sse.combined {RSurveillance}

System sensitivity by combining multiple surveillance components

Description

Calculates overall system sensitivity for multiple components, accounting for lack of independence (overlap) between components

Usage

sse.combined(C = NA, pstar.c, rr, ppr, sep)

Arguments

C

population sizes (number of clusters) for each risk group, NA or vector of same length as rr

pstar.c

cluster level design prevalence (scalar)

rr

cluster level relative risks (vector, length equal to the number of risk strata)

ppr

cluster level population proportions (optional), not required if C is specified (NA or vector of same length as rr)

sep

sep values for clusters in each component and corresponding risk group. A list with multiple elements, each element is a dataframe of sep values from a separate component, first column= clusterid, 2nd =cluster-level risk group index, 3rd col = sep

Value

list of 2 elements, a matrix (or vector if C not specified) of population-level (surveillance system) sensitivities (binomial and hypergeometric and adjusted vs unadjusted) and a matrix of adjusted and unadjusted component sensitivities for each component

Examples

# example for sse.combined (checked in excel combined components.xlsx)
C<- c(300, 1200)
pstar<- 0.01
rr<- c(3,1)
ppr<- c(0.2, 0.8)
comp1<- data.frame(id=1:100, rg=c(rep(1,50), rep(2,50)), cse=rep(0.5,100)) 
comp2<- data.frame(id=seq(2, 120, by=2), rg=c(rep(1,25), rep(2,35)), cse=runif(60, 0.5, 0.8))
comp3<- data.frame(id=seq(5, 120, by=5), rg=c(rep(1,10), rep(2,14)), cse=runif(24, 0.7, 1))
sep<- list(comp1, comp2, comp3)
sse.combined(C, pstar, rr, sep = sep)
sse.combined(C=NA, pstar, rr, ppr, sep = sep)