Cautionary tales of survival analysis: conflicting analyses from a clinical trial in breast cancer
1997

Cautionary Tales of Survival Analysis in Breast Cancer Trials

Sample size: 410 publication Evidence: moderate

Author Information

Author(s): WM Gregory, K Bolland, J Whitehead, RL Souhami

Primary Institution: ICRF Breast Unit, Guy's Hospital, London

Hypothesis

How does the timing of statistical analysis affect the interpretation of survival data in breast cancer clinical trials?

Conclusion

The timing of analyses in breast cancer trials can significantly alter the perceived effectiveness of treatments, highlighting the importance of long follow-up periods.

Supporting Evidence

  • Data from a completed randomized trial in breast cancer were used to demonstrate variations in survival estimates.
  • False plateaux in survival curves were common when few events occurred.
  • Long follow-up periods are necessary for accurate interpretation of survival data.

Takeaway

When doctors study how long patients live after treatment, the timing of their checks can change the results a lot, so they need to be careful.

Methodology

The study involved reanalyzing data from a completed randomized trial in breast cancer, focusing on the effects of timing on survival analysis.

Potential Biases

Potential biases may arise from the timing of analyses and the interpretation of results based on limited data.

Limitations

The study relies on existing trial data, which may not fully capture all variables affecting survival.

Participant Demographics

Patients were recruited from December 1979 to November 1985, with a total of 410 participants.

Statistical Information

P-Value

0.005

Confidence Interval

0.40-0.95

Statistical Significance

p<0.05

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication