Recalculation of Summarised Odds Ratios for the Relationship between Passive Smoking and Lung Cancer Based on Data in the EPA Report

Abstract

We have recalculated the summarised odds ratios, with 95 % confidence intervals (CIs), for the relationship between exposure to environmental tobacco smoke (ETS) and lung cancer based on the odds ratios for the individual epidemiological studies given in the Environmen tal Protection Agency Report in 1992. Using DerSimonian's method, the summarised odds ratio, with 95% CI, for all the studies combined was estimated as 1.28 (1.13, 1.46). This method was used as the odds ratios for the individual studies indicated heterogeneity. This heterogeneity may arise because of differences in factors such as socio-economic status, hous ing or occupational environment among the countries where the epidemiological studies were carried out. If so, the summarised odds ratio for all the studies combined may not be an important parameter. Using Sugita's method, which enables us to eliminate publication bias indirectly, the summarised odds ratio, with 95% CI, for the 11 studies in the USA was estimated as 1.11 (0.97, 1.28). Without eliminating publication bias, the value was 1.19(1.01, 1.38). Including odds ratios from two new studies, the summarised odds ratio for the USA was found to be nearly equal to the value with publication bias eliminated. The odds ratios, with 95% CIs, summarised above do not demonstrate whether or not ETS is a significant risk factor for lung cancer. The association between ETS and lung cancer is certainly weak.