Formula n(sens) = Z^2 Se (1 - Se) / (d^2 x prevalence); n(spec) = Z^2 Sp (1 - Sp) / (d^2 x (1 - prevalence)). The larger governs.
Assumptions Prevalence in the sample matches the stated value. A single reference standard applied to everyone. Limitations Aims at precision, not at a hypothesis test of accuracy. Verification bias and spectrum effects are not modelled. References Buderer NMF. Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity. Acad Emerg Med. 1996;3(9):895-900. Lwanga SK, Lemeshow S. Sample Size Determination in Health Studies: A Practical Manual. Geneva: World Health Organization; 1991. Runs in your browser with fixed formulas. No AI, no AI credits, and your inputs are not sent anywhere. This is a planning aid and not a substitute for review by a qualified statistician.