Skip to content

Bayer-Group/BIC-BReasy

Repository files navigation

BReasy

The BReasy R Shiny app provides a handy platform for structured benefit-risk assessment using clinical study or pooled data. With the provided forest plot estimates for all relevant efficacy and safety outcomes can be shown in one graph. In the graphical display the presented outcomes can be separated into efficacy and safety. To enable a structured and transparent approach to assessing the benefit-risk profile, the outcomes shown in the forest plot are based on the same analysis sets and data scopes. Furthermore, a presentation by study (in case of pooled data), stratum and subgroups is possible to assess the benefit-risk profile in subpopulations of interest.

The following estimates for comparison can be presented with the BReasy Shiny app:

  • Hazard Ratios;
  • Excess number of subjects, for example based on Kaplan-Meier differences or risk differences;
  • Odds Ratios;
  • Risk Differences;
  • Relative risks.

An optional display of the following data is possible to be added to the forest plot:

  • The actual estimates displayed in the forest plot;
  • Event and patient counts;
  • Number needed to treat (NNT) / Number needed to harm (NNH).

Additionally, the BReasy app includes the possibility of creating a value tree showing the relevant benefit and risk criteria. Benefit criteria will be colour-coded as green, and the risk criteria will be colour-coded as red. The interactively created value tree can be saved in html format, please see below for an example.

Installation

You can install and load the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("Bayer-Group/BIC-BReasy")
library(breasy)

and start the app with:

#run_app()

Data Manual

BReasy is designed to upload a CSV file containing effect estimates for outcomes of interest.

File Structure

When uploading a CSV file the following formats are allowed and can be chosen in the app:

  • Separator between variables
    • Comma, Semicolon or Tab
  • Quote highlighting variables
    • None, Double Quote (“,‘“’,”) or Single Quote (’)
  • Decimal character
    • None, Point (.) or Comma (,)

In case an incorrect setting is chosen the app might run into an error.

File Format

In order to use the BReasy, the CSV file has to include the following variables:

<style> table, th, td { border: 1px solid black; } </style>
Variable: Description: Type:
Required Variables ESTIMATE Type of estimate as basis for comparison, e.g. crude incidence or Kaplan-Meier cumulative incidence Character
ANALYSIS_SET Analyis set Character
OUTCOME Outcome of interest Character
EFFECT_xx Variable containing the value of the estimate used for comparison xx indicates the estimate chosen: ARD = Risk difference EXCESS = Excess number of subjects HR = Hazard Ratio OR = Odds Ratio RR = Relative risk Numeric
LOWERxx Corresponding lower Confidence interval bound with xx indicating 1-alpha Numeric
UPPERxx Corresponding upper Confidence interval bound with xx indicating 1-alpha Numeric
Optional variables NUMBER_EVENTS_VERUM Number of events in verum group for the combination of required variables estimate, outcome and analysis set and optional variables study, data scope, stratum, subgroup category Integer
NUMBER_PATIENTS_VERUM Number of patients in the verum group for the combination of required variables analysis set and optional variables study, stratum and subgroup category Integer
NUMBER_EVENTS_COMP Number of events in comparator group for the combination of required variables estimate, outcome and analysis set and optional variables study, data scope, stratum, subgroup category Integer
NUMBER_PATIENTS_COMP Number of patients in the comparator group for the combination of required variables analysis set and optional variables study, stratum and subgroup category Integer
STUDY/TRIALNO Study identifier Character
DATA_SCOPE/AVISIT Data scope, e.g. treatment-emergent Character
SUBGROUP Subgroup of Interest - Values of ‘Any’, ‘All’ or ‘Overall’ or ‘None’ can be used for overall estimates Character
SUBLEVEL Subgroup category corresponding to SUBGROUP Character
STRATUM Stratum Character
NNT Number needed to treat (NNT) / Number needed to harm (NNH) Integer

If the data is stored in more than one file of the same format, it is possible to upload several CSV files into BReasy.

The maximal file size is 40 MB.