| Title: | Matching-Adjusted Indirect Comparison |
|---|---|
| Description: | A generalised workflow for generation of subject weights to be used in Matching-Adjusted Indirect Comparison (MAIC) per Signorovitch et al. (2012) <doi:10.1016/j.jval.2012.05.004>, Signorovitch et al (2010) <doi:10.2165/11538370-000000000-00000>. In MAIC, unbiased comparison between outcomes of two trials is facilitated by weighting the subject-level outcomes of one trial with weights derived such that the weighted aggregate measures of the prognostic or effect modifying variables are equal to those of the sample in the comparator trial. The functions and classes included in this package wrap and abstract the process demonstrated in the UK National Institute for Health and Care Excellence Decision Support Unit (NICE DSU)'s example (Phillippo et al, (2016) [see URL]), providing a repeatable and easily specifiable workflow for producing multiple comparison variable sets against a variety of target studies, with preprocessing for a number of aggregate target forms (e.g. mean, median, domain limits). |
| Authors: | Rob Young [aut, cre] |
| Maintainer: | Rob Young <[email protected]> |
| License: | GPL-3 |
| Version: | 0.1.4 |
| Built: | 2025-03-12 05:18:03 UTC |
| Source: | https://github.com/heorltd/maic |
From index patient level data and a set of target baseline characteristics, construct the input matrix to the maic.
createMAICInput(index, target, dictionary, matching.variables, x = FALSE)createMAICInput(index, target, dictionary, matching.variables, x = FALSE)
index |
A matrix or data.frame containing patient-level data |
target |
A list containing target summary data |
dictionary |
A data frame containing the columns "match.id", "target.variable", "index.variable" and "match.type" |
matching.variables |
A character vector indicating the match.id to use |
x |
Return subject level inputs? |
The dictionary is a data frame containing at least 4 vectors:
"match.id" - the name of the match, used to refer to it in the matching.variables list
"target.variable" - the name of the variable in the target values list use to inform the matching. Use dependent on type
"index.variable" - the name of the variable in the index data frame to match on.
"match.type" - A string indicating the match type to use. The following values are accepted:
minimum - records with index values lower than the target variable will be assigned 0 weight
maximum - records with index values greater than the target variable will be assigned 0 weight
median - records with index values greater than the target variable will be assigned a value of 1, those lower 0. The target for matching will be a mean of 0.5
quantile.X - Generalisation of the median code. records with index values greater than the target variable will be assigned a value of 1, those lower 0. The target for matching will be a mean of 0.X
mean - records will match index value directly onto target value
proportion - as mean, with index values encoded as 1 = true, 0 = false. If target proportion is exclusive (0 or 1 exactly) then excluded members of the index population shall receive no weighting.
sd - a matching on the square of the index value on the sum of the square of the target mean and target standard deviation. The target mean is provided by the "supplementary.target.variable"
var - a matching on the square of the index value on the sum of the square of the target mean and the variance specified by the target variable. The target mean is provided by the "supplementary.target.variable"
In addition, the following vector may be necessary:
"supplementary.target.variable" - The name of the variable in the target values list that provides e.g. the mean for sd and var matching.
and, for estimating some p-values on difference (e.g. in proportion)
"sample.size.variable" - The name of the variable in the target values list that provides the number of subjects in the sample. Only for reporting.
It is possible to use these match types to match on other variables by pre-processing the input correctly.
Finally, the matching.variables is a list or character vector containing
match.ids to be acted upon in this MAIC.
An object of class maic.input
target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)
From index patient level data and a set of target baseline characteristics, calculate MAIC weights.
maicMatching( index, target, dictionary, matching.variables, reporting.variables = NULL )maicMatching( index, target, dictionary, matching.variables, reporting.variables = NULL )
index |
A matrix or data.frame containing patient-level data |
target |
A list containing target summary data |
dictionary |
A data frame containing the columns "match.id", "target.variable", "index.variable" and "match.type" |
matching.variables |
A character vector indicating the match.id to use |
reporting.variables |
A optional character vector of matches to report
upon (defaults to |
The dictionary is a data frame containing at least 4 vectors:
"match.id" - the name of the match, used to refer to it in the matching.variables list
"target.variable" - the name of the variable in the target values list use to inform the matching. Use dependent on type
"index.variable" - the name of the variable in the index data frame to match on.
"match.type" - A string indicating the match type to use. The following values are accepted:
minimum - records with index values lower than the target variable will be discarded
maximum - records with index values greater than the target variable will be discarded
median - records with index values greater than the target variable will be assigned a value of 1, those lower 0, and those equal 0.5. The target for matching will be a mean of 0.5
quantile.X - Generalisation of the median code. records with index values greater than the target variable will be assigned a value of 1, those lower 0. The target for matching will be a mean of 0.X
mean - records will match index value directly onto target value
proportion - as mean, with index values encoded as 1 = true, 0 = false. If target proportion is exclusive (0 or 1 exactly) then excluded members of the index population shall receive no weighting.
sd - a matching on the square of the index value on the sum of the square of the target mean and target standard deviation. The target mean is provided by the "supplementary.target.variable"
var - a matching on the square of the index value on the sum of the square of the target mean and the variance specified by the target variable. The target mean is provided by the "supplementary.target.variable"
In addition, the following vector may be necessary:
"supplementary.target.variable" - The name of the variable in the target values list that provides e.g. the mean for sd and var matching.
It is possible to use these match types to match on other variables, e.g. variance, by pre-processing the input correctly.
Finally, the matching.variables is a list or character vector containing
match.ids to be acted upon in this MAIC.
An object of class MaicAnalysis, with components weights
and aggregate, containing the weights vector and the covariate
aggregate data respectively
target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) weightObj <- maicMatching( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"))target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) weightObj <- maicMatching( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"))
This function calculates the weights to apply to records for
Matching-Adjusted Indirect Comparison (MAIC), from either a raw input
matrix or a maic.input object
maicWeight(x, opt = TRUE, keep.x = TRUE, ...)maicWeight(x, opt = TRUE, keep.x = TRUE, ...)
x |
Either a |
opt |
return the optim object as attribute |
keep.x |
return the input matrix as an attribute |
... |
Optional arguments to |
A numeric vector of weights corresponding to the rows in the input matrix
target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)
This function calculates the raw, target and achieved covariates given a set of weights. Note that for mean values, bootstrapped standard errors are used and so downstream values (such as p-values for difference) may differ from run to run if the random number stream is not consistent
reportCovariates( index, target, dictionary, matching.variables, weights, tidy = TRUE, var.method = c("ML", "unbiased") )reportCovariates( index, target, dictionary, matching.variables, weights, tidy = TRUE, var.method = c("ML", "unbiased") )
index |
A matrix or data.frame containing patient-level data |
target |
A list containing target summary data |
dictionary |
A data frame containing the columns "match.id", "target.variable", "index.variable" and "match.type" |
matching.variables |
A character vector indicating the match.id to use |
weights |
A numeric vector with weights corresponding to the index data rows |
tidy |
A boolean - return as a data frame (otherwise list) |
var.method |
Estimator type passed through to |
An object of class maic.covariates
target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)target <- c("Air.Flow" = 60, "Water.Temp" = 21, "Prop.Acid.Conc.LT.90" = 0.7, "min.air.flow" = 55) stackloss$match.conc.lt.90 <- ifelse(stackloss$Acid.Conc. < 90, 1, 0) dict <- data.frame( "match.id" = c("airflow", "watertemp", "acidconc", "min.airflow"), "target.variable" = c("Air.Flow", "Water.Temp", "Prop.Acid.Conc.LT.90", "min.air.flow"), "index.variable" = c("Air.Flow", "Water.Temp", "match.conc.lt.90", "Air.Flow"), "match.type" = c("mean", "mean", "proportion", "min"), stringsAsFactors = FALSE) ipmat <- createMAICInput( index = stackloss, target = target, dictionary = dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow")) wts <- maicWeight(ipmat) rcv <- reportCovariates( stackloss, target, dict, matching.variables = c("airflow", "watertemp", "acidconc", "min.airflow"), wts)