Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

A comparison of tumour incidence analyses applicable in single‐sacrifice animal experiments

A comparison of tumour incidence analyses applicable in single‐sacrifice animal experiments Monte Carlo methods are used to evaluate the operating characteristics of several trend tests for comparing incidence rates of occult tumours. The candidate tests are applicable in long‐term animal experiments that have just one sacrifice time and no information on cause of death. When survival decreases with dose and the tumour is lethal, the tests that do not adjust for survival or that assume tumours are non‐lethal behave conservatively, whereas the test that assumes tumours are rapidly lethal rejects the hypothesis of equal incidence rates too frequently. The only test which consistently operates at the correct level is one that specifies a constant difference between the death rates for animals with and without the tumour. Surprisingly, the test that specifies a constant ratio of these death rates often is conservative. Finally, a test based on a simple modification of the unadjusted test, which accounts for differential mortality by scaling down the size of the risk set, performs reasonably well in many cases. Among the tests that operate at the proper level, the constant risk difference test consistently exhibits high power across a wide range of situations. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Statistics in Medicine Wiley

A comparison of tumour incidence analyses applicable in single‐sacrifice animal experiments

Dinse, Gregg E.
Statistics in Medicine , Volume 13 (5‐7) – Mar 15, 1994
Download PDF
20 pages

Loading next page...
 
/lp/wiley/a-comparison-of-tumour-incidence-analyses-applicable-in-single-dGII2LFt2m

References (34)

Publisher
Wiley
Copyright
Copyright © 1994 John Wiley & Sons, Ltd.
ISSN
0277-6715
eISSN
1097-0258
DOI
10.1002/sim.4780130530
Publisher site
See Article on Publisher Site

Abstract

Monte Carlo methods are used to evaluate the operating characteristics of several trend tests for comparing incidence rates of occult tumours. The candidate tests are applicable in long‐term animal experiments that have just one sacrifice time and no information on cause of death. When survival decreases with dose and the tumour is lethal, the tests that do not adjust for survival or that assume tumours are non‐lethal behave conservatively, whereas the test that assumes tumours are rapidly lethal rejects the hypothesis of equal incidence rates too frequently. The only test which consistently operates at the correct level is one that specifies a constant difference between the death rates for animals with and without the tumour. Surprisingly, the test that specifies a constant ratio of these death rates often is conservative. Finally, a test based on a simple modification of the unadjusted test, which accounts for differential mortality by scaling down the size of the risk set, performs reasonably well in many cases. Among the tests that operate at the proper level, the constant risk difference test consistently exhibits high power across a wide range of situations.

Journal

Statistics in Medicine – Wiley

Published: Mar 15, 1994

There are no references for this article.