n.rb.varse: Risk-based sample size for varying unit sensitivity

In RSurveillance: Design and Analysis of Disease Surveillance Activities

Description

Calculates sample size for risk-based sampling for a single risk factor and varying unit sensitivity, using binomial method

Usage

n.rb.varse(pstar, rr, ppr, spr, se, spr.rg, sep)

Arguments

pstar	design prevalence (scalar)
rr	relative risk values (vector, length equal to the number of risk strata)
ppr	population proportions for each risk group, vector of same length as rr
spr	planned surveillance proportions for each risk group, vector of same length as rr
se	unit sensitivities (vector of group values)
spr.rg	proportions of samples for each sensitivity value in each risk group (matrix with rows = risk groups, columns = sensitivity values), row sums must equal 1
sep	required population sensitivity (scalar)

Value

list of 3 elements, a matrix of sample sizes for each risk and sensitivity group, a vector of EPI values and a vector of mean sensitivity for each risk group

Examples

# examples for n.rb.varse
m<- rbind(c(0.8, 0.2), c(0.5, 0.5), c(0.7, 0.3))
n.rb.varse(0.01, c(5, 3, 1), c(0.1, 0.1, 0.8), c(0.4, 0.4, 0.2), c(0.92, 0.8), m, 0.95)

m<- rbind(c(0.8, 0.2), c(0.6, 0.4))
n.rb.varse(0.05, c(3, 1), c(0.2, 0.8), c(0.7, 0.3), c(0.95, 0.8), m, 0.95)

m<- rbind(c(1), c(1))
n.rb.varse(0.05, c(3, 1), c(0.2, 0.8), c(0.7, 0.3), c(0.95), m, 0.99)

RSurveillance documentation built on July 2, 2020, 2:33 a.m.