This article describes the requirements of the input data (effects table) needed to generate a dot-forest plot and a trade-off plot.
The example effects table that is used to generate the dot-forest plot and trade-off plot in the manuscript can be downloaded here [TO BE ADDED].
Depending on the type of an outcome (e.g., continuous, binary), the summary statistics used for reporting a treatment response (e.g., proportion, incidence rate, event rate) or a treatment effect (e.g., risk difference, relative risk, odds ratio), and the method for populating a confidence interval (e.g., calculated or supplied), different sets of columns are required in different situations. The description and requiredness of each column is listed as below.
The following variables are required columns, regardless of the conditions.
Factor: A character vector containing whether an outcome is a “Benefit” or a “Risk”.
Grouped_Outcome: A character vector containing the name of grouped outcomes, e.g., Infections.
Outcome: A character vector containing the name of outcomes, e.g., Herpes viral infections, upper respiratory tract infections.
Statistics: A character vector containing the summary statistics of outcomes, e.g., %, mean change from baseline.
Type: A character vector containing whether an outcome is a “Binary” or a “Continuous” variable.
Outcome_Status: A character vector containing whether an outcome is an “Identified” or a “Potential” outcome.
Filter: A character vector containing the filter for subgroup data, should be “None” if no filtre is applicable. Example: None; Sex.
Category: A character vector containing the category for filtering subgroup data, should be “All” if no filter is applicable. Example: All; Male, Female.
Trt1: A character vector containing the name of active treatments.
Trt2: A character vector containing controlled term “Placebo”.
Drug_Status: A character vector containing whether a treatment is an “Approved” or a “Test” drug.
Note that the variables Grouped_Outcome,
Statistics, and Outcome_Status are not
required for generating a dot-forest plot or a trade-off plot, but are
listed as required columns because they are key for generating a value
tree, which is a starting point for all subsequent benefit-risk
assessments. The variable Drug_Status is only required for
generating a trade-off plot.
The requiredness of the following variables are conditional on the type of an outcome (e.g., continuous, binary) and the summary statistics used for reporting a treatment response (e.g., proportion, event rate, incidence rate). They should be filled in only for the specific summary statistic related to an outcome.
Rate_Type: A numeric vector containing whether an AE rate is “EventRate” or “IncRate”. Required for risk outcomes displayed in exposure-adjusted event rate or incidence rate.
Prop1: A numeric vector containing the proportion in active treatment. Required for binary outcomes displayed in proportions; can be populated by nSub1/N1 if both nSub1 and N1 are provided.
IncRate1: A numeric vector containing the exposure-adjusted incidence rate per 100 PYs in active treatment. Required for risk outcomes displayed in exposure-adjusted incidence rates; can be populated by nSub1/PYAR1*100 if both nSub1 and PYAR1 are provided.
EventRate1: A numeric vector containing the exposure-adjusted event rate per 100 PYs in active treatment. Required for risk outcomes displayed in exposure-adjusted event rates; can be populated by nEvent1/PEY1*100 if both nEvent1 and PEY1 are provided.
Mean1: A numeric vector containing the mean in active treatment. Required for continuous outcomes.
Prop2: A numeric vector containing the proportion in comparator treatment. Required for binary outcomes displayed in proportions; can be populated by nSub2/N2 if both nSub1 and N1 are provided.
IncRate2: A numeric vector containing the exposure-adjusted incidence rate per 100 PYs in comparator treatment. Required for risk outcomes displayed in exposure-adjusted incidence rates; can be populated by nSub2/PYAR2*100 if both nSub1 and PYAR1 are provided.
EventRate2: A numeric vector containing the exposure-adjusted event rate per 100 PYs in comparator treatment. Required for risk outcomes displayed in exposure-adjusted event rates; can be populated by nEvent2/PEY2*100 if both nEvent1 and PEY1 are provided.
Mean2: A numeric vector containing the mean in comparator treatment. Required for continuous outcomes.
The requiredness of the following variables are conditional on the summary statistics used for reporting a treatment effect (e.g., risk difference, relative risk, odds ratio) and the method for populating a confidence interval (e.g., calculated or supplied), in addition to the type of an outcome (e.g., continuous, binary) and the summary statistics used for reporting a treatment response (e.g., proportion, event rate, incidence rate).
N1: An integer vector containing the total number of subjects in active treatment. Required when needing to calculate confidence intervals within the package for proportions.
100PYAR1: A numeric vector containing 100 patient-years at risk in active treatment. Required when needing to calculate confidence intervals within the app for exposure-adjusted incidence rates.
100PEY1: A vector containing 100 patient-years of exposure in active treatment. Required when needing to calculate confidence intervals within the app for exposure-adjusted event rates.
Sd1: A numeric vector containing the standard deviation in active treatment. Required when needing to calculate confidence intervals within the app for continuous outcomes; can be populated by Se1/SQRT(N1) if Se1 and N1 are provided.
N2: An integer vector containing the total number of subjects in comparator treatment. Required when needing to calculate confidence intervals within the package for proportions.
100PYAR2: A numeric vector containing 100 patient-years at risk in comparator treatment. Required when needing to calculate confidence intervals within the app for exposure-adjusted incidence rates.
100PEY2: A numeric vector containing 100 patient-years of exposure in comparator treatment. Required when needing to calculate confidence intervals within the app for exposure-adjusted event rates.
Sd2: A numeric vector containing the standard deviation in comparator treatment. Required when needing to calculate confidence intervals within the app for continuous outcomes; can be populated by Se2/SQRT(N2) if Se2 and N2 are provided.
Diff_LowerCI: A numeric vector containing the lower confidence interval for difference in proportions and continuous outcomes. Required when using supplied confidence intervals for difference in proportions and continuous outcomes.
Diff_UpperCI: A numeric vector containing the upper confidence interval for difference in proportions and continuous outcomes. Required when using supplied confidence intervals for difference in proportions and continuous outcomes.
Diff_IncRate_LowerCI: A numeric vector containing the lower confidence interval for difference in exposure-adjusted incidence rates. Required when using supplied confidence intervals for difference in exposure-adjusted incidence rates.
Diff_IncRate_UpperCI: A numeric vector containing the upper confidence interval for difference in exposure-adjusted incidence rates. Required when using supplied confidence intervals for difference in exposure-adjusted incidence rates.
Diff_EventRate_LowerCI: A numeric vector containing the lower confidence interval for difference in exposure-adjusted event rates. Required when using supplied confidence intervals for difference in exposure-adjusted event rates.
Diff_EventRate_UpperCI: A numeric vector containing the upper confidence interval for difference in exposure-adjusted event rates. Required when using supplied confidence intervals for difference in exposure-adjusted event rates.
RelRisk_LowerCI: A numeric vector containing the lower confidence interval for relative risk of binary outcomes. Required when using supplied confidence intervals for relative risk of binary outcomes.
RelRisk_UpperCI: A numeric vector containing the upper confidence interval for relative risk of binary outcomes. Required when using supplied confidence intervals for relative risk of binary outcomes.
OddsRatio_LowerCI: A numeric vector containing the lower confidence interval for odds ratio of binary outcomes. Required when using supplied confidence intervals for odds ratio of binary outcomes.
OddsRatio_UpperCI: A numeric vector containing the upper confidence interval for odds ratio of binary outcomes. Required when using supplied confidence intervals for odds ratio of binary outcomes.
The following variables are supplementary columns. They are not required but can be used to derive other columns. 39) nSub1: An integer vector containing the number of subjects with events in active treatment. Not required; can be used to calculate Prop1 by nSub1/N1.
Dur1: A numeric vector containing the duration of treatment in active treatment. Not required; can be used to estimate 100PYAR1 and 100PEY1.
nEvent1: An integer vector containing the number of events in active treatment. Not required; can be used to calculate EventRate1 by nEvent1/100PEY1.
Se1: A numeric vector containing the standard error in active treatment. Not required; can be used to calculate Sd1 by Se1*SQRT(N1).
nSub2: An integer vector containing the number of subjects with events in comparator treatment. Not required; can be used to calculate Prop2 by nSub2/N2.
Dur2: A numeric vector containing the duration of treatment in comparator treatment. Not required; can be used to estimate 100PYAR2 and 100PEY2.
nEvent2: An integer vector containing the number of events in comparator treatment. Not required; can be used to calculate EventRate2 by nEvent2/100PEY2.
Se2: A numeric vector containing the standard error in comparator treatment. Not required; can be used to calculate Sd2 by Se2*SQRT(N2).
The following variables are for documentation purpose and are not required.
MCDA_Weight: A numeric vector containing the MCDA weight.
Population: A character vector containing the population for the analysis (e.g., ITT, Safety Set).
Data_Source: A character vector containing the source of data (e.g., Reference CSR Table xxx).
Quality: A character vector containing the quality of data.
Notes: A character vector containing notes.
A decision tree is created to aid the understanding of the requiredness of columns that are conditionally required.
[TO BE ADDED]