Statistical Analysis of Prostate Brachytherapy Treatment Data

Using data obtained over a period of 15 years from Brachytherapy procedures used on patients to treat prostate cancer, statistical analysis will be carried out to confirm the hypothesis that Brachtherapy treatments offer the best outcomes at the lowest cost incomparison to other treatment modalities. The first stage of the project will be to review the scope and objectives of the study with the stakeholders and finalize the methodology for this analysis. It is suggested that this analysis start initially with logistic regression, which would be used for estimating the probability of being in remission after a certain period of time following treatment. We then propose to continue with a more detailed analysis using transition models. Such models are used to analyze the data where a strong dependence is suspected between successive observations on a patient. These models distinguish themselves from survival models by allowing for transitions into and out of a set of states defined by the response variable instead of focusing on the time for a certain event to occur. The probability estimates derived from this model could enable the determination of expected years gained in remission by one treatment over the other as well as a comparison of expected costs for these treatments.

Ahalya Sivathayalan
Faculty Supervisor: 
Dr. Patrick Farrell