A Novel Statistical Method to Examine The Cancer Disparities and the Effectiveness of Cancer Intervention



Journal Title

Journal ISSN

Volume Title



Cancer remains the second leading cause of death in the U.S. and worldwide. To thoroughly understand the incidence and mortality of cancer, statistical modeling is crucial in analyzing and interpreting data. The age-period-cohort (APC) analysis has been utilized to distinguish and estimate the age, period, and cohort trends of cancer incidence and mortality. This dissertation mainly focuses on testing the differences in age trends of APC models and their applications in cancer research. In Chapter 2, the APC models and a statistical likelihood-ratio (LR) test are discussed with full details. A rigorous LR test on the differences in age trends is developed following the APC models and the intrinsic estimator (IE) method. The performance of the likelihood-ratio test on age trends is assessed through simulations. Chapter 3 compares data of cancer through APC analysis and examines cancer racial/gender disparities with the LR test. In Chapter 4, a novel statistical test procedure to examine the effectiveness of cancer intervention is described. The robustness of the test is evaluated through sensitivity analysis. Chapter 5 applies the newly developed procedure to cancer incidence data to assess the effectiveness of the intervention programs of a few types of cancer with known intervention programs. Chapter 6 further applies the procedure to the mortality data of these cancer types to assess the effectiveness of the intervention programs on cancer mortality. In this dissertation, the data for eight cancer sites, namely breast (ER-positive and ER-negative), colon and rectum, cervical, leukemia, liver, lung and bronchus, pancreatic, and prostate cancer, between the years 1973 and 2015 are analyzed to illustrate the newly developed statistical tests.



Age-Period-Cohort analysis, Likelihood-ratio test, Cancer screening