Saddlepoint p-values and confidence intervals for a class of two sample permutation tests for current status and...

Area of Science:

• Biostatistics
• Cancer Research
• Statistical Modeling

Background:

• Current status and panel count data are common in cancer and tumorigenicity studies.
• Permutation tests are widely used for analyzing this type of data.
• Existing methods often rely on computationally intensive simulations or less accurate approximations.

Purpose of the Study:

• To develop an accurate and efficient method for calculating exact mid-p-values for permutation tests.
• To replace computationally expensive permutation simulations with analytical saddlepoint computations.
• To provide a more accurate alternative to normal approximations for analyzing current status and panel count data.

Main Methods:

• The double saddlepoint method was adapted to compute exact mid-p-values.
• Analytical saddlepoint computations replaced permutation simulations.
• The method was applied to real tumorigenicity panel count data.
• A simulation study compared saddlepoint approximation with normal asymptotic approximation.

Main Results:

• The saddlepoint method provides extremely accurate mid-p-values, often more accurate than normal approximations.
• Analytical computations are significantly faster than permutation simulations.
• The method facilitates accurate confidence interval calculations for treatment effects.
• The approach is effective for analyzing recurrent event data.

Conclusions:

• The saddlepoint method offers a fast, accurate, and reliable approach for analyzing current status and panel count data.
• This method improves upon traditional permutation tests and normal approximations in biostatistical and cancer research.
• It enables more precise estimation of treatment effects and recurrent event rates.