Statistical issues in randomized trials of cancer screening

Abstract

Background: The evaluation of randomized trials for cancer screening involves special statistical considerations not found in therapeutic trials. Although some of these issues have been discussed previously, we present important recent and new methodologies.

Methods: Our emphasis is on simple approaches.

Results: We make the following recommendations: (1) Use death from cancer as the primary endpoint, but review death records carefully and report all causes of death; (2) Use a simple "causal" estimate to adjust for nonattendance and contamination occurring immediately after randomization; (3) Use a simple adaptive estimate to adjust for dilution in follow-up after the last screen

Conclusion: The proposed guidelines combine recent methodological work on screening endpoints and noncompliance/contamination with a new adaptive method to adjust for dilution in a study where follow-up continues after the last screen. These guidelines ensure good practice in the design and analysis of randomized trials of cancer screening.

Figure 1

Effect of Follow-up on Estimated Reduction in Breast Cancer Deaths Data are from the HIP Study of breast cancer screening. The plot shows point estimates and 95% confidence intervals for estimated reduction in breast cancer deaths, per 10,000 compliers (participants who would have receive breast cancer screening if offered) due to screening. "Fixed" refers to fixing the follow-up time before examining the data. The estimated reduction is computed as negative d_causal(t), where t is the fixed follow-up time. "Adaptive" is the proposed method that bases the follow-up time on the maximum, over time, of a Z-statistic, where confidence intervals are computed by bootstrapping. The estimated reduction is computed as negative d_causal(t*), where t* is the follow-up time based on the adaptive approach.