PVdagger

The goal of PVdagger is to support the visualization and sharing of causal assumptions in pharmacovigilance, using Directed Acyclic Graphs.

Installation

You can install the development version of PVdagger from GitHub with:

# install.packages("pak")
pak::pak("PVverse/pv_dagger")

Example

This is how to draw the causal inquiry, together with the two possible scenarios.

library(PVdagger)
#> Loading required package: DiagrammeR
create_dag("D","E", scenario ="inquiry")
create_dag("D", "E", scenario = "causal")
create_dag("D", "E", scenario = "non-causal")

This is how to draw confounders and colliders

create_dag("D", "E", label_inquiry = "", scenario = "non-causal",
           confounder_path = list(nodes = c("C"), signs = c("",""),
                                  label=""))
create_dag("D", "E", label_inquiry = "", scenario = "non-causal",
           collider_path = list(nodes = c("F"), signs = c("",""),
                                label=""))

This is how to add measurements and draw ascertainment bias

create_dag("D", "E", scenario = "inquiry", add_measurements = TRUE,
           label_inquiry ="")
create_dag("D", "E", scenario = "non-causal", add_measurements = TRUE,
           ascertainment_drug = "",
           label_inquiry ="")
create_dag("D", "E", scenario = "non-causal", add_measurements = TRUE,
           ascertainment_event = "",
           label_inquiry ="")

This is how to add reporting

create_dag("D", "E", scenario = "inquiry", add_measurements = TRUE, add_reporting = TRUE,
           label_inquiry ="")

This is how to add reporting biases

create_dag("D", "E", scenario = "non-causal",notoriety_bias = "Notoriety", add_measurements = TRUE,
           label_inquiry ="Causal Inquiry",drug_competition_bias = "D2",
           event_competition_bias="E2",
           background_dilution =  list(drug="D3",event="E3"))

“”