Letters

Mentioning smoking as a cause of death on death certificates

BMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7170.1456 (Published 21 November 1998) Cite this as: BMJ 1998;317:1456

Relatives can be asked in death registries about smoking habit of dead person

  1. T H Lam, Professor,
  2. S Y Ho, Researcher,
  3. A J Hedley, Professor (hrmrlth{at}hkucc.hku.hk),
  4. K H Mak, Consultant in community medicine
  1. Department of Community Medicine, University of Hong Kong, Hong Kong, China
  2. Department of Health, Wanchai, Hong Kong, China
  3. Pilgrim Hospital, Boston, Lincolnshire PE21 9QS

    EDITOR—Robinson et al report that most British doctors do not put smoking as a cause of death.1 We should recognise, however, that doctors will have difficulty in identifying a clear link between smoking and death in individual patients. The recommendation in the report of the Scientific Committee on Tobacco and Health that smoking status should be “recorded as part of the death registration process to aid monitoring the evolving epidemic of tobacco related diseases” is more practical.2 We agree that smoking status should be recorded on all death certificates and not only as a specified cause of death, but further legislative changes to introduce this will incur further delays. We cannot expect more than one additional question in the certificate, and this should ask whether the dead person had been a regular smoker. A smoking history should be available in the medical record, but if it is not then the doctor should ask the relatives. Whether doctors are prepared to do this reliably is questionable and needs further study.

    In a large case-control study in China the smoking habits of people who had died were obtained from their surviving spouses in home interviews a few years after the death,3 but this method is expensive. As a quick alternative we are obtaining information on the smoking habit of people who have died by interviewing their relatives at death registries. This has been part of a case-control study throughout 1998.4 Relatives approaching a registration counter are invited to complete a questionnaire about the person who has died and about a control, who can be a surviving spouse or another living relative. The process takes only 10-15 minutes while the person waits for the death certificate. All people who have died at the age of30 are included.

    Up to 5 October 1998 information on smoking habit at about 10 years before death was obtained from all death registries in Hong Kong for 11 632 male and 9464 female people who had died. Ninety four per cent of the relatives were willing to participate. Altogether 8032 (69%) of the men and 2052 of the women who had died had smoked. Attributable risks due to smoking can be calculated by comparing the smoking status of the dead subjects with that of the controls. We conclude that this method is cheap, efficient, and acceptable to the relatives and that it has potential for immediate implementation in other countries, particularly those where death certification is unreliable.

    References

    Relatives don't like this being done

    1. Paula Johnson, Specialist registrar in respiratory medicine (mmxphj{at}mmn1.medical.nottingham.ac.uk)
    1. Department of Community Medicine, University of Hong Kong, Hong Kong, China
    2. Department of Health, Wanchai, Hong Kong, China
    3. Pilgrim Hospital, Boston, Lincolnshire PE21 9QS

      EDITOR—Robinson et al suggest that smoking should be listed as a cause of death on death certificates.1 In 1992, as a senior house officer in respiratory medicine, I began to do this. Within a few weeks it led to several complaints from distressed relatives, who inferred that doctors were apportioning blame to the person who had died. In one case I was even asked to reissue a certificate, with smoking removed from the causes of death. I have not since attempted to put smoking on death certificates, and I suspect that my experiences were not unique.

      References

      View Abstract

      Sign in

      Log in through your institution

      Subscribe