Letter

ATAC trial did not report interim results

BMJ 2003; 327 doi: 10.1136/bmj.327.7422.1050 (Published 30 October 2003)
Cite this as: BMJ 2003;327:1050.1

Access to the full text of this article requires a subscription or payment. Please log in or subscribe below.

  1. Michael Baum, visiting professor of medical humanities,
  2. Aman Buzdar, doctor,
  3. Anthony Howell, professor of medical oncology
  1. Clinical Trials Group, Department of Surgery, University College London, London W1W 7EJ
  2. Department of Medical Oncology, Medical Breast Cancer Service, University of Texas, MD Anderson Cancer Center, Houston, TX 77030, USA
  3. Department of Medical Oncology, Cancer Research UK, University of Manchester, Manchester M20 4BX

    EDITOR—We agree with Goodare et al that it is inappropriate to publish interim analyses.1 However, the results of the ATAC (arimidex, tamoxifen, alone or in combination) trial published in the Lancet were the first major analysis as fully defined in the protocol, not an interim analysis.2

    The statistical power of the trial, defined in the protocol, determined that this analysis would be triggered at 1056 events. Using first event as the primary end point in breast cancer trials is …

    Access to the full text of this article requires a subscription or payment

    Subscribe

    Article access

    Article access for 1 day

    Purchase this article for £20 $30 €32*

    The PDF version can be downloaded as your personal record

    * Prices do not include VAT

    Rapid responses

    THIS WEEK'S POLL

    BMJ most popular