simPWSurv enables simulation of a clinical trial with essentially arbitrary
patterns of enrollment, failure rates and censoring.
The piecewise exponential distribution allows a simple method to specify a distribtuion
and enrollment pattern
where the enrollment, failure and dropout rate changes over time.
While the main purpose may be to generate a trial that can be analyzed at a single point in time or
using group sequential methods, the routine can also be used to simulate an adaptive trial design.
Enrollment, failure and dropout rates are specified by treatment group, stratum and time period.
Fixed block randomization is used; blocks must include treatments provided in failure and dropout
specification.
Default arguments are set up to allow very simple implementation of a non-proportional hazards assumption
for an unstratified design.
Usage
simPWSurv(
n = 100,
strata = tibble(Stratum = "All", p = 1),
block = c(rep("Control", 2), rep("Experimental", 2)),
enroll_rate = tibble(rate = 9, duration = 1),
failRates = tibble(Stratum = rep("All", 4), period = rep(1:2, 2), Treatment =
c(rep("Control", 2), rep("Experimental", 2)), duration = rep(c(3, 1), 2), rate =
log(2)/c(9, 9, 9, 18)),
dropoutRates = tibble(Stratum = rep("All", 2), period = rep(1, 2), Treatment =
c("Control", "Experimental"), duration = rep(100, 2), rate = rep(0.001, 2))
)Arguments
- n
Number of observations. If length(n) > 1, the length is taken to be the number required.
- strata
A tibble with strata specified in
Stratum, probability (incidence) of each stratum inp- block
Vector of treatments to be included in each block
- enroll_rate
Enrollment rates; see details and examples
- failRates
Failure rates; see details and examples; note that treatments need to be the same as input in block
- dropoutRates
Dropout rates; see details and examples; note that treatments need to be the same as input in block
Value
a tibble with the following variables for each observation
Stratum,
enroll_time (enrollment time for the observation),
Treatment (treatment group; this will be one of the values in the input block),
failTime (failure time generated using rpwexp()),
dropoutTime (dropout time generated using rpwexp()),
cte (calendar time of enrollment plot the minimum of failure time and dropout time),
fail (indicator that cte was set using failure time; i.e., 1 is a failure, 0 is a dropout).
Examples
library(dplyr)
library(tibble)
# example 1
simPWSurv(n = 20)
#> # A tibble: 20 × 7
#> # Groups: Stratum, Treatment [2]
#> Stratum enroll_time Treatment failTime dropoutTime cte fail
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 All 0.00680 Experimental 4.82 153. 4.83 1
#> 2 All 0.112 Control 28.6 1909. 28.7 1
#> 3 All 0.117 Experimental 62.0 557. 62.2 1
#> 4 All 0.229 Control 3.61 373. 3.84 1
#> 5 All 0.644 Experimental 0.732 1445. 1.38 1
#> 6 All 0.708 Experimental 92.0 751. 92.7 1
#> 7 All 0.892 Control 7.18 297. 8.07 1
#> 8 All 0.994 Control 8.04 231. 9.04 1
#> 9 All 1.04 Control 3.57 2096. 4.61 1
#> 10 All 1.11 Experimental 0.901 70.9 2.01 1
#> 11 All 1.22 Experimental 70.8 1985. 72.0 1
#> 12 All 1.30 Control 17.4 277. 18.7 1
#> 13 All 1.33 Control 7.98 2216. 9.31 1
#> 14 All 1.37 Experimental 64.5 668. 65.9 1
#> 15 All 1.46 Control 11.0 921. 12.4 1
#> 16 All 1.54 Experimental 3.50 1061. 5.04 1
#> 17 All 1.72 Experimental 3.27 931. 4.98 1
#> 18 All 1.91 Experimental 21.8 55.1 23.7 1
#> 19 All 1.92 Control 1.45 640. 3.37 1
#> 20 All 1.97 Control 4.51 208. 6.48 1
# example 2
# 3:1 randomization
simPWSurv(n = 20,
block = c(rep("Experimental",3), "Control"))
#> # A tibble: 20 × 7
#> # Groups: Stratum, Treatment [2]
#> Stratum enroll_time Treatment failTime dropoutTime cte fail
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 All 0.0278 Experimental 25.3 1524. 25.3 1
#> 2 All 0.0779 Experimental 57.9 1299. 58.0 1
#> 3 All 0.456 Control 6.68 270. 7.13 1
#> 4 All 0.498 Experimental 2.78 567. 3.27 1
#> 5 All 0.596 Experimental 68.7 1856. 69.3 1
#> 6 All 0.645 Control 8.25 65.0 8.89 1
#> 7 All 0.769 Experimental 16.5 450. 17.3 1
#> 8 All 0.812 Experimental 14.7 1039. 15.5 1
#> 9 All 1.06 Experimental 19.6 1159. 20.6 1
#> 10 All 1.08 Control 10.7 92.3 11.8 1
#> 11 All 1.11 Experimental 96.8 251. 97.9 1
#> 12 All 1.18 Experimental 1.36 174. 2.54 1
#> 13 All 1.21 Experimental 39.3 19.3 20.5 0
#> 14 All 1.41 Control 7.86 966. 9.28 1
#> 15 All 1.42 Experimental 14.9 20.0 16.3 1
#> 16 All 1.44 Experimental 1.47 73.9 2.91 1
#> 17 All 1.44 Experimental 0.434 2016. 1.88 1
#> 18 All 1.83 Experimental 25.0 4048. 26.8 1
#> 19 All 1.96 Control 11.3 1787. 13.3 1
#> 20 All 1.99 Experimental 0.934 452. 2.92 1
# example 3
# Simulate 2 strata; will use defaults for blocking and enroll_rate
simPWSurv(n = 20,
# 2 strata,30% and 70% prevalence
strata = tibble(Stratum = c("Low","High"), p = c(.3, .7)),
failRates = tibble(Stratum = c(rep("Low", 4), rep("High", 4)),
period = rep(1:2, 4),
Treatment = rep(c(rep("Control", 2),
rep("Experimental", 2)), 2),
duration = rep(c(3,1), 4),
rate = c(.03, .05, .03, .03, .05, .08, .07, .04)),
dropoutRates = tibble(Stratum = c(rep("Low", 2), rep("High", 2)),
period = rep(1, 4),
Treatment = rep(c("Control", "Experimental"), 2),
duration = rep(1, 4),
rate = rep(.001, 4)))
#> # A tibble: 20 × 7
#> # Groups: Stratum, Treatment [4]
#> Stratum enroll_time Treatment failTime dropoutTime cte fail
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 Low 0.0140 Control 0.325 437. 0.339 1
#> 2 High 0.262 Control 3.04 933. 3.30 1
#> 3 High 0.299 Experimental 9.48 1223. 9.77 1
#> 4 High 0.314 Experimental 2.57 428. 2.88 1
#> 5 Low 0.316 Control 17.7 2022. 18.1 1
#> 6 High 0.321 Control 17.5 1484. 17.8 1
#> 7 High 0.326 Control 3.71 1302. 4.03 1
#> 8 High 0.778 Experimental 4.46 761. 5.24 1
#> 9 Low 0.803 Experimental 13.9 266. 14.7 1
#> 10 High 1.17 Experimental 6.47 368. 7.64 1
#> 11 High 1.21 Control 20.1 172. 21.3 1
#> 12 High 1.21 Control 44.8 320. 46.1 1
#> 13 Low 1.30 Experimental 21.6 320. 22.9 1
#> 14 High 1.55 Experimental 77.1 1104. 78.7 1
#> 15 High 1.82 Control 10.7 7.97 9.79 0
#> 16 High 1.83 Experimental 9.91 168. 11.7 1
#> 17 High 1.86 Control 43.9 282. 45.8 1
#> 18 High 2.21 Experimental 9.89 233. 12.1 1
#> 19 High 2.25 Experimental 35.9 1860. 38.2 1
#> 20 High 2.36 Control 8.80 34.3 11.2 1
# example 4
# If you want a more rectangular entry for a tibble
failRates <- bind_rows(
tibble(Stratum = "Low" , period = 1, Treatment = "Control" , duration = 3, rate = .03),
tibble(Stratum = "Low" , period = 1, Treatment = "Experimental", duration = 3, rate = .03),
tibble(Stratum = "Low" , period = 2, Treatment = "Experimental", duration = 3, rate = .02),
tibble(Stratum = "High", period = 1, Treatment = "Control" , duration = 3, rate = .05),
tibble(Stratum = "High", period = 1, Treatment = "Experimental", duration = 3, rate = .06),
tibble(Stratum = "High", period = 2, Treatment = "Experimental", duration = 3, rate = .03))
dropoutRates <- bind_rows(
tibble(Stratum = "Low" , period=1, Treatment = "Control" , duration = 3, rate = .001),
tibble(Stratum = "Low" , period=1, Treatment = "Experimental", duration = 3, rate = .001),
tibble(Stratum = "High", period=1, Treatment = "Control" , duration = 3, rate = .001),
tibble(Stratum = "High", period=1, Treatment = "Experimental", duration = 3, rate = .001))
simPWSurv(n = 12,
strata = tibble(Stratum = c("Low","High"), p = c(.3, .7)),
failRates = failRates,
dropoutRates = dropoutRates)
#> # A tibble: 12 × 7
#> # Groups: Stratum, Treatment [4]
#> Stratum enroll_time Treatment failTime dropoutTime cte fail
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 Low 0.00555 Control 5.89 950. 5.90 1
#> 2 Low 0.0413 Experimental 19.5 298. 19.5 1
#> 3 High 0.219 Experimental 8.41 740. 8.63 1
#> 4 High 0.298 Experimental 8.22 95.1 8.52 1
#> 5 High 0.393 Control 12.4 1303. 12.8 1
#> 6 High 0.423 Control 19.6 717. 20.0 1
#> 7 High 0.498 Control 1.88 869. 2.38 1
#> 8 Low 0.589 Experimental 28.5 1741. 29.1 1
#> 9 High 0.667 Control 10.2 18.5 10.9 1
#> 10 Low 0.720 Control 16.6 681. 17.3 1
#> 11 Low 0.724 Control 28.0 92.2 28.7 1
#> 12 High 0.888 Experimental 3.39 583. 4.28 1