ABSTRACT
Population attributable fraction (PAF) is an important public health parameter that is familiar to epidemiologists. The calculation of PAF is frequently based on the relative risk (RR) of a risk factor for a certain disease and the prevalence of exposure to the risk factor in the total population (Pe). We introduce here the method for making a nomogram linking PAF to RR and Pe. The nomogram will be very handy for public health professionals in estimating PAF quickly and accurately when information on RR and Pe is available.
Subject(s)
Humans , Epidemiologic Methods , Risk FactorsABSTRACT
<p><b>OBJECTIVE</b>Previously documented indirect adjustment methods could only adjust for the confounding effects from cigarette smoking. The aim of this paper is to introduce a novel method for dealing with the effects of smoking in occupational, epidemiological cohort studies using a 'smoking adjustment factor'.</p><p><b>METHODS</b>A retrospective cohort study among male silicotic workers purely exposed to silica dusts in Hong Kong (1981 - 1999) was used as an example. 'Smoking adjustment factor' in occupationally exposed smoking and nonsmoking sub-cohorts was expressed as 1/(1- PAR% )xRR and 1/1- PAR% respectively. Relative exposure effect and Synergy index were estimated to assess the multiplicative and additive interactions.</p><p><b>RESULTS</b>'Smoking adjustment factor' for non-smoking and smoking silicotic workers was 1/0.33 and 1/1.62 respectively. Lung cancer standardized mortality ratio(SMR) of all cohort members was reduced from 1.61(95% CI: 1.22-2.10) to 1.08(95% CI:0.81-1.41) after indirectly adjusted for smoking effect. Results from our novel indirect method were in line with that from Axelson' s approach. Relative silicosis effect and synergy index were estimated to be 0.63 (95% CI:0.08-0.79) and 0.90 (95 % CI:0.42-1.94) ,suggesting a significant but negative multiplicative interaction between smoking and silicosis on the risk of lung cancer mortality.</p><p><b>CONCLUSION</b>The merit of this new method was the ability to adjust for the confounding effect and evaluate the interactive effect with smoking. However, comparability of age distribution between occupationally exposed smoking and nonsmoking sub-cohorts was a prerequisite for the accurate estimations of the smoking indirectly adjusted SMR, relative exposure effect, and/or synergy index.</p>