本ページについて

このページには、第4回データサイエンスラウンドテーブル会議 (2017年3月9日開催) で「抗がん剤の第III相試験における生存時間解析に関する統計的課題」において挙げていただいた疑問点の説明資料を掲載する予定です。
当日の発表スライドやディスカッション結果は上記のリンクから入手できます。

Rank preserving structural failure time (RPFST) 法の簡単な解説

Rのパッケージ (rpsftm) があり、vignette の解説が事例付きでわかりやすいと思われます。

掲載予定。

競合リスクモデルのシミュレーション結果 (Dignam, et al, 2012) の再考

掲載予定。

Restricted mean survival time (RMST) の解析プログラム等の紹介

宇野一先生のHP に R, Stata, SAS 用のプログラムが紹介されている。

survRM2 package の機能を拝借した、単群の R 用サンプルコードを以下に示す。


  require("survRM2")

  time <- c(27.8606,14.9723,1.2585,6.4632,8.4133,0.2114,1.9108,0.6967,6.7331,3.5564,
            1.7072,2.5596,9.7294,5.2727,5.5134,8.6262,1.9087,1.5617,2.3439,3.3077,
            0.0198,1.0134,15.5296,7.0067,8.4248,0.2253,2.0488,3.2047,9.7172,2.2049,
            10.5392,0.8608,0.2194,3.5848,14.7897,14.5835,7.8358,3.6135,19.0975,1.0123,
            0.9825,18.6853,0.2339,3.4684,16.3825,1.9767,0.2474,4.4012,8.2612,3.2463)
  event <- c(1,1,1,1,1,1,1,1,1,0,1,1,1,1,0,1,1,1,1,1,0,0,1,1,1,1,1,1,1,1,
            1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,1)

  onermst <- getAnywhere(rmst1)$objs[[1]]
  onermst(time, event, tau = 5, alpha = 0.05)
  

群間比較の場合は vignette に事例等の豊富な記載がある。

SASマクロは %RMST1 (単群), %RMST2 (二群間比較), %rmst2reg (共変量調整) があり、いずれもデータセット名や変数名を指定するだけで計算が可能である。
SASのLIFETEST Procedure には timelim オプションがあり、一見同様の解析が可能に見えるが、データセット加工を行わないと正しい RMST を計算できない場合がある (Thanks to Kosei Tajima and Dr. Takahiro Hasegawa)。 また、標準誤差の計算方法が異なるため、現在最新の SAS/STAT 14.2 以下では、マクロ以外で計算する方法がないと思われる。 標準誤差の計算方法が SAS とその他 (Royston and Parmar, 2013) で異なり、性質が違う点に注意したい(日本製薬工業協会の資料より)。

参考文献

  1. Robins JM, Tsiatis AA. Correcting for non-compliance in randomized trials using rank preserving structural failure time models. Communications in Statistics Theory and Methods 1991; 20: 2609–2631. DOI: 10.1080/03610929108830654.
  2. White IR, Babiker AG, Walker S, Darbyshire JH. Randomisation-based methods for correcting for treatment changes: examples from the Concorde trial. Statistics in Medicine 1999; 18: 2617–2634. DOI: 10.1002/(SICI)1097-0258(19991015)18:19<2617::AID-SIM187>3.0.CO;2-E.
  3. White IR, Walker S, Babiker AG, Darbyshire JH. Impact of treatment changes on the interpretation of the Concorde trial. AIDS 1997; 11: 999–1006. DOI: 10.1097/00002030-199708000-00008.
  4. Dignam JJ, Zhang Q, Kocherginsky M. The use and interpretation of competing risks regression models. Clinical Cancer Research 2012; 18(8): 2301–2308. DOI: 10.1158/1078-0432.CCR-11-2097.
  5. Royston P, Parmar MKB. Restricted mean survival time: an alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Medical Research Methodolgy 2013; 13: 152. doi: 10.1186/1471-2288-13-152.

履歴

  • 2018/06/13 RMST の標準誤差についての記載を訂正。
  • 2017/05/25 RMST の解析プログラムについての記載を追加。
  • 2017/03/28 公開