Testing for sufficient follow-up in survival data

✍ Scribed by Pao-sheng Shen

• Book ID: 104301440
• Publisher: Elsevier Science
• Year: 2000
• Tongue: English
• Weight: 109 KB
• Volume: 49
• Category: Article
• ISSN: 0167-7152
• DOI: 10.1016/s0167-7152(00)00063-8

✦ Synopsis

proposed a nonparametric test, called the ˆ n-test, for testing insu cient versus su cient follow-up. It was noted by Maller and Zhou (1996, Survival Analysis with longterm survivors, Wiley, New York) that the ˆ n-test can have Type I error much larger than the nominal value. In this note, it is pointed out that the ˆ n-test can be used instead to test the hypothesis of no immune proporiton when follow-up is su cient. To test whether there is su cient follow-up, a modiÿed test, called the ˜ n-test, is proposed and shown to be consistent. Simulations are conducted to investigate the performances of the ˜ n-test and the other existing test, the qn-test, developed by . Our investigation shows that both the ˜ n and qn-test can have a substantially smaller distortion of nominal signiÿcance level compared to the ˆ n-test.