Inferring Effects on Tumor Frequencies and Times to Observation in the Analysis of Tumor Multiplicity Data
✍ Scribed by Ralph L. Kodell; James J. Chen
- Publisher
- John Wiley and Sons
- Year
- 2001
- Tongue
- English
- Weight
- 113 KB
- Volume
- 43
- Category
- Article
- ISSN
- 0323-3847
No coin nor oath required. For personal study only.
✦ Synopsis
A sequential likelihood-ratio strategy for analyzing tumor multiplicity data from animal bioassays is described and illustrated. The likelihood function is composed of a negative binomial component for the number of induced tumors per animal and a Weibull component for the time to observation of such tumors. The method can distinguish whether differences in tumor responses between treatment groups are due to differences in numbers of induced tumors or to differences in times to observation of such tumors. Although adapted from the underlying model of Kokoska (Kokoska, 1987, Biometrics 43, 525-534;, Anticancer Research 13, 1357-1364), the proposed testing methodology can give different results from the Kokoska method, as is shown in an example with real data. Several extensions of the model are described.