Effects of varying key assumptions on estimates of future mortality caused by second hand smoke (SHS)
| Assumption | Alternative conditions | Effect on number of future deaths per year |
| 1. SHS raises risk of stroke death (RR for men 2.1, for women 1.7) | SHS does not affect the risk of dying from stroke | Decreased by 26% (new n=243.2) |
| 2. Best estimate RR from meta-analyses or original papers | Upper 95% confidence limit of all RRs | Increased by 52% (n=490.4) |
| Lower 95% confidence limit of all RRs | Decreased by 46% (n=174.6) | |
| 3. Ex-smokers susceptible to effects of SHS | Ex-smokers not susceptible to effects of SHS: proportion of deaths among never smokers: heart disease 0.22 (males), 0.28 (females)12; stroke 0.3113 | Decreased by 45% (n=180.0) |
| 4. Current smokers not susceptible to SHS | Regard all deaths from lung cancer, heart disease, and stroke as influenced by SHS | Increased by 87% (n=606.8) |
| 5. Workplace exposure to SHS affects risk of heart disease beyond age 65 | Workplace exposure to SHS does not affect risk beyond age 65 | Decreased by 6% (n=306.4) |
| 6. Exposure to SHS at work excludes exposures at tea and lunch breaks, pe = 0.135 | Exposure at work defined as “more than 1 cigarette per day smoked around you at work, including lunch and tea breaks”, pe=0.356 (males), 0.246 (females)9 | Increased by 16% (n=390.5) |
| 7. No change in future exposures to SHS | Prevalence of SHS exposure reduced by: 50% in the home; 90% at work | Decreased by 55% (n=146.0) |
| 8. Effect of exposure to SHS at home is independent of exposure at work | Exposure to SHS at work does not add to risk of disease in those exposed also at home (effectively, prevalence of exposure at work reduced by 20%) | Decreased by 4% (n=312.3) |
RR, relative risk.