Related Experiment Videos
Analyzing counts, durations, and recurrences in clinical trials
1Department of Economics, Concordia University, Montreal, PQ, Canada. jamesm@vax2.concordia.ca
Journal of Biopharmaceutical Statistics
|July 19, 2001
Summary
This study reveals that both pyridoxine and theotepa effectively treat bladder cancer, utilizing advanced statistical models for clinical trial data analysis. Findings suggest durations and tumor counts are more reliable indicators than recurrence numbers.
Area of Science:
- Urology
- Biostatistics
- Clinical Trial Analysis
Background:
- The Byar and Blackard bladder cancer data set (1978) has been underutilized.
- Previous analyses have not fully integrated recurrence and duration data consistently.
- Existing methodologies lack comprehensive modeling for clinical trial data.
Purpose of the Study:
- To determine the efficacy of pyridoxine and theotepa in treating bladder cancer.
- To develop a consistent statistical methodology for analyzing clinical trial data.
- To illustrate advanced procedures for analyzing recurrence and duration data.
Main Methods:
- Modeled recurrence counts using Poisson and negative binomial distributions.
- Applied Poisson and autoregressive models to analyze tumor recurrence durations.
- Utilized autoregressive Weibull and negative binomial distributions for durations and tumor counts.
Main Results:
- Autoregressive models provided a better fit for duration data than simple Poisson models.
- Duration and tumor count data proved more reliable for inference than recurrence counts.
- Both pyridoxine and theotepa demonstrated significant effectiveness in treating bladder cancer.
Conclusions:
- The study provides conclusive evidence for the efficacy of pyridoxine and theotepa against bladder cancer.
- Advanced statistical modeling enhances the reliability of clinical trial data interpretation.
- Findings challenge previous conclusions regarding the effectiveness of these bladder cancer treatments.