You are in: Home › Statistics › Deaths › Death Certificates and Coding the Causes of Death ›

Death Certificates and Coding the Causes of Death

This section describes briefly how causes of death are recorded on death certificates. It mentions the bases on which figures can be produced, and indicates how the causes of death are coded.

The Death Certificate

The death cartificate is completed by a registered medical practitioner. While there are some general notes of guidance on how the form should be filled in, including advice on points that should be covered and examples of the kinds of terminology that should (or should not) be used, the information that is recorded is a matter for the doctor's clinical judgement.

The certificate has two parts:

• Part 1 should show the immediate cause of death, and then work back in time to the disease or condition that started the process. It should therefore identify:
  • the disease or condition that led directly to the death;
  • any antecedent or intermediate causes of that event or disease or condition (i.e. which occurred earlier in the chain of events that led to the death); and, eventually, going back to -
  • the underlying cause of death - this is defined (on pages 33-34 of Volume 2 of the International Standard Classification of Diseases and Related Health Problems) as: "(a) the disease or injury which initiated the chain of morbid events leading directly to death or (b) the circumstances of the accident or violence which produced the fatal injury".
• For example:
  • the disease or condition that led directly to death might be a compound fracture of the skull or a ruptured liver;
  • whereas the underlying cause of death might be a gunshot wound, or being injured in a road accident.

• Part 2 can be used to record other significant diseases, conditions or accidents which contributed to the occurence of the death , but were not part of the main sequence leading to the death. However, Part 2 should not be used to list all the conditions that were present at death. For example:
  • a person with diabetes who died of lung cancer might have died sooner than would have been the case if he/she did not have diabetes - if that is thought to be the case, diabetes should be recorded as contributing to the death; but
  • if the person also had osteoarthiritis, it is unlikely that it would have contributed to the death, so it should not be mentioned in Part 2.

The Bases on Which Figures can be Produced

As a result, figures for a particular cause of death can be produced on three bases:

• "underlying cause" - cases where it was the disease or injury which initiated the chain of morbid events leading directly to death or the accident/act which produced the fatal injury;
• "contributory factor" - cases where it was not the underlying cause of death, but it did contribute to the occurence of the death - e.g. it did not cause the death, but may have hastened its occurence.
• "any mention" - i.e. whether it appeared to be the underlying cause of the death, or was just a factor which contributed to, or may have hastened, the occurence of the death.

Figures are normally provided on the basis of the "underlying cause". Every death has just one "underlying cause" coded, and so is counted only once in figures which are produced on the "underlying cause" basis.

A death may have several other causes coded as contributory factors, so could be counted several times in figures which are produced on the "contributory factor" or "any mention" bases. For that reason, figures are normally provided on the basis of the "underlying cause".

Coding the Causes of Death

For deaths registered from 1 January 2000, the causes of death have been coded in accordance with the International Statistical Classification of Diseases and Related Health Problems (Tenth Revision), which may be referred to as "ICD10". Deaths which were registered in 1999 were coded using both ICD9 and ICD10, in order to obtain an indication of the effect on the figures of the change in the classification. The definitive figures for 1999 are those produced using ICD9.

The classification of the underlying cause of death is based on the information collected on the death certificate together with any additional information provided by the certifying doctor or, in some cases, Procurators Fiscal and the Crown Office.

The application of the ICD10 rules can lead to the underlying cause being coded to something other than the final entry in Part 1 of the death certificate - for example:

• if this is something which is "ill-defined" for ICD purposes (e.g. "organ failure"), in which case a more specific cause of death is coded, if one is available;
• if it is an obvious consequence of one or more of the other conditions which were reported (either in Part 1 or Part 2) - e.g. "any disease described as secondary should be assumed to be a consequence of the most probably primary cause entered on the certificate";
• if it is "linked by a provision in the classification or in the notes for underlying cause mortality coding with one or more of the other conditions on the certificate", in which case the appropriate combination is coded.

In January 1996, GROS introduced an automated method of coding the causes of death, using software provided by the US National Centre for Health Statistics. While this handles "straightforward" cases correctly, experienced coding staff are still needed to check the results and to code the more complicated cases.