Computing Comorbidity Scores

Computing comorbidity scores such as the weighted Charlson score (Charlson, 1987 ) and the Elixhauser comorbidity score (Elixhauser, 1998 ) using ICD-9-CM or ICD-10 codes (Quan, 2005 ).

2018-10-08

comorbidity is an R package for computing comorbidity scores such as the weighted Charlson score and the Elixhauser comorbidity score; both ICD-10 and ICD-9 coding systems are supported.

Installation

comorbidity is on CRAN. You can install it as usual with:

install.packages("comorbidity")

Alternatively, you can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("ellessenne/comorbidity")

Simulating ICD-10 codes

The comorbidity packages includes a function named sample_diag() that allows simulating ICD diagnostic codes in a straightforward way. For instance, we could simulate ICD-10 codes:

# load the comorbidity package
library(comorbidity)
# set a seed for reproducibility
set.seed(1)
# simulate 50 ICD-10 codes for 5 individuals
x <- data.frame(
  id = sample(1:5, size = 50, replace = TRUE),
  code = sample_diag(n = 50),
  stringsAsFactors = FALSE
)
x <- x[order(x$id, x$code), ]
print(head(x, n = 15), row.names = FALSE)
#>  id code
#>   1 C838
#>   1  H30
#>   1 I260
#>   1 I469
#>   1 K611
#>   1 L949
#>   1  V09
#>   2 B677
#>   2 C081
#>   2 I446
#>   2 K225
#>   2  M41
#>   2 M430
#>   2 T635
#>   2 U016

It is also possible to simulate from two different versions of the ICD-10 coding system. The default is to simulate ICD-10 codes from the 2011 version:

set.seed(1)
x1 <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30),
  stringsAsFactors = FALSE
)
set.seed(1)
x2 <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30, version = "ICD10_2011"),
  stringsAsFactors = FALSE
)
# should return TRUE
all.equal(x1, x2)
#> [1] TRUE

Alternatively, you could use the 2009 version:

set.seed(1)
x1 <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30, version = "ICD10_2009"),
  stringsAsFactors = FALSE
)
set.seed(1)
x2 <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30, version = "ICD10_2011"),
  stringsAsFactors = FALSE
)
# should not return TRUE
all.equal(x1, x2)
#> [1] "Component \"code\": 29 string mismatches"

Simulating ICD-9 codes

ICD-9 codes can be easily simulated too:

set.seed(2)
x9 <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30, version = "ICD9_2015"),
  stringsAsFactors = FALSE
)
x9 <- x9[order(x9$id, x9$code), ]
print(head(x9, n = 15), row.names = FALSE)
#>  id  code
#>   1 01161
#>   1  2535
#>   1 37854
#>   1 46451
#>   1 83664
#>   1 94433
#>   1  9711
#>   1 E8038
#>   1 E8498
#>   1  V092
#>   2 01092
#>   2 01301
#>   2  2337
#>   2 37033
#>   2 67450

Computing comorbidity scores

The main function of the comorbidity package is named comorbidity(), and it can be used to compute any supported comorbidity score; scores can be specified by setting the score argument, which is required.

Say we have 3 individuals with a total of 30 ICD-10 diagnostic codes:

set.seed(1)
x <- data.frame(
  id = sample(1:3, size = 30, replace = TRUE),
  code = sample_diag(n = 30),
  stringsAsFactors = FALSE
)

We could compute the Charlson score, index, and each comorbidity domain:

charlson <- comorbidity(x = x, id = "id", code = "code", score = "charlson", icd = "icd10")
charlson
#>   id ami chf pvd cevd dementia copd rheumd pud mld diab diabwc hp rend canc
#> 1  1   0   0   0    0        0    0      0   0   1    0      0  0    0    0
#> 2  2   0   0   0    0        0    0      0   0   0    0      0  0    1    1
#> 3  3   0   0   0    0        0    0      0   0   0    0      0  0    0    0
#>   msld metacanc aids score index wscore windex
#> 1    0        0    0     1   1-2      1    1-2
#> 2    0        0    0     2   1-2      4    3-4
#> 3    0        0    0     0     0      0      0

The default is to assume ICD-10 codes:

charlson.default <- comorbidity(x = x, id = "id", code = "code", score = "charlson")
all.equal(charlson, charlson.default)
#> [1] TRUE

Alternatively, we could compute the Elixhauser score:

elixhauser <- comorbidity(x = x, id = "id", code = "code", score = "elixhauser", icd = "icd10")
elixhauser
#>   id chf carit valv pcd pvd hypunc hypc para ond cpd diabunc diabc hypothy rf
#> 1  1   0     0    0   0   0      0    0    0   0   0       0     0       0  0
#> 2  2   0     0    0   0   0      0    0    0   0   0       0     0       0  1
#> 3  3   0     0    0   0   0      0    0    0   0   0       0     0       0  0
#>   ld pud aids lymph metacanc solidtum rheumd coag obes wloss fed blane dane
#> 1  1   0    0     0        0        0      0    0    0     0   0     0    0
#> 2  0   0    0     0        0        1      0    0    0     0   0     0    0
#> 3  0   0    0     0        0        0      0    0    0     0   0     0    0
#>   alcohol drug psycho depre score index wscore windex
#> 1       0    0      0     0     1   1-4      4    1-4
#> 2       0    0      0     0     2   1-4     13    >=5
#> 3       0    0      0     0     0     0      0      0

Conversely, say we have 5 individuals with a total of 100 ICD-9 diagnostic codes:

set.seed(3)
x <- data.frame(
  id = sample(1:5, size = 100, replace = TRUE),
  code = sample_diag(n = 100, version = "ICD9_2015"),
  stringsAsFactors = FALSE
)

The Charlson and Elixhauser comorbidity codes can be easily computed:

We could compute the Charlson score, index, and each comorbidity domain:

charlson9 <- comorbidity(x = x, id = "id", code = "code", score = "charlson", icd = "icd9")
charlson9
#>   id ami chf pvd cevd dementia copd rheumd pud mld diab diabwc hp rend canc
#> 1  1   0   0   0    0        0    0      0   0   0    0      0  0    0    1
#> 2  2   0   0   0    0        0    0      0   0   0    0      1  0    0    1
#> 3  3   0   0   0    0        0    0      0   0   0    1      0  0    0    1
#> 4  4   1   0   0    0        0    0      0   0   0    0      0  0    0    1
#> 5  5   0   0   0    0        0    0      0   0   0    0      0  0    0    1
#>   msld metacanc aids score index wscore windex
#> 1    0        0    0     1   1-2      2    1-2
#> 2    1        0    0     3   3-4      7    >=5
#> 3    0        0    0     2   1-2      3    3-4
#> 4    0        0    0     2   1-2      3    3-4
#> 5    0        0    0     1   1-2      2    1-2

Alternatively, we could compute the Elixhauser score:

elixhauser9 <- comorbidity(x = x, id = "id", code = "code", score = "elixhauser", icd = "icd9")
elixhauser9
#>   id chf carit valv pcd pvd hypunc hypc para ond cpd diabunc diabc hypothy rf
#> 1  1   0     0    0   0   0      0    0    0   0   0       0     0       0  0
#> 2  2   0     0    0   0   0      0    0    0   0   0       0     1       0  0
#> 3  3   0     0    0   0   0      0    0    0   0   0       0     1       0  0
#> 4  4   0     0    0   0   0      0    1    0   0   0       0     0       0  0
#> 5  5   0     0    0   1   0      0    0    0   0   0       0     0       0  0
#>   ld pud aids lymph metacanc solidtum rheumd coag obes wloss fed blane dane
#> 1  0   0    0     0        0        1      0    0    0     0   0     0    0
#> 2  1   0    0     1        0        1      0    0    0     0   0     0    0
#> 3  0   0    0     1        0        0      0    0    0     0   0     0    0
#> 4  0   0    0     0        0        1      0    0    0     0   0     0    0
#> 5  0   0    0     0        0        1      0    0    0     0   0     0    0
#>   alcohol drug psycho depre score index wscore windex
#> 1       0    0      0     0     1   1-4      7    >=5
#> 2       0    1      0     0     5   >=5      7    >=5
#> 3       1    0      0     0     3   1-4      2    1-4
#> 4       0    0      0     1     3   1-4      1    1-4
#> 5       0    0      0     0     2   1-4     13    >=5

Citation

If you find comorbidity useful, please cite it in your publications:

citation("comorbidity")
#> 
#> To cite the comorbidity package in publications, please use:
#> 
#>   Gasparini, (2018). comorbidity: An R package for computing
#>   comorbidity scores. Journal of Open Source Software, 3(23), 648,
#>   https://doi.org/10.21105/joss.00648
#> 
#> A BibTeX entry for LaTeX users is
#> 
#>   @Article{,
#>     author = {Alessandro Gasparini},
#>     title = {comorbidity: An R package for computing comorbidity scores},
#>     journal = {Journal of Open Source Software},
#>     year = {2018},
#>     volume = {3},
#>     issue = {23},
#>     pages = {648},
#>     doi = {10.21105/joss.00648},
#>     url = {https://doi.org/10.21105/joss.00648},
#>   }

References

This package is based on the ICD-10-based formulations of the Charlson score and Elixhauser score proposed by Quan et al. in 2005. The ICD-9 formulation of the Charlson score is also from Quan et al. The ICD-9-based Elixhauser score is according to the AHRQ formulation (Moore et al., 2017). Weights for the Charlson score are based on the original formulation by Charlson et al. in 1987, while weights for the Elixhauser score are based on work by van Walraven et al. Finally, the categorisation of scores and weighted scores is based on work by Menendez et al.

• Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Medical Care 2005; 43(11):1130-1139. DOI: 10.1097/01.mlr.0000182534.19832.83
• Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. Journal of Chronic Diseases 1987; 40:373-383. DOI: 10.1016/0021-9681(87)90171-8
• Moore BJ, White S, Washington R, Coenen N, and Elixhauser A. Identifying increased risk of readmission and in-hospital mortality using hospital administrative data: the AHRQ Elixhauser comorbidity index. Medical Care 2017; 55(7):698-705. DOI: 10.1097/MLR.0000000000000735
• Elixhauser A, Steiner C, Harris DR and Coffey RM. Comorbidity measures for use with administrative data. Medical Care 1998; 36(1):8-27. DOI: 10.1097/00005650-199801000-00004
• van Walraven C, Austin PC, Jennings A, Quan H and Forster AJ. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Medical Care 2009; 47(6):626-633. DOI: 10.1097/mlr.0b013e31819432e5
• Menendez ME, Neuhaus V, van Dijk CN, Ring D. The Elixhauser comorbidity method outperforms the Charlson index in predicting inpatient death after orthopaedic surgery. Clinical Orthopaedics and Related Research 2014; 472(9):2878-2886. DOI: 10.1007/s11999-014-3686-7

Copyright

The icon for the hex sticker was made by Freepik from Flaticon and is licensed by CC 3.0 BY.