update_treat_stageshift: For stage-shifted cases in an early detection scenario,...

In cancerpolicy/bcimodel: Model the impact of early detection and treatment interventions for breast cancer

Description

Takes treatments from a paired scenario with no early detection and updates treatments only for those advanced-stage cases who were stage-shifted in the early detection scenario

Usage

update_treat_stageshift(policies, shifts, treats)

Arguments

policies	Data frame specifying policies - see ex1$pol. If 'pairnum' = NA, no treatments are updated. If 'pairnum' is numeric, treatments from that number scenario will be the starting point, with early-detected cases shifted according to the next input
shifts	List of treatment-shift indicators (see shifttreatment_indicator)
treats	List of original treatment matrices (see treatments_by_policy)

Value

List of updated treatment matrices

Examples

library(bcimodel)
data(ex1) 
# ex1$nh shows that there are 4 stage-subgroups. Use a fake random distribution of groups 1:4 for the population before stage-shifting.
popdistr <- matrix(sample.int(4, size=40, replace=TRUE), nrow=20, ncol=2)

# Create stageshift indicator matrices for all 3 scenarios: no stage shifts for #1 and #2, but 30% stageshift for #3. Use a small population of size 20, and 2 sims
stageshifts <- list(base=matrix(0, nrow=20, ncol=2),
                    tam=matrix(0, nrow=20, ncol=2), 
                    tamshift=stageshift_indicator(0.85, 20, 2))

# Get the actual stages - only policy #3 has stage-shifting
stages <- lapply(stageshifts, shift_stages, original=popdistr, map=ex1$map)

# First 2 scenarios have no shifting, same treatment. Randomly distribute treatment
# Scenario 3 has shifting. NAs indicate that treatments will be same as the scenario
# with no early detection but same treatments
treats <- treatments_by_policy(policies=ex1$pol, 
                          treat_chars=ex1$tx, 
                          stagegroups=stages, 
                          map=ex1$map, 
                          pop_size=20, nsim=2)

# Indicators of treatment shifting
treatshifts <- lapply(ex1$pol$num,
                          shifttreatment_indicator,
                          type=ex1$pol$pairnum,
                          shifts=stageshifts, 
                          basecase=popdistr, 
                          map=ex1$map)

newtreat <- update_treat_stageshift(ex1$pol, treatshifts, treats) 
# Two cases have been shifted in #3 compared to #2 
lapply(treatshifts, table)
lapply(newtreat, table)

cancerpolicy/bcimodel documentation built on June 30, 2019, 12:39 a.m.