PS-Integrated Methods for Incorporating RWE in Clinical Studies

Description

High-quality real-world data can be transformed into scientific real-world evidence for regulatory and healthcare decision-making using proven analytical methods and techniques. For example, propensity score (PS) methodology can be applied to select a subset of real-world data containing patients that are similar to those in the current clinical study in terms of baseline covariates, and to stratify the selected patients together with those in the current study into more homogeneous strata. Then, statistical methods such as the power prior approach or composite likelihood approach can be applied in each stratum to draw inference for the parameters of interest. This package provides functions that implement the PS-integrated real-world evidence analysis methods such as Wang et al. (2019) doi:10.1080/10543406.2019.1657133, Wang et al. (2020) doi:10.1080/10543406.2019.1684309, and Chen et al. (2020) doi:10.1080/10543406.2020.1730877.

This package provides R functions for conducting clinical studies with real-world evidence (RWE) incorporated in the study design and analysis.

PS-integrated power prior

We extend the Bayesian power prior approach for a single-arm study (the current study) to leverage external real-world data (RWD). We use propensity score methodology to pre-select a subset of real-world data containing patients who are similar to those in the current study in terms of covariates, and to stratify the selected patients together with those in the current study into more homogeneous strata. The power prior approach is then applied in each stratum to obtain stratum-specific posterior distributions, which are combined to complete the Bayesian inference for the parameters of interest.

PS-integrated composite likelihood

A propensity score-integrated composite likelihood (PSCL) approach is developed for cases in which the control arm of a two-arm randomized controlled trial (RCT) (treated vs. control) is augmented with patients from real-world data (RWD) containing both clinical outcomes and covariates at the patient level. The PSCL approach first estimates the propensity score for every patient as the probability of the patient being in the RCT rather than the RWD, and then stratifies all patients into strata based on the estimated propensity scores. Within each propensity score stratum, a composite likelihood function is specified and utilized to down-weight the information contributed by the RWD source. Estimates of the stratum-specific parameters are obtained by maximizing the composite likelihood function. These stratum-specific estimates are then combined to obtain an overall population-level estimate of the parameter of interest.

Author(s)

Maintainer: Wei-Chen Chen wccsnow@gmail.com

Authors:

• Chenguang Wang

Other contributors:

• Trustees of Columbia University (tools/make_cpp.R, R/stanmodels.R) [copyright holder]

References

Chen WC, Wang C, Li H, Lu N, Tiwari R, Xu Y, Yue LQ. Propensity score-integrated composite likelihood approach for augmenting the control arm of a randomized controlled trial by incorporating real-world data. *Journal of Biopharmaceutical Statistics*. 2020; 30(3):508-520.

Wang C, Lu N, Chen WC, Li H, Tiwari R, Xu Y, Yue LQ. Propensity score-integrated composite likelihood approach for incorporating real-world evidence in single-arm clinical studies. *Journal of Biopharmaceutical Statistics*. 2020; 30(3):495-507.

Wang C, Li H, Chen WC, Lu N, Tiwari R, Xu Y, Yue LQ. Propensity score-integrated power prior approach for incorporating real-world evidence in single-arm clinical studies. *Journal of Biopharmaceutical Statistics*. 2019; 29(5):731-748.

See Also

Useful links:

• https://github.com/olssol/psrwe

• Report bugs at https://github.com/olssol/psrwe/issues

Example dataset

Description

Example dataset of a single arm study.

Usage

data(ex_dta)

Format

A data frame with the following variables:

Group

current, rwd

Y_Bin

Binary outcome

Y_Con

Continuous outcome

Y_Surv

Survival outcome in days

Status

Event status (0=alive, 1=dead)

V1

Baseline covariate

V2

Baseline covariate

V3

Baseline covariate

V4

Baseline covariate

V5

Baseline covariate

V6

Baseline covariate

V7

Baseline covariate

Example dataset

Description

Example dataset of a randomized study.

Usage

data(ex_dta_rct)

Format

A data frame with the following variables:

Group

current, rwd

Arm

control, treatment

Y_Con

Continuous outcome

Y_Surv

Survival outcome in days

Status

Event status (0=alive, 1=dead)

V1

Baseline covariate

V2

Baseline covariate

V3

Baseline covariate

V4

Baseline covariate

V5

Baseline covariate

V6

Baseline covariate

V7

Baseline covariate

Distance between two distributions

Description

Calculate difference measures using different metrics.

Usage

get_distance(
  cov0,
  cov1,
  metric = c("ovl", "ksd", "astd", "std", "abd", "ley", "mhb", "omkss")
)

Arguments

Value

A real value of the distance.

Examples

x <- rnorm(100,  mean = 0, sd = 1)
y <- rnorm(1000, mean = 1, sd = 2)
get_distance(x, y, "ovl")
get_distance(x, y, "abd")

Plot PS distributions

Description

S3 method for visualizing PS adjustment

Usage

## S3 method for class 'PSRWE_DTA'
plot(x, plot_type = c("ps", "balance", "diff"), ...)

Arguments

Value

A plot of class in ggplot2

Plot PS distributions

Description

S3 method for visualizing PS adjustment based on matching.

Usage

## S3 method for class 'PSRWE_DTA_MAT'
plot(x, ...)

Arguments

Value

A plot of class in ggplot2

See Also

plot.PSRWE_DTA

Plot estimation results for power prior approach

Description

S3 method plotting estimation results

Usage

## S3 method for class 'PSRWE_RST'
plot(x, ...)

Arguments

Value

A plot of class in ggplot2

Print borrow information

Description

Print summary information of borrowing

Usage

## S3 method for class 'PSRWE_BOR'
print(x, ...)

Arguments

Value

A list from x$Borrow

See Also

psrwe_borrow

Print PS estimation results

Description

Print summary information of PS estimation results

Usage

## S3 method for class 'PSRWE_DTA'
print(x, ...)

Arguments

Value

A list from summary(x) with additional information

See Also

summary.PSRWE_DTA

Print PS estimation results

Description

Print summary information of PS estimation results

Usage

## S3 method for class 'PSRWE_DTA_MAT'
print(x, ...)

Arguments

Value

A list from summary(x) with additional information

See Also

summary.PSRWE_DTA_MAT

Print estimation results

Description

Print summary information of outcome mean estimation results

Usage

## S3 method for class 'PSRWE_RST'
print(x, ...)

Arguments

Value

None (invisible NULL)

See Also

summary.PSRWE_RST

Print outcome analysis results

Description

Print detailed information of outcome analysis results

Usage

## S3 method for class 'PSRWE_RST_OUTANA'
print(x, show_details = FALSE, show_rct = FALSE, show_pred_tps = NULL, ...)

Arguments

Value

None (invisible NULL)

Get number of subjects borrowed from each statum

Description

Based on PS distances or number of current control subjects, split the total number of subjects to be borrowed from the external data source for each stratum

Usage

psrwe_borrow(
  dtaps,
  total_borrow,
  method = c("distance", "inverse_distance", "n_current", "n_external"),
  .drop_arg_fml = FALSE,
  ...
)

Arguments

Value

A class PSRWE_BORR list. It appends the following items to the dtaps:

Proportion

Proportion splitting the number of total borrow among strata.

N_Borrow

The number of subjects to be borrowed in each stratum.

Alpha

Weight parameter value in each stratum.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
                     v_covs = paste("V", 1:7, sep = ""),
                     v_grp = "Group",
                     cur_grp_level = "current")

ps_borrow <- psrwe_borrow(total_borrow = 20, dta_ps)
ps_borrow

## Use different similarity metric
ps_borrow_omkss <- psrwe_borrow(total_borrow = 20, dta_ps,
                                 metric = "omkss")
ps_borrow_omkss

Confidence/Credible Interval for PS-Integrated Estimation

Description

Estimate the confidence/credible interval for the PS-integrated approach.

Usage

psrwe_ci(
  dta_psrst,
  method_ci = c("wald", "wilson"),
  conf_int = 0.95,
  conf_type = c("log_log", "plain"),
  ...
)

Arguments

Details

method_ci = "wilson" is for binary outcomes only. conf_type = "log_log" is for ps_km only.

Value

A list with class name PSRWE_EST.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
ps_rst <- psrwe_compl(ps_borrow, v_outcome = "Y_Con")
rst <- psrwe_ci(ps_rst)
rst

PS-Integrated Composite Likelihood Estimation

Description

Estimate the mean of the outcome based on PS-integrated composite likelihood approach. Variance is estimated by the Jackknife method. Applies to the case when there is only one external data source.

Usage

psrwe_compl(
  dta_psbor,
  v_outcome = "Y",
  outcome_type = c("continuous", "binary"),
  stderr_method = c("jk", "sjk", "cjk", "sbs", "cbs", "none"),
  n_bootstrap = 200,
  ...
)

Arguments

Details

stderr_method include jk as default using the Jackknife method within each stratum, sjk for simple Jackknife method for combined estimates such as point estimates in single-arm or treatment effects in RCT, or cjk for complex Jackknife method including refitting PS model, matching, trimming, calculating borrowing parameters, and combining overall estimates. Note that sjk may take a while longer to finish and cjk will take even much longer to finish. The sbs and cbs is for simple and complex Bootstrap methods.

Value

A data frame with class name PSRWE_RST. It contains the composite estimation of the mean for each stratum as well as the jackknife estimation for each subject. The results can be further summarized by its S3 method summary. The results can also be analyzed by psrwe_outana for outcome analysis and inference.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
rst       <- psrwe_compl(ps_borrow, v_outcome = "Y_Con")
rst

Estimate propensity scores

Description

Estimate propensity scores using logistic regression or random forest model.

Usage

psrwe_est(
  data,
  ps_fml = NULL,
  ps_method = c("logistic", "randomforest"),
  v_covs = "V1",
  v_grp = "Group",
  cur_grp_level = 1,
  v_arm = NULL,
  ctl_arm_level = NULL,
  stra_ctl_only = TRUE,
  nstrata = 5,
  trim_ab = c("both", "above", "below", "none"),
  .drop_arg_fml = FALSE,
  ...
)

Arguments

Value

A list of class PSRWE_DAT with items:

data

Original data with column _ps_ for estimated PS scores and _strata_ for PS stratum added.

ps_fml

PS formula for estimated PS scores.

is_rct

Whether the current study is a randomized study.

nstrata

Number of strata.

Examples

data(ex_dta)
psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")

Inference for the PS-Integrated Estimation

Description

Inference for the PS-integrated approach.

Usage

psrwe_infer(
  dta_psrst,
  alternative = c("less", "greater", "two_sided"),
  mu = 0,
  method_pval = c("wald", "score", "score_weighted"),
  ...
)

Arguments

Value

A list with class name PSRWE_EST.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
ps_rst <- psrwe_compl(ps_borrow, v_outcome = "Y_Con")
rst <- psrwe_infer(ps_rst)
rst

PS matching

Description

Match patients in the external data source with patients in the current study based on PS using the nearest neighbor method.

Usage

psrwe_match(
  dta_ps,
  ratio = 3,
  strata_covs = NULL,
  caliper = 1,
  seed = NULL,
  method = c("nnwor", "optm"),
  .drop_arg_fml = FALSE,
  ...
)

Arguments

Value

A list of class PSRWE_DTA_MAT with items:

data

Original data with column _ps_ for estimated PS scores, match_id for matched current study subject ID, and _strata_ for PS stratum added.

ps_fml

PS formula for estimated PS scores.

nstrata

Number of strata.

ratio

Matching ratio.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
                     v_covs = paste("V", 1:7, sep = ""),
                     v_grp = "Group",
                     cur_grp_level = "current")

## With neighest neighbor
.remark_nnmatch <- function() {
  dta_ps_mat <- psrwe_match(dta_ps, ratio = 2, strata_covs = "V1",
                            seed = 123)
  dta_ps_mat
}
## Unmark below to run
# .remark_nnmatch()

## With optmatch
.remark_optmatch <- function() {
    warning("The optmatch package may restrict use (academic license).")
    dta_ps_opt <- psrwe_match(dta_ps, ratio = 2,
                               strata_covs = "V1",
                               method = "optm", seed = 123)
    dta_ps_opt
}
## Unmark below to run if optmatch is available.
# .remark_optmatch()

Outcome Analysis for PS-Integrated Estimation

Description

Report outcome analysis for the PS-integrated approach.

Usage

psrwe_outana(
  dta_psrst,
  method_ci = c("wald", "wilson"),
  conf_type = c("log_log", "plain"),
  conf_int = 0.95,
  alternative = c("less", "greater", "two_sided"),
  mu = 0,
  method_pval = c("wald", "score", "score_weighted"),
  ...
)

Arguments

Details

This function is mainly for summarizing and reporting the outcome analysis for the PS-integrated estimation. The input dta_psrst can be generated from the functions psrwe_powerp, psrwe_compl, and psrwe_survkm. See the functions psrwe_ci and psrwe_infer for the options of outcome analyses.

Value

A list with class name PSRWE_EST_OUTANA.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
ps_rst <- psrwe_compl(ps_borrow, v_outcome = "Y_Con")
rst <- psrwe_outana(ps_rst)
rst

Get posterior samples based on PS-power prior approach

Description

Draw posterior samples of the parameters of interest for the PS-power prior approach

Usage

psrwe_powerp(
  dta_psbor,
  v_outcome = "Y",
  outcome_type = c("continuous", "binary"),
  prior_type = c("fixed", "random"),
  ...,
  seed = NULL
)

Arguments

Value

A class PSRWE_RST list with the following objects

Observed

Observed mean and SD of the outcome by group, arm and stratum

Control

A list of estimated mean and SD of the outcome by stratum in the control arm

Treatment

A list of estimated mean and SD of the outcome by stratum in the treatment arm for RCT

Effect

A list of estimated mean and SD of the treatment effect by stratum for RCT

Borrow

Borrowing information from dta_psbor

stan_rst

Result from STAN sampling

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
rst <- psrwe_powerp(ps_borrow, v_outcome = "Y_Con", seed = 123)

Get posterior samples based on PS-power prior approach (WATT)

Description

Draw posterior samples of the parameters of interest for the PS-power prior approach with weights of ATT (WATT)

Usage

psrwe_powerp_watt(
  dta_psbor,
  v_outcome = "Y",
  outcome_type = c("continuous", "binary"),
  mcmc_method = c("rstan", "analytic", "wattcon"),
  tau0_method = c("Wang2019", "weighted"),
  ipw_method = c("Heng.Li", "Xi.Ada.Wang"),
  ...,
  seed = NULL
)

Arguments

Value

A class PSRWE_RST list with the following objects

Observed

Observed mean and SD of the outcome by group, arm and stratum

Control

A list of estimated mean and SD of the outcome by stratum in the control arm

Treatment

A list of estimated mean and SD of the outcome by stratum in the treatment arm for RCT

Effect

A list of estimated mean and SD of the treatment effect by stratum for RCT

Borrow

Borrowing information from dta_psbor

stan_rst

Result from STAN sampling

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current",
       nstrata = 1)
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
rst <- psrwe_powerp_watt(ps_borrow, v_outcome = "Y_Bin", seed = 123)

PS-Integrated Kaplan-Meier Estimation

Description

Estimate the mean of a survival outcome at a given time point based on PS-integrated Kaplan-Meier approach. Variance can be estimated by the Jackknife methods. Apply to the case when there is only one external data source.

Usage

psrwe_survkm(
  dta_psbor,
  pred_tp,
  v_time = "time",
  v_event = "event",
  stderr_method = c("naive", "jk", "sjk", "cjk", "sbs", "cbs", "none"),
  n_bootstrap = 200,
  ...
)

Arguments

Details

stderr_method includes naive as default which mostly follows the Greenwood formula, jk using the Jackknife method within each stratum, sjk using simple Jackknife method for combined estimates such as point estimates in single-arm or treatment effects in RCT, or cjk for complex Jackknife method including refitting PS model, matching, trimming, calculating borrowing parameters, and combining overall estimates. Note that sjk may take a while longer to finish and cjk will take even longer to finish. The sbs and cbs is for simple and complex Bootstrap methods.

Value

A data frame with class name PSRWE_RST. It contains the composite estimation of the mean for each stratum as well as the Jackknife estimation. The results can be further summarized by its S3 method summary. The results can also be analyzed by psrwe_outana for outcome analysis and inference.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
       v_covs = paste("V", 1:7, sep = ""),
       v_grp = "Group",
       cur_grp_level = "current")
ps_borrow <- psrwe_borrow(total_borrow = 30, dta_ps)
rst       <- psrwe_survkm(ps_borrow,
                          pred_tp = 365,
                          v_time = "Y_Surv",
                          v_event = "Status")
rst

PS-Integrated Log-Rank Test For Comparing Time-to-event Outcomes

Description

Log-rank test evaluates a two-arm RCT for up to a given time point. Variance can be estimated by Jackknife methods. Apply to the case when there is only one external data source and two-arm RCT.

Usage

psrwe_survlrk(
  dta_psbor,
  pred_tp,
  v_time = "time",
  v_event = "event",
  stderr_method = c("naive", "jk", "sjk", "cjk", "sbs", "cbs", "none"),
  n_bootstrap = 200,
  ...
)

Arguments

Details

stderr_method includes naive as default which mostly follows the Greenwood formula, jk using the Jackknife method within each stratum, sjk using simple Jackknife method for combined estimates such as point estimates in single-arm or treatment effects in RCT, or cjk for complex Jackknife method including refitting PS model, matching, trimming, calculating borrowing parameters, and combining overall estimates. Note that sjk may take a while longer to finish and cjk will take even longer to finish. The sbs and cbs is for simple and complex Bootstrap methods.

Value

A data frame with class name PSRWE_RST_TESTANA. It contains the test statistics of each stratum as well as the Jackknife estimation. The results can be further summarized by its S3 method summary. The results can also be analyzed by psrwe_outana for outcome analysis and inference.

Examples

data(ex_dta_rct)
dta_ps_rct <- psrwe_est(ex_dta_rct,
                        v_covs = paste("V", 1:7, sep = ""),
                        v_grp = "Group", cur_grp_level = "current",
                        v_arm = "Arm", ctl_arm_level = "control",
                        ps_method = "logistic", nstrata = 5,
                        stra_ctl_only = FALSE)
ps_bor_rct <- psrwe_borrow(dta_ps_rct, total_borrow = 30)
rst_lrk <- psrwe_survlrk(ps_bor_rct,
                         pred_tp = 365, 
                         v_time = "Y_Surv",
                         v_event = "Status")
rst_lrk

PS-Integrated Restricted Mean Survival Time (RMST) Test For Comparing Time-to-event Outcomes

Description

RMST test evaluates a two-arm RCT for up to a given time point. Variance can be estimated by the Jackknife method. Apply to the case when there is only one external data source and two-arm RCT.

Usage

psrwe_survrmst(
  dta_psbor,
  pred_tp,
  v_time = "time",
  v_event = "event",
  stderr_method = c("naive", "jk", "sjk", "cjk", "sbs", "cbs", "none"),
  n_bootstrap = 200,
  ...
)

Arguments

Details

stderr_method includes naive as default which mostly follows the Greenwood formula, jk using the Jackknife method within each stratum, sjk using simple Jackknife method for combined estimates such as point estimates in single-arm or treatment effects in RCT, or cjk for complex Jackknife method including refitting PS model, matching, trimming, calculating borrowing parameters, and combining overall estimates. Note that sjk may take a while longer to finish and cjk will take even longer to finish. The sbs and cbs is for simple and complex Bootstrap methods.

Value

A data frame with class name PSRWE_RST_TESTANA. It contains the test statistics of each stratum as well as the Jackknife estimation. The results can be further summarized by its S3 method summary. The results can also be analyzed by psrwe_outana for outcome analysis and inference.

Examples

data(ex_dta_rct)
dta_ps_rct <- psrwe_est(ex_dta_rct,
                        v_covs = paste("V", 1:7, sep = ""),
                        v_grp = "Group", cur_grp_level = "current",
                        v_arm = "Arm", ctl_arm_level = "control",
                        ps_method = "logistic", nstrata = 5,
                        stra_ctl_only = FALSE)
ps_bor_rct <- psrwe_borrow(dta_ps_rct, total_borrow = 30)
rst_rmst <- psrwe_survrmst(ps_bor_rct,
                           pred_tp = 365,
                           v_time = "Y_Surv",
                           v_event = "Status")
rst_rmst

Composite Likelihood Estimation

Description

Estimate parameter of interest based on composite likelihood for a single PS stratum

Usage

rwe_cl(
  dta_cur,
  dta_ext,
  n_borrow = 0,
  outcome_type = c("continuous", "binary"),
  equal_sd = TRUE
)

Arguments

Value

Maximum composite likelihood estimator of the mean

Examples

x <- rnorm(100,  mean = 0, sd = 1)
y <- rnorm(1000, mean = 1, sd = 2)
rwe_cl(x, y, n_borrow = 20, equal_sd = FALSE)

Create strata

Description

Cut a sequence of numbers into bins.

The cut points are chosen such that there will be equal numbers in each bin for x. By default, values of y that are outside the range of x will be excluded from the bins, unless they are in the keep_inx.

Usage

rwe_cut(
  x,
  y = x,
  breaks = 5,
  keep_inx = NULL,
  trim_ab = c("both", "above", "below", "none")
)

Arguments

Value

A vector of stratum assignment for y. The y's that are outside the range of x and not in keep_inx are assigned NA in the result.

Examples

x <- rnorm(100,  mean = 0, sd = 1)
y <- rnorm(1000, mean = 1, sd = 2)
rwe_cut(x, y, breaks = 5)

Kaplan-Meier Estimation

Description

Estimate survival probability based on Kaplan-Meier estimator for a single PS stratum

Usage

rwe_km(
  dta_cur,
  dta_ext = NULL,
  n_borrow = 0,
  pred_tps = NULL,
  stderr_method = "naive"
)

Arguments

Value

Estimation of survival probabilities at time pred_tps

Log-rank Estimation

Description

Estimate log-rank estimates for a single PS stratum

Usage

rwe_lrk(
  dta_cur,
  dta_ext,
  dta_cur_trt,
  n_borrow = 0,
  pred_tps = NULL,
  stderr_method = "naive"
)

Arguments

Value

Estimation of log-rank estimates at time pred_tps

RMST Estimation

Description

Estimate RMST estimates for a single PS stratum

Usage

rwe_rmst(
  dta_cur,
  dta_ext,
  dta_cur_trt,
  n_borrow = 0,
  pred_tps = NULL,
  stderr_method = "naive"
)

Arguments

Value

Estimation of RMST estimates at time pred_tps

Call STAN models

Description

Call STAN models. Called by psrwe_powerp.

Usage

rwe_stan(
  lst_data,
  stan_mdl = c("powerps", "powerpsbinary", "powerp", "powerps_wattcon"),
  chains = 4,
  iter = 2000,
  warmup = 1000,
  control = list(adapt_delta = 0.95),
  ...
)

Arguments

Value

Result from STAN sampling

Summarize PS estimation and stratification results

Description

Get the number of subjects and the distances of PS distributions for each PS stratum.

Usage

## S3 method for class 'PSRWE_DTA'
summary(
  object,
  metric = c("ovl", "ksd", "std", "abd", "ley", "mhb", "omkss"),
  min_n0 = 10,
  ...
)

Arguments

Value

A list with columns:

Summary

A data frame with Stratum, number of subjects in RWD, current study, number of subjects in control and treatment arms for RCT studies, and distance in PS distributions.

Overall

A data frame with the overall number of not-trimmed subjects in RWD, number of patients in the current study, number of subjects in control and treatment arms for RCT studies, and distance in PS distributions.

N

Vector of total number of total RWD patients, number of trimmed RWD patients, and total number of current study patients.

ps_fml

PS model.

Distance_metric

Metric used for calculating the distance.

Examples

data(ex_dta)
dta_ps <- psrwe_est(ex_dta,
                     v_covs = paste("V", 1:7, sep = ""),
                     v_grp = "Group",
                     cur_grp_level = "current")
dta_ps

## With different similarity metric
print(dta_ps, metric = "omkss")
dta_ps_sum <- summary(dta_ps, metric = "omkss")

Summarize PS estimation and matching results

Description

Get number of subjects for each PS stratum.

Usage

## S3 method for class 'PSRWE_DTA_MAT'
summary(object, ...)

Arguments

Value

A list with columns:

Summary

A data frame with Stratum (defined by covariates), number of subjects in RWD, current study, number of subjects in control and treatment arms for RCT studies.

Overall

A data frame with the overall number of not-trimmed subjects in RWD, number of patients in the current study, number of subjects in control and treatment arms for RCT studies.

N

Vector of total number of total RWD patients, number of trimmed RWD patients, total number of current study patients, number of current control patients with less than ratio matched RWD subjects.

ps_fml

PS model.

N_Match

Number of current control subjects matched with ratio, 0 and other number of RWD subjects.

ratio

Matching ratio.

Summarize overall estimation results

Description

S3 method summarizing overall estimation results

Usage

## S3 method for class 'PSRWE_RST'
summary(object, ...)

Arguments

Value

A list with data frames for the borrowing and estimation results.

Summary outcome analysis results

Description

Summary information of outcome analysis results

Usage

## S3 method for class 'PSRWE_RST_OUTANA'
summary(object, pred_tps = NULL, ...)

Arguments

Value

A list of class PSRWE_RST_OUTANA with additiona information