Learn R Programming

WARDEN: Workflows for health technology Assessments in R using Discrete EveNts

Introduction

WARDEN is a user-friendly package that facilitates the use of discrete event simulations with or without resource constraints for cost-effectiveness analysis. The package supports a flexible, transparent, practical approach to discrete event simulation while keeping an acceptable performance.

The current version supports:

  • Discrete event simulation models, Markov/semi-Markov models and hybrid models using parallel and non-parallel engines
  • Constrained and unconstrained models (with or without interactivity between patients)
  • Seamlessly integrating data.frames and other objects into the model
  • Delayed execution of the main inputs to facilitate readability of the model
  • Implementation of structural and parameter uncertainty
  • Helper functions to facilitate drawing of time to events and the use of hazard ratios, as well as other functions to facilitate transparency
  • Performing cost-effectiveness and uncertainty analysis

It is recommended that the user checks the vignettes, first the simple Sick-Sicker-Dead model and then the more complex model for early breast cancer. The markov example shows how to run a cohort Markov model while using the same modeling framework. Similarly, a simulation based Markov model could be run. Structural and parametric uncertainty are explored in the corresponding vignette. The IPD vignette shows how WARDEN can be used when individual patient data is available. It is also suggested to check the constrained resources DES example on the website.

Documentation

Have a look at the package home site for more details on documentation and specific tutorials.

For more details on the code, check our Github repository, or our CRAN site.

Installation

WARDEN can now be installed directly from CRAN or this repo via

install.packages("WARDEN") #CRAN version 

# install.packages("devtools")
devtools::install_github("jsanchezalv/WARDEN", ref="main") #github version

Citation

If you use WARDEN, please contact the authors for the most up to date appropriate citation.

WARDEN

Copy Link

Version

Install

install.packages('WARDEN')

Monthly Downloads

230

Version

2.0.1

License

GPL (>= 3)

Maintainer

Javier Sanchez Alvarez

Last Published

December 11th, 2025

Functions in WARDEN (2.0.1)

add_tte

Define events and the initial event time
disc_cycle_v

Cycle discounting for vectors
disc_cycle

Cycle discounting
pick_val_v

Select which values should be applied in the corresponding loop for several values (vector or list).
pop_and_return_event

Pop and return the next event
modify_event

Modify or add events for a patient
modify_item

Modify the value of existing items
qcond_norm

Conditional quantile function for normal distribution
qcond_weibull

Conditional quantile function for weibull distribution
pcond_gompertz

Survival Probaility function for conditional Gompertz distribution (lower bound only)
pick_psa

Helper function to create a list with random draws or whenever a series of functions needs to be called. Can be implemented within pick_val_v.
queue_empty

Check if the event queue is empty
queue_size

Get the Size of the Event Queue
qcond_gamma

Conditional quantile function for gamma distribution
qcond_gompertz

Quantile function for conditional Gompertz distribution (lower bound only)
draw_tte

Draw a time to event from a list of parametric survival functions
disc_instant_v

Calculate instantaneous discounted costs or qalys for vectors
evpi_des

Calculate the Expected Value of Perfect Information (EVPI) for a DES model with a PSA result
rpoisgamma_rcpp

Draw time to event (tte) from a Poisson or Poisson-Gamma (PG) Mixture/Negative Binomial (NB) Process using C++
rpoisgamma

Draw time to event (tte) from a Poisson or Poisson-Gamma (PG) Mixture/Negative Binomial (NB) Process
extract_psa_result

Extract PSA results from a treatment
qtimecov

Draw time-to-event with time-dependent covariates and luck adjustment
resource_discrete

Create a discrete resource
queue_create

Create a New Event Queue
rgamma_mse

Draw from a gamma distribution based on mean and se
with_write_flags_lang

Evaluate a language object with write-tracking via active bindings
has_event

Check if a patient has a specific event in the queue
luck_adj

Perform luck adjustment
rdirichlet_prob

Draw from a dirichlet distribution based on mean transition probabilities and standard errors
summary_results_sim

Summary of PSA outputs for a treatment
sens_iterator

Create an iterator based on sens of the current iteration within a scenario (DSA)
rdirichlet

Draw from a dirichlet distribution based on number of counts in transition. Adapted from brms::rdirichlet
shared_input

Shared input object
tte.df

Example TTE IPD data
discrete_resource_clone

Clone independent discrete resources
get_event

Get a specific event time
print.resource_discrete

Print method for resource_discrete
pop_event

Remove the next event from the queue
next_event_pt

Get the next events in the queue for a specific patient
next_event

Get the next events in the queue
replicate_profiles

Replicate profiles data.frame
remove_event

Remove events for a patient
.resource_discrete_from_xptr

Creates a discrete resource from an external pointer
qbeta_mse

Draw from a beta distribution based on mean and se (quantile)
modify_item_seq

Modify the value of existing items
new_event

Add events to the queue for a patient
summary_results_det

Deterministic results for a specific treatment
summary_results_sens

Summary of sensitivity outputs for a treatment
random_stream

Creates an environment (similar to R6 class) of random uniform numbers to be drawn from
rbeta_mse

Draw from a beta distribution based on mean and se
qcond_lnorm

Conditional quantile function for lognormal distribution
qcond_llogis

Conditional quantile function for loglogistic distribution
qcond_weibullPH

Conditional quantile function for WeibullPH (flexsurv)
rcond_gompertz

Draw from a conditional Gompertz distribution (lower bound only)
qgamma_mse

Use quantiles from a gamma distribution based on mean and se
rcond_gompertz_lu

Draw from a conditional Gompertz distribution (lower and upper bound)
qcond_exp

Conditional quantile function for exponential distribution
run_sim

Run the simulation
run_sim_parallel

Run simulations in parallel mode (at the simulation level)
add_item2

Define parameters that may be used in model calculations (uses expressions)
ceac_des

Calculate the cost-effectiveness acceptability curve (CEAC) for a DES model with a PSA result
cond_dirichlet

Calculate conditional dirichlet values
create_indicators

Creates a vector of indicators (0 and 1) for sensitivity/DSA analysis
adj_val

Adjusted Value Calculation
add_reactevt

Define the modifications to other events, costs, utilities, or other items affected by the occurrence of the event
add_item

Define or append model inputs
cond_mvn

Calculate conditional multivariate normal values
ast_as_list

Transform a substituted expression to its Abstract Syntax Tree (AST) as a list
disc_ongoing

Calculate discounted costs and qalys between events
disc_ongoing_v

Calculate discounted costs and qalys between events for vectors
disc_instant

Calculate instantaneous discounted costs or qalys
extract_elements_from_list

Extracts items and events by looking into assignments, modify_event and new_event
extract_from_reactions

Extract all items and events and their interactions from the event reactions list