A Decision-Analytic Approach to Addressing the Evidence About Football and Chronic Traumatic Encephalopathy.

Doubts can be raised about almost any assertion that a particular exposure can lead to an increase in a given adverse health effect. Even some of the most well-accepted causal associations in public health, such as that linking cigarette smoking to increased lung cancer risk, have intriguing research questions remaining to be answered. The inquiry whether an exposure causes a disease is never wholly a yes/no question but ought to follow from an appraisal of the weight of evidence supporting the positive conclusion in light of any coherent theories casting doubt on this evidence and the data supporting these. More importantly, such an appraisal cannot be made sensibly without considering the relative consequences to public health and economic welfare of specific actions based on unwarranted credulity (false positives) versus unwarranted skepticism (false negatives). Here we appraise the weight of evidence for the premise that repeated head impacts (RHIs) in professional football can increase the incidence of chronic traumatic encephalopathy (CTE) and, in turn, cause a variety of cognitive and behavioral symptoms. We first dismiss four logical fallacies that should not affect the appraisal of the weight of evidence. We then examine four alternative hypotheses in which RHI is not associated with CTE or symptoms (or both), and we conclude that the chances are small that the RHI→ CTE→ symptoms link is coincidental or artifactual. In particular, we observe that there are many specific interventions for which, even under a skeptical appraisal of the weight of evidence, the costs of a false positive are smaller than the false negative costs of refusing to intervene.

Assessment and management of residential radon health risks: a report from the health Canada radon workshop.

Epidemiologic studies of uranium miners and other underground miners have consistently shown miners exposed to high levels of radon to be at increased risk of lung cancer. More recently, concern has arisen about lung cancer risks among people exposed to lower levels of radon in homes. The current Canadian guideline for residential radon exposure was set in 1988 at 800 Bq/m(3). Because of the accumulation of a considerable body of new scientific evidence on radon lung cancer risks since that time, Health Canada sponsored a workshop to review the current state-of-the-science on radon health risks. The specific objectives of the workshop were (1) to collect and assess scientific information relevant to setting national radon policy in Canada, and (2) to gather information on social, political, and operational considerations in setting national policy. The workshop, held on 3-4 March 2004, was attended by 38 invited scientists, regulators, and other stakeholders from Canada and the United States. The presentations on the first day dealt primarily with scientific issues. The combined analysis of North American residential radon and lung cancer studies was reviewed. The analysis confirmed a small but detectable increase in lung cancer risk at residential exposure levels. Current estimates suggest that radon in homes is responsible for approximately 10% of all lung cancer deaths in Canada, making radon the second leading cause of lung cancer after tobacco smoking. This was followed by a perspective from an UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation) working group on radon. There were two presentations on occupational exposures to radon and two presentations considered the possibility of radon as a causative factor for cardiovascular disease and for cancer in other organs besides the lung. The possible contribution of environmental tobacco smoke to lung cancers in nonsmokers was also considered. Areas for future research were identified. The second day was devoted to policy and operational issues. The presentations began with a perspective from the U.S. Environmental Protection Agency, followed by a history of radon policy development in Canada. Subsequent presentations dealt with the cost-effectiveness of radon mitigation, Canadian building codes and radon, and a summary of radon standards from around the world. Provincial representatives and a private consultant were given opportunities to present their viewpoints. A number of strategies for reducing residential radon exposure in Canada were recognized, including testing and mitigation of existing homes (on either a widespread or targeted basis) and changing the building code to require that radon mitigation devices be installed at the time a new home is constructed. The various elements of a comprehensive national radon policy were set forth.

Approximations and heuristics for the "cause-modified" life table.

Life-table analysis can help to gauge the lifetime impacts that accrue from modifications to (age-specific) baseline mortality. Modifications of interest include those stemming from risk-factor-related exposures or from interventions. The specific algorithm used in these analyses can be called a cause-modified life table (a generalization of the cause-deleted life table). The author presents an approach for approximating that algorithm and uses it to obtain remarkably simplified expressions for approximating three indices of common interest: life-years lost (LYL), excess lifetime risk ratio (ELRR), and risk of exposure-induced death (REID). These efforts are restricted to the special case of multiplicative increases to baseline mortality (modeled as an excess rate ratio, ERR). The simplified expressions effectively "break open" what is often treated as a "black-box" calculation. Several insights result. For a practical range of risk factor impacts (ERRs), each index can be related to the ERR as a function of a baseline summary statistic and a "characteristic number" specific to the population and cause of interest. Conveniently, those numbers help form "rules of thumb" for translating among the three indices and suggest heuristics for extrapolating indices across populations and causes of death.

Residential radon in Canada: an uncertainty analysis of population and individual lung cancer risk.

Following a comprehensive evaluation of the health risks of radon, the U.S. National Research Council (US-NRC) concluded that the radon inside the homes of U.S. residents is an important cause of lung cancer. To assess lung cancer risks associated with radon exposure in Canadian homes, we apply the new (US-NRC) techniques, tailoring assumptions to the Canadian context. A two-dimensional uncertainty analysis is used to provide both population-based (population attributable risk, PAR; excess lifetime risk ratio, ELRR; and life-years lost, LYL) and individual-based (ELRR and LYL) estimates. Our primary results obtained for the Canadian population reveal mean estimates for ELRR, PAR, and LYL are 0.08, 8%, and 0.10 years, respectively. Results are also available and stratified by smoking status (ever versus never). Conveniently, the three indices (ELRR, PAR, and LYL) reveal similar output uncertainty (geometric standard deviation, GSD approximately 1.3), and in the case of ELRR and LYL, comparable variability and uncertainty combined (GSD approximately 4.2). Simplifying relationships are identified between ELRR, LYL, PAR, and the age-specific excess rate ratio (ERR), which suggest a way to scale results from one population to another. This insight is applied in scaling our baseline results to obtain gender-specific estimates, as well as in simplifying and illuminating sensitivity analysis.

Residual Perchloroethylene in Dry-Cleaned Acetate: The Effect of Pressing and Extent of Inter-Dry-Cleaner Variability.

Measurements of residual perchloroethylene (PCE), a dry-cleaning solvent associated with human health effects, were made in dry-cleaned acetate cloth to enable improved characterizations of both occupational and environmental exposure. A limited sample size (25 acetate cloths) was used to explore the extent of inter-dry-cleaner variability in residual PCE and to characterize the effect of the pressing operation on residual PCE. A new method, which uses carbon-disulfide as the direct extracting agent, proved effective in the analysis of residual PCE, with a recovery-efficiency ≈ 75%. Inter-dry-cleaner variability of residual PCE, although marginally statistically significant, was relatively low, showing only a fourfold range compared to a 5-order-of-magnitude range obtained from Kawauchi and Nishiyama1. Pairwise comparison of residual PCE in nonpressed versus pressed acetate samples revealed a statistically significant reduction (p < 0.008), which amounted to a consistent (among dry-cleaners) pressing-related removal efficiency of 75 ± 4%. A preliminary assessment of the source term associated with the pressing operation (mass PCE liberated per kg cloth dry-cleaned, SPCE ≈ 30 mg/kg) indicates a minor contribution to the average ambient air concentrations within dry-cleaning establishments.