Use with simulated portfolio data to generate under-reporting stats for specified scenarios.
Arguments
- df_portf
dataframe as returned by
sim_test_data_portfolio- extra_ur_sites
numeric, set maximum number of additional under-reporting sites, see details Default: 3
- ur_rate
numeric vector, set under-reporting rates for scenarios Default: c(0.25, 0.5)
- r
integer, denotes number of simulations, default = 1000
- poisson_test
logical, calculates poisson.test pvalue
- prob_lower
logical, calculates probability for getting a lower value
- parallel
logical, use parallel processing see details, Default: FALSE
- progress
logical, show progress bar, Default: TRUE
- site_aggr_args
named list of parameters passed to
site_aggr, Default: list()- eval_sites_args
named list of parameters passed to
eval_sites, Default: list()- check
logical, perform data check and attempt repair with
Value
dataframe with the following columns:
- study_id
study identification
- site_number
site identification
- n_pat
number of patients at site
- n_pat_with_med75
number of patients at site with visit_med75
- visit_med75
median(max(visit)) * 0.75
- mean_ae_site_med75
mean AE at visit_med75 site level
- mean_ae_study_med75
mean AE at visit_med75 study level
- n_pat_with_med75_study
number of patients at site with visit_med75 at study excl site
- extra_ur_sites
additional sites with under-reporting patients
- frac_pat_with_ur
ratio of patients in study that are under-reporting
- ur_rate
under-reporting rate
- pval
p-value as returned by
poisson.test- prob_low
bootstrapped probability for having mean_ae_site_med75 or lower
- pval_adj
adjusted p-values
- prob_low_adj
adjusted bootstrapped probability for having mean_ae_site_med75 or lower
- pval_prob_ur
probability under-reporting as 1 - pval_adj, poisson.test (use as benchmark)
- prob_low_prob_ur
probability under-reporting as 1 - prob_low_adj, bootstrapped (use)
Details
The function will apply under-reporting scenarios to each site. Reducing the number of AEs by a given under-reporting (ur_rate) for all patients at the site and add the corresponding under-reporting statistics. Since the under-reporting probability is also affected by the number of other sites that are under-reporting we additionally calculate under-reporting statistics in a scenario where additional under reporting sites are present. For this we use the median number of patients per site at the study to calculate the final number of patients for which we lower the AEs in a given under-reporting scenario. We use the furrr package to implement parallel processing as these simulations can take a long time to run. For this to work we need to specify the plan for how the code should run, e.g. plan(multisession, workers = 18)
See also
sim_test_data_study
get_config
sim_test_data_portfolio
sim_ur_scenarios
get_portf_perf
Examples
# \donttest{
df_visit1 <- sim_test_data_study(n_pat = 100, n_sites = 10,
frac_site_with_ur = 0.4, ur_rate = 0.6)
df_visit1$study_id <- "A"
df_visit2 <- sim_test_data_study(n_pat = 100, n_sites = 10,
frac_site_with_ur = 0.2, ur_rate = 0.1)
df_visit2$study_id <- "B"
df_visit <- dplyr::bind_rows(df_visit1, df_visit2)
df_site_max <- df_visit %>%
dplyr::group_by(study_id, site_number, patnum) %>%
dplyr::summarise(max_visit = max(visit),
max_ae = max(n_ae),
.groups = "drop")
df_config <- get_config(df_site_max)
df_config
#> # A tibble: 20 × 6
#> study_id ae_per_visit_mean site_number max_visit_sd max_visit_mean n_pat
#> <chr> <dbl> <chr> <dbl> <dbl> <int>
#> 1 0001 0.381 0001 2.85 19.9 10
#> 2 0001 0.381 0002 3.31 18.1 10
#> 3 0001 0.381 0003 3.14 18.1 10
#> 4 0001 0.381 0004 4.74 20.7 10
#> 5 0001 0.381 0005 5.20 19.2 10
#> 6 0001 0.381 0006 3.30 21 10
#> 7 0001 0.381 0007 4.07 19.9 10
#> 8 0001 0.381 0008 4.53 18.5 10
#> 9 0001 0.381 0009 2.95 21.7 10
#> 10 0001 0.381 0010 3.36 20.2 10
#> 11 0002 0.491 0001 3.03 20.9 10
#> 12 0002 0.491 0002 2.44 18.8 10
#> 13 0002 0.491 0003 3.62 20.3 10
#> 14 0002 0.491 0004 3.79 19.8 10
#> 15 0002 0.491 0005 4.19 19.3 10
#> 16 0002 0.491 0006 5.34 18.5 10
#> 17 0002 0.491 0007 4.23 18.9 10
#> 18 0002 0.491 0008 3.28 18.9 10
#> 19 0002 0.491 0009 5.33 18 10
#> 20 0002 0.491 0010 3.48 19.1 10
df_portf <- sim_test_data_portfolio(df_config)
df_portf
#> # A tibble: 3,746 × 8
#> study_id ae_per_visit_mean site_number max_visit_sd max_visit_mean patnum
#> <chr> <dbl> <chr> <dbl> <dbl> <chr>
#> 1 0001 0.381 0001 2.85 19.9 0001
#> 2 0001 0.381 0001 2.85 19.9 0001
#> 3 0001 0.381 0001 2.85 19.9 0001
#> 4 0001 0.381 0001 2.85 19.9 0001
#> 5 0001 0.381 0001 2.85 19.9 0001
#> 6 0001 0.381 0001 2.85 19.9 0001
#> 7 0001 0.381 0001 2.85 19.9 0001
#> 8 0001 0.381 0001 2.85 19.9 0001
#> 9 0001 0.381 0001 2.85 19.9 0001
#> 10 0001 0.381 0001 2.85 19.9 0001
#> # ℹ 3,736 more rows
#> # ℹ 2 more variables: visit <int>, n_ae <int>
df_scen <- sim_ur_scenarios(df_portf,
extra_ur_sites = 2,
ur_rate = c(0.5, 1))
#> aggregating site level
#> prepping for simulation
#> generating scenarios
#> getting under-reporting stats
#> evaluating stats
df_scen
#> # A tibble: 140 × 14
#> study_id site_number extra_ur_sites frac_pat_with_ur ur_rate n_pat
#> <chr> <chr> <dbl> <dbl> <dbl> <int>
#> 1 0001 0001 0 0 0 10
#> 2 0001 0001 0 0.122 0.5 10
#> 3 0001 0001 0 0.122 1 10
#> 4 0001 0001 1 0.222 0.5 10
#> 5 0001 0001 1 0.222 1 10
#> 6 0001 0001 2 0.322 0.5 10
#> 7 0001 0001 2 0.322 1 10
#> 8 0001 0002 0 0 0 10
#> 9 0001 0002 0 0.104 0.5 10
#> 10 0001 0002 0 0.104 1 10
#> # ℹ 130 more rows
#> # ℹ 8 more variables: n_pat_with_med75 <dbl>, visit_med75 <dbl>,
#> # mean_ae_site_med75 <dbl>, mean_ae_study_med75 <dbl>,
#> # n_pat_with_med75_study <int>, prob_low <dbl>, prob_low_adj <dbl>,
#> # prob_low_prob_ur <dbl>
df_perf <- get_portf_perf(df_scen)
df_perf
#> # A tibble: 27 × 5
#> fpr thresh extra_ur_sites ur_rate tpr
#> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 0.001 0.529 0 0 0.05
#> 2 0.001 0.529 1 0 0.05
#> 3 0.001 0.529 2 0 0.05
#> 4 0.001 0.529 0 0.5 1
#> 5 0.001 0.529 1 0.5 1
#> 6 0.001 0.529 2 0.5 0.95
#> 7 0.001 0.529 0 1 1
#> 8 0.001 0.529 1 1 1
#> 9 0.001 0.529 2 1 1
#> 10 0.01 0.517 0 0 0.05
#> # ℹ 17 more rows
# }