Use this URL to cite or link to this record in EThOS:
Title: Competing risks methodology applied to the analysis of cancer therapy outcome
Author: Ataman, Ozlem
ISNI:       0000 0001 3431 4700
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 2003
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The relationships between potential prognostic factors and various types of failures in different cancer sites were studied in a competing risks model. The methods that have been applied here were based on time to the first observed event. The clinical data analyzed came from the randomized trials of continuous hyperfractionated accelerated radiotherapy (CHART) vs. conventional radiotherapy in non-small cell lung cancer (NSCLC) and head and neck squamous cell carcinoma (HNSCC). Clinical outcome was analyzed in 549 NSCLC patients and in 309 HNSCC patients. Competing risks analysis was performed using the biomedical data package (BMDP) statistical software with an accelerated failure-time model and a log- logistic hazard function. The final reduced models included age, sex, clinical stage and treatment for NSCLC and proliferative pattern, bcl2, cyclin D1, TRTCD31, Ki-67 scores and T and N stage for HNSCC. From these models, prognostic Indies for local, regional and distant failure were estimated for each individual patient and prognostic groups were formed to identify patients at risk for different types of failures. Cumulative incidence (CI) and the Kaplan-Meier estimate (KM) are two estimates for quantifying the occurrence of an endpoint over time in the competing risks framework. Properties of the two estimators for treatment failure and late morbidity over time have been explored for different prognostic groups using the CHART NSCLC and HNSCC studies. The CI estimate showed unexpected variations in estimating tumour outcome and late side effects in unfavourable prognostic groups where there was a higher incidence of competing risks. Without a comprehensive understanding of the assumptions of KM method, the clinical interpretations must be made with caution. The KM and the Cl methods should be used as complementary analyses. The natural behaviour of the tumour site and the competing events under study should be clearly understood by clinicians.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available