|Title:||Comparative analysis of clinical trials using long-term survivor models|
Survival analysis (Biometry)
Hong Kong Polytechnic University -- Dissertations
Department of Applied Mathematics
|Pages:||xii, 154 leaves : ill. ; 30 cm|
|Abstract:||This study ascertains the benefits of using long-term survivor models (mixture models) in analysis of clinical trials. Survival times of 195 patients with pharynx cancer are analysed with mixture models advocated by Mailer and Zhou (1996). Statistical test using both non-parametric and parametric methods suggest immunes is present in the sample (at 5% level of significance) and mixture models should apply. Weibull mixture model provides an excellent description of patients' survival pattern. The mean absolute percentage errors of fitting using Weibull mixture model is 8.34% while it is 17.97% for ordinary Weibull model. This represents more than 100% improvement in goodness of fit. Other measurements of goodness of fit such as log-likelihood and correlation coefficient for censored data yield consistent result. The mixture model predicts that in long run 20.9% of the patients are free from pharynx cancer after treatments. In the trials, 100 patients are treated with standard radiation treatment (radiotherapy) alone while the other 95 patients are treated with an additional chemotherapeutic agent (neo-adjuvant chemotherapy). Comparison of treatment effects using traditional approaches did not detect significant difference between the two regimens (at 5% level of significance). Parametric test based on Weibull mixture model, however, reveals significant difference between them. Patients short-tern survival is comprised by the neo-adjuvant and their median survival time is 397 days (i.e. 13 months) compared with 553 days (i.e. 18 months) for those treated with radiation alone. However, patients receiving the neo-adjuvant have a better long-term prospective as higher proportion is cured (23.2%). For patients treated with radiation alone only 17.2% is cured.|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: