False-positive and false-negative rates for carcinogenicity screens

Area of Science:

• Toxicology
• Carcinogenesis Research
• Biostatistics

Background:

• Chemical carcinogen screening programs are vital for public health.
• Some screening strategies exhibit high false-positive error rates.
• Interpreting screening results requires careful consideration of potential biases.

Purpose of the Study:

• To compute upper bounds on false-positive error rates for various chemical carcinogen screening strategies.
• To emphasize the importance of biological context in interpreting screening outcomes.
• To highlight the limitations of relying solely on statistical significance.

Main Methods:

• Utilized designs from the National Cancer Institute (NCI).
• Incorporated historical spontaneous tumor rates from control animals.
• Calculated upper bounds for false-positive error rates.

Main Results:

• False-positive results are less likely at tissue sites with low spontaneous tumor rates.
• The location of a significant tumor increase is a critical factor.
• Statistical significance alone is insufficient for definitive conclusions.

Conclusions:

• Biological knowledge, including tumor site and dose-response relationships, is essential for validating screening results.
• Corroborative evidence across different species or sexes strengthens findings.
• A comprehensive approach integrating statistical and biological data enhances the reliability of carcinogen screening.