Abstract: Poster Presentations |

Estimating the Standard Mortality Ratio for Lung Cancer due to the Confounding Effects of Smoking FREE TO VIEW

Robert M. Ross, MD*
Author and Funding Information

Baylor College of Medicine, Houston, TX


Chest. 2004;126(4_MeetingAbstracts):850S. doi:10.1378/chest.126.4_MeetingAbstracts.850S
Text Size: A A A
Published online


PURPOSE:  The Standard Mortality Ratio (SMR) is frequently used to assess the risk of lung cancer associated with exposure to various substances. However, it is virtually impossible to adequately control for THE major confounder, smoking. This abstract presents a mathematical formula and tableTable #1

Standard Mortality Ratios (SMR) for lung cancer solely due to smoking for study groups that have more ever-smokers than the general population. Values displayed are for examples of different prevalences of ever-smokers and different relative-risks associated with ever-smoking between the study group and the general population. For example, if a study group has an ever-smoking prevalence of 60% and those people have an average relative-risk which is 15-fold that of non-smokers while the general population has an ever-smoking prevalence of 40% and their relative-risk is 10-fold, the smoking-related SMR would be 2.04 (a).

Relative-Risk-of-Death from Lung Cancer for Ever-Smokers compared to Non-Smokers for a Study Group compared to the General PopulationRelationship A Study Group 15-fold General Population 15-foldRelationship B Study Group 20-fold General Population 15-foldRelationship C Study Group 15-fold General Population 10-foldRelationship D Study Group 20-fold General Population 10-foldPrevalence of Ever-Smokers in the General Popula- tion60%1.151.301.451.521.721.931.691.912.132.232.532.8350%1.181.351.531.551.792.031.711.962.222.252.602.9540%1.211.421.641.591.882.171.742.04(a)2.352.282.703.1130%1.271.541.811.652.022.381.782.162.542.322.843.3520%1.371.742.111.762.262.761.862.362.862.393.073.7510%20%30%10%20%30%10%20%30%10%20%30%

Absolute Increase in Prevalence of Ever-Smokers in a study Group compared to the General Population

that determine what the SMR would be, based solely on differences in smoking-related factors between a study group and the general population.

METHODS:  Figure #1presents the equation. In order to accurately determine the smoking-related SMR, the prevalence of ever-smokers and their relative-risk, for both the study group and the general population, must be known. This is not possible. However, presently approximately 50% of men are ever-smokers and their relative-risk for lung cancer is about 15 times that of non-smokers. The percentage and relative-risk are decreasing. Different study groups have ever-smoker prevalences up to about 80% and relative-risks up to about 20. From these ranges for the 4 variables, estimates can made and the smoking-related SMR obtained from table #1.

RESULTS:  To use table #1, first choose the sub-column that most closely reflects the relationship between the ever-smoking-relative-risks for the study group and the general population (for example, relationship B). The smoking-related SMR is then found at the intersection of the row showing the prevalence of ever-smokers in the general population and the column indicating the absolute percent increase in ever-smokers in the study group compared to the general population. Small differences can cause a significant increase in the SMR. However, this is not due to the substance being studied.

CONCLUSION:  A study group that has more ever-smokers, who on average have more smoke exposure or have other factors that increase their risk of death from lung cancer will have an increased SMR. Further, the SMR increase is magnified as the prevalence and relative-risk associated with ever-smoking decreases in the general population.

CLINICAL IMPLICATIONS:  Elevated SMR’s are generally inadequate as the sole criteria to determine if exposure to a substance increases the risk of lung cancer.

DISCLOSURE:  R.M. Ross, None.

Wednesday, October 27, 2004

12:30 PM- 2:00 PM




Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543