make.ruleoutTable.pre: make.ruleoutTable.pre

In n8thangreen/IDEAdectree: The IDEA study clinical data health economics analysis with decision trees

Description

make.ruleoutTable.pre creates the times and costs of a diagnositic pathway for suspected active TB with and without an initial rule-out test, split by Dosanjh category.

Usage

make.ruleoutTable.pre(thresh = seq(from = 1, to = 0.8, by = -0.01),
  Ctest = c(200, 300), FNcost = 0, FNtime = 42, ruleouttime = 1L,
  pathreturn = 1L, qaly = 0.67, A = 55, npatients = nrow(data),
  cat4percent = table(data$DosanjhGrouped)[4] * 100/npatients,
  comb = NA, cat3TB = TRUE, cat4propfollowup = 0,
  prop_highrisk = 0.4, stat = "Median", model = "pretest.fixed")

Arguments

thresh	test sensitivity and specificity
Ctest	cost of rule-out test
FNcost	false negative cost of true diagnosis or cost to start of standard pathway
FNtime	false negative time to true diagnosis or start of pathway
ruleouttime	time taken for rule-ou test result
pathreturn	does a ruled-out individual return to the pathway? false positives return to standard pathway Y=1/N=0
qaly	QALY for disease
A	cost of one day in full health
npatients	number of patients in cohort
cat4percent	percent of patients in Dosanjh category 4
comb	combined sensitivity, specificity and rule-out test cost array
cat3TB	are Dosanjh category 3 patients all active TB?
cat4propfollowup	proportion of negative Dosanjh category 4 patients, not immediately on standard pathway, who are followed-up at 6 weeks (alpha). We assume that all active TB cases are certain to be followed-up.
prop_highrisk	Minimum predictive probability of TB risk score (delta)
stat	Which statistic to use for time and cost estimate (e.g. Mean, Median, 1st Qu.)
model	Which model structure/tree design to use ( pretest.fixed is include highest risk proportion with clinical judgement before rule-out test (fixed proportion risk factor independent) **USED IN MAIN PAPER**, pretest.var is include highest risk proportion with clinical judgement before rule-out test (risk factor dependent), posttest.fixed is test everyone first and include randomly selected proportion then include highest risk proportions, posttest.var is test everyone first then remove highest risk proportion, pretest.var.sensspec.var is remove highest risk proportion before rule-out test and modify sensitivity and specificity proportions wrt subset case-mix.)
[previously,	proportion of patients put on standard pathway by clinical judgment (gamma)],

Value

list

n8thangreen/IDEAdectree documentation built on Feb. 10, 2020, 11:35 a.m.