Related Experiment Videos
Significance tests for cancer screening trials
1Mathematical Statistics and Applied Mathematics Section, National Cancer Institute, Bethesda, Maryland 20892.
Biometrics
|September 1, 1989
Summary
Cancer screening aims to reduce mortality by early detection. Comparing case mortality rates in trials is more powerful than population rates, especially for cancers where incidence isn't reduced by screening.
Area of Science:
- Biostatistics
- Epidemiology
- Oncology
Background:
- Cancer screening programs aim to decrease cancer mortality through early tumor detection.
- Screening can reduce both cancer incidence and mortality for some cancers, while for others, it primarily reduces mortality.
Purpose of the Study:
- To derive optimal statistical tests for comparing cancer mortality rates in screening trials.
- To evaluate the efficiency of different statistical tests under various scenarios of cancer incidence.
Main Methods:
- Utilized Poisson models to analyze cancer incidence and mortality data.
- Employed Fisher's exact test and Pearson chi-square approximation for mortality rate comparisons.
- Assessed the asymptotic relative efficiencies of test statistics.
Main Results:
- Testing equality of case mortality rates using Fisher's exact test or its Pearson chi-square approximation is nearly optimal when cancer incidence rates are equal.
- These tests are fully efficient when cancer incidence rates are unequal.
- Comparing case mortality rates is more powerful than comparing population mortality rates in screening trials.
Conclusions:
- Fisher's exact test and its approximation are effective for analyzing cancer screening trial data, particularly for mortality outcomes.
- Comparing case mortality offers a more powerful statistical approach in specific cancer screening trial designs.
- The findings are illustrated using data from a breast cancer screening trial.