3046\u7d71\u8a08\u3067\u60a9\u3080\u306e\u306f\u7d42\u308f\u308a\u306b\u3057\u307e\u305b\u3093\u304b\uff1f\u00a0<\/a><\/strong><\/span><\/p>\r\n\"\"<\/a>\r\n

    \u21911\u4e07\u4eba\u4ee5\u4e0a\u306e\u533b\u7642\u5f93\u4e8b\u8005\u304c\u8cfc\u8aad\u4e2d<\/span><\/strong><\/span><\/p><\/div>

    \u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u5206\u6790\u306e\u6e96\u5099<\/h2>\n\n\n\n

    R\u3067\u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30\u5206\u6790\u3092\u884c\u3046\u306b\u306f\u3044\u304f\u3064\u304b\u306e\u30d1\u30c3\u30b1\u30fc\u30b8\u3092\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3059\u308b\u5fc5\u8981\u304c\u3042\u308b\u3002<\/p>\n\n\n\n

    \u4ee5\u4e0b\u306e\u30b9\u30af\u30ea\u30d7\u30c8\u3092\u4f5c\u52d5\u3055\u305b\u308b\u306b\u306f\u4ee5\u4e0b\u306e\u30d1\u30c3\u30b1\u30fc\u30b8\u3092\u3042\u3089\u304b\u3058\u3081\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3057\u3066\u304a\u304f\u3002<\/p>\n\n\n\n

      \n
    • riskRegression ( ARR, LRR, CSC )<\/li>\n\n\n\n
    • cmprsk ( crr )<\/li>\n\n\n\n
    • prodlim ( Hist )<\/li>\n\n\n\n
    • timereg ( comp.risk )<\/li>\n<\/ul>\n\n\n\n

      ( ) \u5185\u306f\u3001\u4ee5\u4e0b\u3067\u767b\u5834\u3059\u308b\u3001\u5404\u30d1\u30c3\u30b1\u30fc\u30b8\u306b\u542b\u307e\u308c\u308b\u95a2\u6570\u540d\u3002<\/p>\n\n\n\n

      \u307e\u305f\u3001Melanoma \u3068\u3044\u3046\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306f\u3001riskRegression \u30d1\u30c3\u30b1\u30fc\u30b8\u306b\u542b\u307e\u308c\u308b\u30b5\u30f3\u30d7\u30eb\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3002<\/p>\n\n\n\n

      \u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u65b9\u6cd5\u306f\u4ee5\u4e0b\u306e\u3068\u304a\u308a<\/p>\n\n\n\n

      install.packages<\/span>(<\/span>\"riskRegression\"<\/span>)<\/span>\n<\/code><\/pre>\n\n\n\n
      \u3068 R \u30b3\u30f3\u30bd\u30fc\u30eb\u306b\u66f8\u3044\u3066\u30a8\u30f3\u30bf\u30fc\u3057\u3001Japan \u306e Mirror Server \u3092\u9078\u3093\u3067 OK \u3092\u30af\u30ea\u30c3\u30af\u3002<\/p>\n\n\n\n
      \u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u304c\u6e08\u3093\u3060\u3089\u3001library()\u3067\u547c\u3073\u51fa\u3057\u3066\u304a\u304f\u3002\u30c7\u30d5\u30a9\u30eb\u30c8\u3067\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3055\u308c\u3066\u3044\u308bsurvival\u3082\u547c\u3073\u51fa\u3057\u3066\u304a\u304f\u3002<\/p>\n\n\n\n
      library<\/span>(<\/span>riskRegression)<\/span>\nlibrary<\/span>(<\/span>cmprsk)<\/span>\nlibrary<\/span>(<\/span>prodlim)<\/span>\nlibrary<\/span>(<\/span>timereg)<\/span>\nlibrary<\/span>(<\/span>survival)<\/span>\n<\/code><\/pre>\n\n\n\n
      \u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30 1\uff1a \u7d76\u5bfe\u30ea\u30b9\u30af\u56de\u5e30<\/h2>\n\n\n\n
      \u6d78\u6f64\u306e\u30ea\u30b9\u30af\u3092\u63a8\u5b9a\u3059\u308b\u305f\u3081\u3001\u5e74\u9f62\u3092\u5171\u5909\u91cf\u3068\u3057\u3066\u30e2\u30c7\u30eb\u3092\u4f5c\u6210\u3057\u89e3\u6790\u3059\u308b\u3002<\/p>\n\n\n\n
      ARR()\u306e\u4e2d\u306bEvent History\u5909\u6570\u3092\u4f5c\u308bHist()\u3092\u4f7f\u3046\u3002<\/p>\n\n\n\n
      status==1\uff08\u30e1\u30e9\u30ce\u30fc\u30de\u306b\u3088\u308b\u6b7b\u4ea1\u3002\u691c\u8a0e\u3057\u305f\u3044\u30a8\u30f3\u30c9\u30dd\u30a4\u30f3\u30c8\uff09\u306e\u30ea\u30b9\u30af\u3092\u8a08\u7b97\u3057\u3066\u3082\u3089\u3046\u305f\u3081\u306bcause=1\u3092\u3064\u3051\u308b\u3002<\/p>\n\n\n\n
      ARR1 <-<\/span> ARR<\/span>(<\/span>Hist<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>invasion+<\/span>age,<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>)<\/span>\nsummary<\/span>(<\/span>ARR1)<\/span>\n<\/code><\/pre>\n\n\n\n
      riskRegression()\u3092\u4f7f\u3046\u5834\u5408link=”relative”\u3068\u3059\u308b\u3002<\/p>\n\n\n\n
      \u7d50\u679c\u306f\u540c\u3058\u306b\u306a\u308b\u3002<\/p>\n\n\n\n
      rR1 <-<\/span> riskRegression<\/span>(<\/span>Hist<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>invasion+<\/span>age,<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>,<\/span>link=<\/span>\"relative\"<\/span>)<\/span>\nsummary<\/span>(<\/span>rR1)<\/span>\n<\/code><\/pre>\n\n\n\n
      ><\/span> summary<\/span>(<\/span>ARR1)<\/span>\nriskRegression:<\/span> Competing risks regression model\nIPCW estimation. The weights are based on\nthe Kaplan-<\/span>Meier estimate for<\/span> the censoring distribution.\nLink function<\/span>:<\/span> 'relative'<\/span> yielding absolute risk ratios,<\/span> see help<\/span>(<\/span>riskRegression)<\/span>.\n\uff08\u4e2d\u7565\uff09\nTime constant regression coefficients:<\/span>\nFactor    Coef exp<\/span>(<\/span>Coef)<\/span> StandardError       z         CI_95    Pvalue\ninvasionlevel.1   0.833<\/span>     2.301<\/span>         0.319<\/span>   2.612<\/span> [<\/span>1.231<\/span>;<\/span>4.300<\/span>]<\/span> 0.0089994<\/span>\ninvasionlevel.2   1.313<\/span>     3.717<\/span>         0.338<\/span>   3.880<\/span> [<\/span>1.915<\/span>;<\/span>7.213<\/span>]<\/span> 0.0001044<\/span>\nage 0.00440<\/span>   1.00441<\/span>       0.00795<\/span> 0.55335<\/span> [<\/span>0.989<\/span>;<\/span>1.020<\/span>]<\/span> 0.5800251<\/span>\nNote:<\/span> The values exp<\/span>(<\/span>Coef)<\/span> are absolute risk ratios\n<\/code><\/pre>\n\n\n\n
      \u7d50\u679c\u306f\u4e0a\u8a18\u306e\u901a\u308a\u3067\u3001\u6d78\u6f64\u30ec\u30d9\u30eb\u30bc\u30ed\u306b\u6bd4\u3079\uff11\u306f2.3\u500d\u306e\u30e1\u30e9\u30ce\u30fc\u30de\u6b7b\u4ea1\u306e\u30ea\u30b9\u30af\u30012\u306f3.7\u500d\u306e\u30ea\u30b9\u30af\u3068\u8a00\u3048\u308b\u3002<\/p>\n\n\n\n
      \u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30 2\uff1a \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u30ea\u30b9\u30af\u56de\u5e30<\/h2>\n\n\n\n
      \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u30ea\u30b9\u30af\u56de\u5e30\u306fLRR()\u3092\u4f7f\u3046\u3002<\/p>\n\n\n\n
      LRR1 <-<\/span> LRR<\/span>(<\/span>Hist<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>invasion+<\/span>age,<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>)<\/span>\nsummary<\/span>(<\/span>LRR1)<\/span>\n<\/code><\/pre>\n\n\n\n
      riskRegression()\u3092\u4f7f\u3046\u5834\u5408\u306flink=”logistic”\u3068\u3059\u308b\u3002\u4e0a\u8a18\u3068\u540c\u3058\u7d50\u679c\u306b\u306a\u308b\u3002<\/p>\n\n\n\n
      rR.L1 <-<\/span> riskRegression<\/span>(<\/span>Hist<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>invasion+<\/span>age,<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>,<\/span>link=<\/span>\"logistic\"<\/span>)<\/span>\nsummary<\/span>(<\/span>rR.L1)<\/span>\n<\/code><\/pre>\n\n\n\n
      ><\/span> summary<\/span>(<\/span>LRR1)<\/span>\nriskRegression:<\/span> Competing risks regression model\nIPCW estimation. The weights are based on\nthe Kaplan-<\/span>Meier estimate for<\/span> the censoring distribution.\nLink function<\/span>:<\/span> 'logistic'<\/span> yielding odds ratios,<\/span> see help<\/span>(<\/span>riskRegression)<\/span>.\n\uff08\u4e2d\u7565\uff09\nTime constant regression coefficients:<\/span>\nFactor    Coef exp<\/span>(<\/span>Coef)<\/span> StandardError       z          CI_95    Pvalue\ninvasionlevel.1   1.058<\/span>     2.881<\/span>         0.392<\/span>   2.701<\/span> [<\/span>1.337<\/span>;<\/span> 6.210<\/span>]<\/span> 0.0069102<\/span>\ninvasionlevel.2   1.818<\/span>     6.160<\/span>         0.477<\/span>   3.808<\/span> [<\/span>2.416<\/span>;<\/span>15.701<\/span>]<\/span> 0.0001401<\/span>\nage 0.00384<\/span>   1.00385<\/span>       0.01165<\/span> 0.32964<\/span> [<\/span>0.981<\/span>;<\/span> 1.027<\/span>]<\/span> 0.7416742<\/span>\nNote:<\/span> The values exp<\/span>(<\/span>Coef)<\/span> are odds ratios\n<\/code><\/pre>\n\n\n\n
      \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u30ea\u30b9\u30af\u56de\u5e30\u306e\u7d50\u679c\u3068\u3057\u3066\u306e\u30aa\u30c3\u30ba\u6bd4\u306f\u3001\u7d76\u5bfe\u30ea\u30b9\u30af\u56de\u5e30\u306e\u7d50\u679c\u306b\u6bd4\u3079\u308b\u3068\u5927\u304d\u304f\u306a\u308a\u3001\u6d78\u6f64\u30ec\u30d9\u30eb\uff11\u306f\u30bc\u30ed\u306b\u6bd4\u3079\u30aa\u30c3\u30ba\u6bd42.9\u30012\u306f\u30aa\u30c3\u30ba\u6bd46.2\u3068\u306a\u308b\u3002<\/p>\n\n\n\n
      \u7af6\u5408\u30ea\u30b9\u30af\u56de\u5e30 3\uff1a Fine-Gray \u56de\u5e30<\/h2>\n\n\n\n
      Fine-Gray\u56de\u5e30\u306fcomp.risk()\u3092\u4f7f\u3046\u3002Time to Event\u30c7\u30fc\u30bf\u306b\u5909\u63db\u3059\u308bEvent()\u3092\u4f7f\u3046\u3002<\/p>\n\n\n\n
      cr1 <-<\/span> comp.risk<\/span>(<\/span>Event<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>const<\/span>(<\/span>invasion)<\/span>+<\/span>const<\/span>(<\/span>age),<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>,<\/span>model=<\/span>\"prop\"<\/span>)<\/span>\nsummary<\/span>(<\/span>cr1)<\/span>\n<\/code><\/pre>\n\n\n\n
      riskRegression()\u3067\u3082link=”prop”\u3068\u3059\u308b\u3068\u540c\u3058\u8a08\u7b97\u304c\u3067\u304d\u308b\u3002<\/p>\n\n\n\n
      rR.FG1 <-<\/span> riskRegression<\/span>(<\/span>Hist<\/span>(<\/span>time,<\/span>status)<\/span>~<\/span>invasion+<\/span>age,<\/span>\ndata=<\/span>Melanoma,<\/span>cause=<\/span>1<\/span>,<\/span>link=<\/span>\"prop\"<\/span>)<\/span>\nsummary<\/span>(<\/span>rR.FG1)<\/span>\n<\/code><\/pre>\n\n\n\n
      ><\/span> summary<\/span>(<\/span>cr1)<\/span>\nCompeting risks Model\nNo test for<\/span> non-<\/span>parametric terms\nParametric terms :<\/span>\nCoef.      SE Robust SE     z    P-<\/span>val lower2.5%\nconst<\/span>(<\/span>invasion)<\/span>level.1 0.93900<\/span> 0.35300<\/span>   0.35300<\/span> 2.660<\/span> 0.007740<\/span>    0.2470<\/span>\nconst<\/span>(<\/span>invasion)<\/span>level.2 1.55000<\/span> 0.40100<\/span>   0.40100<\/span> 3.870<\/span> 0.000108<\/span>    0.7640<\/span>\nconst<\/span>(<\/span>age)<\/span>             0.00414<\/span> 0.00984<\/span>   0.00984<\/span> 0.421<\/span> 0.674000<\/span>   -<\/span>0.0151<\/span>\nupper97.5%\nconst<\/span>(<\/span>invasion)<\/span>level.1     1.6300<\/span>\nconst<\/span>(<\/span>invasion)<\/span>level.2     2.3400<\/span>\nconst<\/span>(<\/span>age)<\/span>                 0.0234<\/span>\n><\/span> summary<\/span>(<\/span>rR.FG1)<\/span>\nriskRegression:<\/span> Competing risks regression model\nIPCW estimation. The weights are based on\nthe Kaplan-<\/span>Meier estimate for<\/span> the censoring distribution.\nLink function<\/span>:<\/span> 'proportional'<\/span> yielding sub-<\/span>hazard ratios <\/span>(<\/span>Fine &<\/span> Gray 1999<\/span>),<\/span> see help<\/span>(<\/span>riskRegression)<\/span>.\n\uff08\u4e2d\u7565\uff09\nTime constant regression coefficients:<\/span>\nFactor    Coef exp<\/span>(<\/span>Coef)<\/span> StandardError       z          CI_95   Pvalue\ninvasionlevel.1   0.939<\/span>     2.558<\/span>         0.353<\/span>   2.663<\/span> [<\/span>1.282<\/span>;<\/span> 5.106<\/span>]<\/span> 0.007736<\/span>\ninvasionlevel.2   1.551<\/span>     4.716<\/span>         0.401<\/span>   3.872<\/span> [<\/span>2.151<\/span>;<\/span>10.339<\/span>]<\/span> 0.000108<\/span>\nage 0.00414<\/span>   1.00415<\/span>       0.00984<\/span> 0.42116<\/span> [<\/span>0.985<\/span>;<\/span> 1.024<\/span>]<\/span> 0.673635<\/span>\nNote:<\/span> The values exp<\/span>(<\/span>Coef)<\/span> are sub-<\/span>hazard ratios <\/span>(<\/span>Fine &<\/span> Gray 1999<\/span>)<\/span>\n<\/code><\/pre>\n\n\n\n
      \u7d50\u679c\u306f\u3001\u6d78\u6f64\u30ec\u30d9\u30eb\u30bc\u30ed\u306b\u6bd4\u30791\u306e\u5834\u5408\u306f2.6\u500d\u306e\u30ea\u30b9\u30af\u30012\u306e\u5834\u5408\u306f4.7\u500d\u306e\u30ea\u30b9\u30af\u3068\u306a\u308b\u3002<\/p>\n\n\n\n
      \u89e3\u8aac\u52d5\u753b\uff1aEZR\u3067 Fine-Gray \u56de\u5e30<\/h3>\n\n\n\n