Set bcea() Parameters: Constructor and Setters

BCEA

Bayesian statistics

Health economics

Authors

Gianluca Baio

Nathan Green

Published

September 23, 2020

There are several arguments passed to bcea() to specify the form of the analysis. These are

bcea(eff,
     cost,
     ref = 1,
     interventions = NULL,
     .comparison = NULL,
     Kmax = 50000,
     wtp = NULL,
     plot = FALSE)

Those of interest here are:

ref is the reference intervention group to compare against the other groups.
.comparisons are the groups to compare against ref. The default is all of the non-ref groups. This is a new argument in the latest release of BCEA to make it more flexible and consistent with other functions. A preceding dot is used to keep it back-compatible with previous versions of BCEA. Argument c is partially matched with both c and comparison otherwise throwing an error.
Kmax is the maximum value of the willingness-to-pay to calculate statistics for.

During an analysis we may want to explore changing some of these parameters and keeping all of the others the same. We can do with with package setter functions.

Changing Reference Group

Load cost-effectiveness data.

data(Vaccine)

We first create bcea object using the constructor function for 2 different reference groups.

he_ref1 <- bcea(eff, cost,
                ref = 1,           
                interventions = treats,
                Kmax = 50000)  
str(he_ref1)
#> List of 24
#>  $ n_sim        : int 1000
#>  $ n_comparators: num 2
#>  $ n_comparisons: int 1
#>  $ delta_e      :'data.frame':   1000 obs. of  1 variable:
#>   ..$ Vaccination: num [1:1000] -0.000148 -0.000152 -0.000192 -0.000504 -0.000394 ...
#>  $ delta_c      :'data.frame':   1000 obs. of  1 variable:
#>   ..$ Vaccination: num [1:1000] -5.84 -3.54 -10.15 -6.45 -6.68 ...
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Vaccination"
#>  $ Kmax         : num 50000
#>  $ k            : num [1:501] 0 100 200 300 400 500 600 700 800 900 ...
#>  $ ceac         : num [1:501, 1] 0.98 0.978 0.977 0.977 0.977 0.976 0.976 0.975 0.975 0.973 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ k   : NULL
#>   .. ..$ ints: chr "Vaccination"
#>  $ ib           : num [1:501, 1:1000, 1] 5.84 5.83 5.81 5.8 5.78 ...
#>   ..- attr(*, "dimnames")=List of 3
#>   .. ..$ k   : NULL
#>   .. ..$ sims: NULL
#>   .. ..$ ints: chr "Vaccination"
#>  $ eib          : num [1:501, 1] 5.04 5.01 4.99 4.96 4.94 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ k   : NULL
#>   .. ..$ ints: chr "Vaccination"
#>  $ kstar        : num 20100
#>  $ best         : int [1:501] 1 1 1 1 1 1 1 1 1 1 ...
#>  $ U            : num [1:1000, 1:501, 1:2] -10.41 -5.83 -5.78 -12.21 -9.79 ...
#>   ..- attr(*, "dimnames")=List of 3
#>   .. ..$ sims: NULL
#>   .. ..$ k   : NULL
#>   .. ..$ ints: chr [1:2] "Status Quo" "Vaccination"
#>  $ vi           : num [1:1000, 1:501] -0.754 3.821 3.871 -2.553 -0.131 ...
#>  $ Ustar        : num [1:1000, 1:501] -10.41 -5.83 -5.78 -12.21 -9.79 ...
#>  $ ol           : num [1:1000, 1:501] 0 0 0 0 0 0 0 0 0 0 ...
#>  $ evi          : num [1:501] 0.0562 0.0573 0.0586 0.0598 0.0611 ...
#>  $ ref          : int 1
#>  $ comp         : int 2
#>  $ step         : num 100
#>  $ interventions: chr [1:2] "Status Quo" "Vaccination"
#>  $ e            : num [1:1000, 1:2] -0.001047 -0.000884 -0.00089 -0.001643 -0.001352 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ : NULL
#>   .. ..$ : chr [1:2] "Status Quo" "Vaccination"
#>  $ c            : num [1:1000, 1:2] 10.41 5.83 5.78 12.21 9.79 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ : NULL
#>   .. ..$ : chr [1:2] "Status Quo" "Vaccination"
#>  - attr(*, "class")= chr [1:2] "bcea" "list"

ceplane.plot(he_ref1)

he_ref2 <- bcea(eff, cost,
                ref = 2,           
                interventions = treats,
                Kmax = 50000)  

str(he_ref2[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Status Quo"
#>  $ ref          : int 2
#>  $ comp         : int 1

Alternatively, we can do the same by modifying the first output.

###setReferenceGroup(he_ref1) <- 2
`setReferenceGroup<-`(he_ref1,2)
#>  $ n_sim        : int 1000
#>  $ n_comparators: num 2
#>  $ n_comparisons: int 1
#>  $ delta_e      :'data.frame':   1000 obs. of  1 variable:
#>   ..$ Status Quo: num [1:1000] 0.0001480642 0.0001515688 0.0001922809 0.0005037002 0.0003943893 ...
#>  $ delta_c      :'data.frame':   1000 obs. of  1 variable:
#>   ..$ Status Quo: num [1:1000] 5.843391 3.538563 10.15072 6.445766 6.680534 ...
#>  $ ICER         : Named num 20097.59
#>  $ Kmax         : num 50000
#>  $ k            : num [1:501] 0 100 200 300 400 500 600 700 800 900 ...
#>  $ ceac         : num [1:501, 1] 0.02 0.022 0.023 0.023 0.023 0.024 0.024 0.025 0.025 0.027 ...
#>  $ ib           : num [1:501, 1:1000, 1] -5.843391 -5.828585 -5.813778 -5.798972 -5.784166 ...
#>  $ eib          : num [1:501, 1] -5.035983 -5.010925 -4.985868 -4.96081 -4.935752 ...
#>  $ kstar        : num 20100
#>  $ best         : int [1:501] 1 1 1 1 1 1 1 1 1 1 ...
#>  $ U            : num [1:1000, 1:501, 1:2] -10.409146 -5.834875 -5.784903 -12.208484 -9.786787 ...
#>  $ vi           : num [1:1000, 1:501] -0.7536823 3.820589 3.8705605 -2.5530205 -0.131323 ...
#>  $ Ustar        : num [1:1000, 1:501] -10.409146 -5.834875 -5.784903 -12.208484 -9.786787 ...
#>  $ ol           : num [1:1000, 1:501] 0 0 0 0 0 0 0 0 0 0 ...
#>  $ evi          : num [1:501] 0.05620997 0.05733657 0.058577 0.05984868 0.06112037 ...
#>  $ ref          : int 2
#>  $ comp         : int 1
#>  $ step         : num 100
#>  $ interventions: chr [1:2] "Status Quo" "Vaccination"
#>  $ e            : num [1:1000, 1:2] -0.0010466668 -0.0008836105 -0.0008898137 -0.0016430238 -0.0013518841 ...
#>  $ c            : num [1:1000, 1:2] 10.409146 5.834875 5.784903 12.208484 9.786787 ...

str(he_ref1[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Vaccination"
#>  $ ref          : int 1
#>  $ comp         : int 2

Changing Kmax

In the same way as above we can change Kmax in 2 equivalent ways.

he_Kmax1 <- bcea(eff, cost,
                 ref = 1,           
                 interventions = treats,
                 Kmax = 50000)  

str(he_Kmax1[c("n_comparators", "ICER", "ref", "comp", "Kmax")])
#> List of 5
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Vaccination"
#>  $ ref          : int 1
#>  $ comp         : int 2
#>  $ Kmax         : num 50000

he_Kmax2 <- bcea(eff, cost,
                 ref = 2,           
                 interventions = treats,
                 Kmax = 2000)  

str(he_Kmax2[c("n_comparators", "ICER", "ref", "comp", "Kmax")])
#> List of 5
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Status Quo"
#>  $ ref          : int 2
#>  $ comp         : int 1
#>  $ Kmax         : num 2000

setKmax(he_Kmax1) <- 2000

str(he_Kmax1[c("n_comparators", "ICER", "ref", "comp", "Kmax")])
#> List of 5
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 20098
#>   ..- attr(*, "names")= chr "Vaccination"
#>  $ ref          : int 1
#>  $ comp         : int 2
#>  $ Kmax         : num 2000

Change Comparison Groups

Lets load some data with more than two groups.

data(Smoking)

Defaults is all other groups which in this case is 2, 3 and 4.

he_comp234 <- bcea(eff, cost,
                   ref = 1,           
                   interventions = treats,
                   Kmax = 50000)  

str(he_comp234[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 4
#>  $ ICER         : Named num [1:3] 159 196 198
#>   ..- attr(*, "names")= chr [1:3] "Self-help" "Individual counselling" "Group counselling"
#>  $ ref          : int 1
#>  $ comp         : int [1:3] 2 3 4

ceplane.plot(he_comp234, wtp = 2000)

Let us compare against only groups 2.

he_comp2 <- bcea(eff, cost,
                 ref = 1,
                 .comparison = 2,
                 interventions = treats,
                 Kmax = 2000)  

str(he_comp2[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 159
#>   ..- attr(*, "names")= chr "Self-help"
#>  $ ref          : int 1
#>  $ comp         : num 2

ceplane.plot(he_comp2, wtp = 2000)

We can achieve the same thing using the appropriate setter.

setComparisons(he_comp234) <- 2

str(he_comp234[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 2
#>  $ ICER         : Named num 159
#>   ..- attr(*, "names")= chr "Self-help"
#>  $ ref          : int 1
#>  $ comp         : num 2

ceplane.plot(he_comp234, wtp = 2000)

We can select multiple comparison groups too. Let us compare against only groups 2 and 4.

he_comp24 <- bcea(eff, cost,
                  ref = 1,
                  .comparison = c(2,4),
                  interventions = treats,
                  Kmax = 2000)  
str(he_comp24[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 3
#>  $ ICER         : Named num [1:2] 159 198
#>   ..- attr(*, "names")= chr [1:2] "Self-help" "Group counselling"
#>  $ ref          : int 1
#>  $ comp         : num [1:2] 2 4

ceplane.plot(he_comp24, wtp = 2000)

setComparisons(he_comp234) <- c(2,4)

str(he_comp234[c("n_comparators", "ICER", "ref", "comp")])
#> List of 4
#>  $ n_comparators: num 3
#>  $ ICER         : Named num [1:2] 159 198
#>   ..- attr(*, "names")= chr [1:2] "Self-help" "Group counselling"
#>  $ ref          : int 1
#>  $ comp         : num [1:2] 2 4

ceplane.plot(he_comp24, wtp = 2000)

Further, a bcea object with all comparison groups can be passed to other functions such as ceplane.plot and ceac.plot with a comparison argument, which will do the modifications using these functions internally instead.

ceplane.plot(he_comp234, comparison = 2, wtp = 2000)

ceplane.plot(he_comp234, comparison = c(2,4), wtp = 2000)