## What's in a word?

Trends in cause of death can be an instructive way of looking at past mortality, although we have previously seen that we have to be very careful that an apparent "trend" is not due to changes in recording.  Leaving aside the problems of shifting classification over time, what of the categories themselves?  Table 1 shows the top ten causes of death for males aged 70–74 in England and Wales in 2000.

Table 1. Top ten causes of death for males aged 70–74 in England & Wales in 2000.  Source: 20th Century Mortality.

ProportionDescriptionSecondary description
13.1%Acute myocardial infarction
9.7%Malignant neoplasm of trachea, bronchus and lungBronchus and lung, unspecified
8.3%Other forms of chronic ischaemic heart diseaseUnspecified
5.2%Chronic airways obstruction, not elsewhere classified
5.1%Other forms of chronic ischaemic heart diseaseCoronary atherosclerosis
4.1%Bronchopneumonia, organism unspecified
4.0%Acute but ill-defined cerebrovascular disease
3.4%Malignant neoplasm of prostate
2.1%Malignant neoplasm without specification of siteOther
1.9%Malignant neoplasm of stomachStomach, unspecified

These ten causes cover 57.0% of deaths of males in this age group in 2000.  However, a curious feature of the supplied descriptions lies in the choice of words:

• 4 "unspecified"
• 1 "not elsewhere classified"
• 1 "ill-defined"
• 1 "without specification of site"
Phrases like "not elsewhere classified" suggest a classification used as a bit of a catch-all.  And how useful is the description "Malignant neoplasm without specification of site"?  Apparently the third most common cancer cause of death for this group is one where the certifying doctor could not state what or where it was.

Assume we have a random variable, $$X$$, with expected value ... Read more