4. Conclusion
Potential misclassification must be considered in sample size calculation for the one-sample binomial test. Even if sensitivity and specificity are high (98-99%), necessary sample sizes may be much higher than without misclassification.