Published 14 October 2009, doi:10.1136/bmj.b4135
Cite this as: BMJ 2009;339:b4135

Endgames

Statistical question

Case-control studies

Philip Sedgwick, senior lecturer in medical statistics

1 Centre for Medical and Healthcare Education, St George’s, University of London, Tooting, London SW17 0RE

p.sedgwick{at}sgul.ac.uk

Researchers used a case-control study to investigate whether passive smoking during childhood was associated with lung cancer in adulthood. Passive smoking in childhood was recorded using self completed questionnaires. Participants were recorded as exposed to smoking if anyone smoked regularly where the participant lived as a child. Those individuals with diagnosed lung cancer were more likely to have been exposed to smoking in childhood than those without diagnosed lung cancer (72.8% v 53.4%; P<0.001).1 Which of the following statements, if any, are true?

a) Cases are those individuals with diagnosed lung cancer
b) Controls are those not exposed to smoking in childhood
c) All children were followed up until adulthood, with passive smoking and lung cancer status recorded
d) From the above study results we can conclude that lung cancer in adulthood is associated with passive smoking in childhood

Show Answers

Answer

a,d—Answer a is true. In a case-control study, patients are assigned to one of two groups on the basis of their disease status: one group with the disease (the cases) and one without (the controls). The cases and controls are then compared to ascertain if certain factors occur more frequently in one group than the other. In this study, the cases are individuals with diagnosed lung cancer and the controls those without. This case-control study is retrospective in design, with individuals asked about exposure to smoking in childhood. The aim of the case-control study was to provide insight into which factors may raise or lessen the risk of lung cancer.

Answer d is true. Lung cancer in adulthood is associated with passive smoking in childhood; that is, those participants with diagnosed lung cancer were more likely to have experienced passive smoking in childhood. However, a distinction must be made between association and causation. In particular, although individuals with diagnosed lung cancer were more likely to have experienced passive smoking in childhood, this finding does not necessarily mean that passive smoking caused the lung cancer and is, therefore, a risk factor.

The most compelling evidence for a causal link between a potential risk factor and disease would come from an experimental study. However, it would be unethical to undertake a trial where children were randomly assigned to passive smoking or no smoking exposure and then observed for the occurrence of lung cancer until adulthood. A series of nine criteria can be used to assess evidence for a causal link in observational studies, including case-control studies.2 Known as the Bradford-Hill criteria, these requirements are widely used to determine if the relationship between a disease and its supposed risk factor is a causal one or a simple association.

Answer b is false. By definition cases and controls are selected on the basis of their disease status, namely whether they have diagnosed lung cancer. Both groups of individuals then self reported their exposure to smoking in childhood.

Answer c is false. A case-control study is retrospective in design, whereas answer c describes a prospective cohort study. In the above study, cases and controls were selected on the basis of lung cancer status and asked about exposure to smoking in childhood

Cite this as: BMJ 2009;339:b4135

References

  1. Kohlmeier L, Arminger G, Bartolomeycik S, Bellach B, Rehm J, Thamm M. Pet birds as an independent risk factor for lung cancer: case-control study. BMJ 1992;305:986-9.[Abstract/Free Full Text]
  2. Hill AB. The environment and disease: association or causation? Proc R Soc Med 1965;58:295-300.[Web of Science][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to StumbleUpon StumbleUpon   Add to Technorati Technorati    What's this?

Relevant Article

Pet birds as an independent risk factor for lung cancer: case-control study.
L. Kohlmeier, G. Arminger, S. Bartolomeycik, B. Bellach, J. Rehm, and M. Thamm
BMJ 1992 305: 986-989. [Abstract] [PDF]

Access jobs at BMJ Careers
Whats new online at Student BMJ