On the sample size for studies based upon McNemar's test.

Area of Science:

• Biostatistics
• Clinical Trials

Background:

• McNemar's test is used for paired dichotomous data.
• Sample size calculation typically requires discordance probability and odds ratio.
• Investigators often lack precise discordance probability estimates.

Purpose of the Study:

• To develop a method for sample size calculation for McNemar's test when discordance probability is unknown.
• To utilize marginal probabilities to estimate sample size requirements.
• To compare the proposed method's sample size estimates with existing literature.

Main Methods:

• Utilizing marginal probabilities to define admissible ranges for off-diagonal cell probabilities.
• Calculating sample size based on these admissible ranges.
• Comparing derived sample sizes with values from Schlesselman and Connett et al.

Main Results:

• The proposed method provides a practical approach to sample size calculation without prior discordance probability.
• Sample size estimates are comparable to those obtained through Monte Carlo simulations.
• The method establishes bounds for cell probabilities based on marginal probabilities.

Conclusions:

• This approach offers a viable alternative for sample size determination in studies using McNemar's test.
• It enhances study planning by accommodating uncertainty in discordance probability.
• The findings align with previous simulation-based results, validating the method.