PDF version

Parental smoking and risk of childhood cancer: hospital-based case–control study in Shiraz

Sir,

I read with interest the paper entitled “Parental smoking and risk of childhood cancer: hospital-based case–control study in Shiraz” by M. Edraki and M. Rambod published in the Eastern Mediterranean Health Journal [1].

At the beginning of the manuscript, the authors highlighted the scientific foundations of the issue through a large review of the literature. However, for a case–control study, basic principles related to the methodology must be respected to ensure valid results in accordance with the recommendations of the Strengthening of Observational Studies in Epidemiology (STROBE) statement [2].

First, the authors did not clearly define cases and controls and only indicated the eligibility criteria. They introduce a selection bias in selecting cases from hospital patients and the control group from public schools. The right approach would be to select the control group from the hospital also. Thus, the 2 populations would be comparable in all respects except for the disease (cancer) [3]. Controls chosen from outside the hospital may be different from cases in a variety of respects apart from cancer.

The study appears to have matched cases and controls on age and sex; however the authors performed an unconditional logistic regression analysis. In matched case–control studies conditional regression it is advisable to use in order to avoid getting conservatively biased estimates.

Values from the calculation of the size of the study sample do not seem correct; considering the same parameters and using software such as Epi-Info, the sample size should be 108 cases and 108 controls instead of 120 [4].

Finally, the authors did not take into account other variables or potential confounders (e.g. alcohol, food, residence) and did not make adjustments during analysis, which may call into question the conclusions of the article.

Case–control studies should meet the principles of conduct and reporting as outlined in the STROBE statement guidelines.

References

1. Edraki M, Rambod M. Parental smoking and risk of childhood cancer: hospital-based case–control study in Shiraz. Eastern Mediterranean Health Journal, 2011, 17(4):303–308.
2. Von Elm E et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement guidelines for reporting observational studies. Lancet, 2007, 370:1453–1457.
3. Rothman KJ, Greenland S, Lash LT. Modern epidemiology, 3rd ed. Philadelphia, PA, Lippincott Williams & Wilkins, 2008.
4. Czernichow P, Chaperon J, Le Contour X. Epidémiologie, connaissances et pratiques [Epidemiology: knowledge and practices]. Paris, Masson, 2001.

Diouf MassambaAssistant Researcher, Public Health Services, Université Cheikh Anta Diop de Dakar, Dakar, Sénégal

Dr Edraki and Dr Rambod were invited to respond to this letter, but no response was received.

Clinical management guidelines for pandemic (H1N1) 2009 virus infection in the Eastern Mediterranean Region: technical basis and overview

Sir,

The recent publication on clinical management guidelines for influenza is very useful and interesting [1]. However, some issues should be further discussed and added to the present guidelines. First, the guidelines should also address possible new or re-emerging influenza

infections, not only the 2009 virus. At present, there are warning signs of a new bird flu occurring and this could be a worldwide problem again [2]. Second, the guidelines exist as literature. The important issue is how to implement them and put into actual practice. Assessment of the adherence of practitioners to the guidelines would be useful to perform and could predict their success. Also, a previous report found that different groups of practitioners managed cases differently despite using the same guideline [3]. Therefore it was recommended that specific guidelines be set for less experienced practitioners such as internists [3].