Heterogeneity & Demographic Analysis

Introduction

Heterogeneity analysis is a way to explore how the results of a model can vary depending on the characteristics of individuals in a population, and demographic analysis estimates the average values of a model over an entire population.

In practice these two analyses naturally complement each other: heterogeneity analysis runs the model on multiple sets of parameters (reflecting different characteristics found in the target population), and demographic analysis combines the results.

For this example we will use the result from the assessment of a new total hip replacement previously described in vignette("d-non-homogeneous", "heemod").

Population characteristics

The characteristics of the population are input from a table, with one column per parameter and one row per individual. Those may be for example the characteristics of the indiviuals included in the original trial data.

For this example we will use the characteristics of 100 individuals, with varying sex and age, specified in the data frame tab_indiv:

tab_indiv
## # A tibble: 100 × 2
##      age   sex
##    <dbl> <int>
##  1    65     1
##  2    52     1
##  3    54     0
##  4    56     0
##  5    50     0
##  6    44     1
##  7    72     1
##  8    46     0
##  9    47     0
## 10    55     1
## # ℹ 90 more rows
library(ggplot2)
ggplot(tab_indiv, aes(x = age)) +
  geom_histogram(binwidth = 2)

Running the analysis

res_mod, the result we obtained from run_model() in the Time-varying Markov models vignette, can be passed to update() to update the model with the new data and perform the heterogeneity analysis.

res_h <- update(res_mod, newdata = tab_indiv)
## No weights specified in update, using equal weights.
## Updating strategy 'standard'...
## Updating strategy 'np1'...

Interpreting results

The summary() method reports summary statistics for cost, effect and ICER, as well as the result from the combined model.

summary(res_h)
## An analysis re-run on 100 parameter sets.
## 
## * Unweighted analysis.
## 
## * Values distribution:
## 
##                                  Min.      1st Qu.     Median        Mean
## standard - Cost          438.70535048  613.9316623 632.979866 696.7592773
## standard - Effect          6.12465030   23.3226486  26.729786  26.0065825
## standard - Cost Diff.               -            -          -           -
## standard - Effect Diff.             -            -          -           -
## standard - Icer                     -            -          -           -
## np1 - Cost               590.76054210  637.9767000 643.532617 661.6910119
## np1 - Effect               6.13624942   23.4706053  27.104563  26.2718381
## np1 - Cost Diff.        -163.38052116 -122.7948420  10.552752 -35.0682654
## np1 - Effect Diff.         0.01159912    0.1948185   0.220806   0.2652556
## np1 - Icer              -353.62679735 -327.6476693  57.461391  97.0346111
##                             3rd Qu.         Max.
## standard - Cost         819.1977737   875.943516
## standard - Effect        29.2164282    31.299481
## standard - Cost Diff.             -            -
## standard - Effect Diff.           -            -
## standard - Icer                   -            -
## np1 - Cost              696.4029317   712.562995
## np1 - Effect             29.4828147    31.532860
## np1 - Cost Diff.         24.0450377   152.055192
## np1 - Effect Diff.        0.3747771     0.462014
## np1 - Icer              115.2176112 13109.195655
## 
## * Combined result:
## 
## 2 strategies run for 60 cycles.
## 
## Initial state counts:
## 
## PrimaryTHR = 1000L
## SuccessP = 0L
## RevisionTHR = 0L
## SuccessR = 0L
## Death = 0L
## 
## Counting method: 'beginning'.
## 
## Values:
## 
##           utility     cost
## standard 26006.58 696759.3
## np1      26271.84 661691.0
## 
## Efficiency frontier:
## 
## np1
## 
## Differences:
## 
##     Cost Diff. Effect Diff.      ICER     Ref.
## np1  -35.06827    0.2652556 -132.2055 standard

The variation of cost or effect can then be plotted.

plot(res_h, result = "effect", binwidth = 5)

plot(res_h, result = "cost", binwidth = 50)

plot(res_h, result = "icer", type = "difference",
     binwidth = 500)

plot(res_h, result = "effect", type = "difference",
     binwidth = .1)

plot(res_h, result = "cost", type = "difference",
     binwidth = 30)

The results from the combined model can be plotted similarly to the results from run_model().

plot(res_h, type = "counts")

Weighted results

Weights can be used in the analysis by including an optional column .weights in the new data to specify the respective weights of each strata in the target population.

tab_indiv_w
## # A tibble: 100 × 3
##      age   sex .weights
##    <dbl> <int>    <dbl>
##  1    54     0    0.159
##  2    66     0    0.128
##  3    63     0    0.768
##  4    67     1    0.357
##  5    76     0    0.376
##  6    59     1    0.431
##  7    49     1    0.658
##  8    52     1    0.159
##  9    64     1    0.946
## 10    30     0    0.463
## # ℹ 90 more rows
res_w <- update(res_mod, newdata = tab_indiv_w)
## Updating strategy 'standard'...
## Updating strategy 'np1'...
res_w
## An analysis re-run on 100 parameter sets.
## 
## * Weights distribution:
## 
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## 0.002929 0.210924 0.464004 0.480499 0.713336 0.997128 
## 
## Total weight: 48.04993
## 
## * Values distribution:
## 
##                                  Min.      1st Qu.      Median        Mean
## standard - Cost          500.08967163  605.0062810 621.5892423 677.0845757
## standard - Effect         10.06345874   23.3435213  26.7297859  25.5830584
## standard - Cost Diff.               -            -           -           -
## standard - Effect Diff.             -            -           -           -
## standard - Icer                     -            -           -           -
## np1 - Cost               607.16692250  635.5509751 640.0588497 656.1271066
## np1 - Effect              10.13073146   23.4946595  27.1045630  25.8289154
## np1 - Cost Diff.        -159.96283707  -99.5031416  18.4696074 -20.9574691
## np1 - Effect Diff.         0.05767389    0.1948185   0.2116899   0.2458569
## np1 - Icer              -351.98058303 -304.0330575  83.6463073  60.4642308
##                             3rd Qu.         Max.
## standard - Cost         786.6690449  871.1621236
## standard - Effect        29.0596426   31.6837747
## standard - Cost Diff.             -            -
## standard - Effect Diff.           -            -
## standard - Icer                   -            -
## np1 - Cost              687.1659033  711.1992865
## np1 - Effect             29.2683350   31.9214350
## np1 - Cost Diff.         30.5446941  107.0772509
## np1 - Effect Diff.        0.3272774    0.4544649
## np1 - Icer              156.7853582 1856.5985016
## 
## * Combined result:
## 
## 2 strategies run for 60 cycles.
## 
## Initial state counts:
## 
## PrimaryTHR = 1000L
## SuccessP = 0L
## RevisionTHR = 0L
## SuccessR = 0L
## Death = 0L
## 
## Counting method: 'beginning'.
## 
## Values:
## 
##           utility     cost
## standard 25583.06 677084.6
## np1      25828.92 656127.1
## 
## Efficiency frontier:
## 
## np1
## 
## Differences:
## 
##     Cost Diff. Effect Diff.      ICER     Ref.
## np1  -20.95747    0.2458569 -85.24254 standard

Parallel computing

Updating can be significantly sped up by using parallel computing. This can be done in the following way:

  • Define a cluster with the use_cluster() functions (i.e. use_cluster(4) to use 4 cores).
  • Run the analysis as usual.
  • To stop using parallel computing use the close_cluster() function.

Results may vary depending on the machine, but we found speed gains to be quite limited beyond 4 cores.