Table 5 Description of the data generation mechanism of the unpaired and paired design in the simulation study (Bin: binomial distribution, MVBin: multivariate binomial distribution, k: number of trials, p: success probability, ρ: dependence between both tests, N: total sample size, nDE: diseased individuals in experimental group, nDC: diseased individuals in comparator group)
From: Blinded sample size re-estimation in a comparative diagnostic accuracy study
Unpaired design | Paired design | |
|---|---|---|
Diseased individuals (nD) according to reference standard | \({n}_{D_E}\;\sim\;Bin(k = N, p = \pi_\mathrm{true})\) \(n_{D_C}\,\sim\,Bin(k = N, p = \pi_\mathrm{true})\) | nD ~ Bin(k = N, p = πtrue) |
True Positive Results (TP) | \(TP_{E}\,\sim Bin(k = n_{D_E}, p = Se_{E})\) \(TP_{C}\,\sim Bin(k = n_{D_C}, p = Se_{C})\) | \((TP_{E}, TP_{C})\,\sim MVBin(k_\mathrm{E} = n_{D_E}, k_\mathrm{C} = n_{D_C},\) pE = SeE, pC = SeC, ρ = TPPR) |
True Negative Results (TN) | \(TN_{E}\,\sim Bin(k = N - n_{D_E}, p = Sp_{E})\) \(TN_{C}\,\sim Bin(k = N - n_{D_C}, p = Sp_{C})\) | \((TN_{E}, TN_{C})\,\sim MVBin(k_\mathrm{E} = N - n_{D_E}, k_\mathrm{C} = N - n_{D_C},\) pE = SpE, pC = SpC, ρ = TNNR) |