Derived variables

STATA CODE

Name of derived variable: UDCOD
Variable label: Underlying cause of death (20 categories)
Categories (with distribution up to December 2008 in LSLOAD72):

1 "Ischaemic heart disease"
2 "Stroke"
3 "Pulmonary Disease"
4 "Lung Cancer"
5 "Other Cancers"
6 "Infectious or parasitic disease"
7 "Diabetes mellitus"
8 "Gastro-intestinal tract disease"
9 "Liver disease"
10 "Mental or behavioural"
11 "Abnormalities or lab results"
12 "Other circulatory disease"
13 "Accidents or self harm"
14 "Muscular disease"
15 "Benign neoplasms"
16 "Nervous system diseases"
17 "Genito-urinary"
18 "Other endocrine"
19 "Skin disease"
20 "Other cause"

52,205
24,094
26,692
11,115
44,406
1,424
2,593
4,872
1,814
4,029
3,199
23,004
5,646
1,514
802
4,028
2,649
314
374
17,804

Derived from: IC10UDE, IC10UFDE, UCDE3DDE, ICD9UDE, ICD9UFDE, UCDE3CDE, DEYRBDE (all in the DETH table)
Study group: People who have died since 1971 i.e. all cases in the DETH table. If living people are also included in the dataset they will receive the default value of 0, which will then mean ‘still alive’.
Likely use: This STATA code groups cause of death into 20 categories, using the ICD (International Classification of Diseases) coding in use at the time the death was recorded. LS users should note that ICD coding changes with each new version; however, the categories used in this variable are sufficiently broad that there is no need for the STATA code to take account of these changes.
Notes:

Be aware that this is not the only way of grouping causes of death; for instance, not all liver diseases appear under 'Liver disease'. As always, the most appropriate coding will depend on the research question.

Users may wish to group these categories further, possibly by adding codes 6-19 to code 20 ('Other cause'); alternatively, codes 1,2 and 12 could form a 'circulatory disease' category, while 8 and 9 could be united as 'digestive diseases', etc.

This variable represents underlying cause of death (the most commonly-used category), but the same principle could be used to group values for contributory cause of death. The training module on mortality offers more information.

Back to variables on cause of death

Last modified 19 March 2010