Epidemiologic perspectives & innovations : EP+I最新文献

It is well known that the incidence odds ratio approximates the risk ratio when the disease of interest is rare, but increasingly overestimates the risk ratio as the disease becomes more common. However when assessing interaction, incidence odds ratios may not approximate risk ratios even when the disease is rare. We use the term "distributional interaction" to refer to interaction that appears when using incidence odds ratios that does not appear, or appears to a lesser degree, when using risk ratios. The interpretational problems that arise from this discrepancy can have important implications in epidemiologic research. Therefore, quantification of the relationship between the interaction odds ratio and the interaction risk ratio is warranted. In this paper, we provide a formula to quantify the differences between incidence odds ratios and risk ratios when they are used to estimate effect modification on a multiplicative scale. Using this formula, we examine the conditions under which these two estimates diverge. Furthermore, we expand this discussion to the implications of using incidence odds ratios to assess effect modification on an additive scale. Finally, we illustrate how distributional interaction arises and the problems that it causes using an example from the literature. Whenever the risk of the outcome variable is non-negligible, distributional interaction is possible. This is true even when the disease is rare (e.g., disease risk is less than 5%). Therefore, when assessing interaction on either an additive or multiplicative scale, caution should be taken in interpreting interaction estimates based on incidence odds ratios.

BACKGROUND: The WINPEPI (PEPI-for-Windows) computer programs for epidemiologists are designed for use in practice and research in the health field and as learning or teaching aids. They aim to complement other statistics packages. The programs are free, and can be downloaded from the Internet. IMPLEMENTATION: There are at present four WINPEPI programs: DESCRIBE, for use in descriptive epidemiology, COMPARE2, for use in comparisons of two independent groups or samples, PAIRSetc, for use in comparisons of paired and other matched observations, and WHATIS, a "ready reckoner" utility program. The programs contain 75 modules, each of which provides a number, sometimes a large number, of statistical procedures. The manuals explain the uses, limitations and applicability of specific procedures, and furnish formulae and references. CONCLUSIONS: WINPEPI provides a wide variety of statistical routines commonly used by epidemiologists, and is a handy resource for many procedures that are not very commonly used or easily found. The programs are in general user-friendly, although some users may be confused by the large numbers of options and results provided. The main limitations are the inability to read data files and the fact that only one of the programs presents graphic results. WINPEPI has a considerable potential as a learning and teaching aid.

Excess Years of Life Lost due to exposure is an important measure of health impact complementary to rate or risk statistics. I show that the total excess Years of Life Lost due to exposure can be estimated unbiasedly by calculating the corresponding excess Years of Potential Life Lost given conditions that describe study validity (like exchangeability of exposed and unexposed) and assuming that exposure is never preventive. I further demonstrate that the excess Years of Life Lost conditional on age at death cannot be estimated unbiasedly by a calculation of conditional excess Years of Potential Life Lost without adopting speculative causal models that cannot be tested empirically. Furthermore, I point out by example that the excess Years of Life Lost for a specific cause of death, like lung cancer, cannot be identified from epidemiologic data without assuming non-testable assumptions about the causal mechanism as to how exposure produces death. Hence, excess Years of Life Lost estimated from life tables or regression models, as presented by some authors for lung cancer or after stratification for age, are potentially biased. These points were already made by Robins and Greenland 1991 reasoning on an abstract level. In addition, I demonstrate by adequate life table examples designed to critically discuss the Years of Potential Life Lost analysis published by Park et al. 2002 that the potential biases involved may be fairly extreme. Although statistics conveying information about the advancement of disease onset are helpful in exposure impact analysis and especially worthwhile in exposure impact communication, I believe that attention should be drawn to the difficulties involved and that epidemiologists should always be aware of these conceptual limits of the Years of Potential Life Lost method when applying it as a regular tool in cohort analysis.

BACKGROUND: Epidemiologic research is often devoted to etiologic investigation, and so techniques that may facilitate mechanistic inferences are attractive. Some of these techniques rely on rigid and/or unrealistic assumptions, making the biologic inferences tenuous. The methodology investigated here is effect decomposition: the contrast between effect measures estimated with and without adjustment for one or more variables hypothesized to lie on the pathway through which the exposure exerts its effect. This contrast is typically used to distinguish the exposure's indirect effect, through the specified intermediate variables, from its direct effect, transmitted via pathways that do not involve the specified intermediates. METHODS: We apply a causal framework based on latent potential response types to describe the limitations inherent in effect decomposition analysis. For simplicity, we assume three measured binary variables with monotonic effects and randomized exposure, and use difference contrasts as measures of causal effect. Previous authors showed that confounding between intermediate and the outcome threatens the validity of the decomposition strategy, even if exposure is randomized. We define exchangeability conditions for absence of confounding of causal effects of exposure and intermediate, and generate two example populations in which the no-confounding conditions are satisfied. In one population we impose an additional prohibition against unit-level interaction (synergism). We evaluate the performance of the decomposition strategy against true values of the causal effects, as defined by the proportions of latent potential response types in the two populations. RESULTS: We demonstrate that even when there is no confounding, partition of the total effect into direct and indirect effects is not reliably valid. Decomposition is valid only with the additional restriction that the population contain no units in which exposure and intermediate interact to cause the outcome. This restriction implies homogeneity of causal effects across strata of the intermediate. CONCLUSIONS: Reliable effect decomposition requires not only absence of confounding, but also absence of unit-level interaction and use of linear contrasts as measures of causal effect. Epidemiologists should be wary of etiologic inference based on adjusting for intermediates, especially when using ratio effect measures or when absence of interacting potential response types cannot be confidently asserted.

Austin Bradford Hill's landmark 1965 paper contains several important lessons for the current conduct of epidemiology. Unfortunately, it is almost exclusively cited as the source of the "Bradford-Hill criteria" for inferring causation when association is observed, despite Hill's explicit statement that cause-effect decisions cannot be based on a set of rules. Overlooked are Hill's important lessons about how to make decisions based on epidemiologic evidence. He advised epidemiologists to avoid over-emphasizing statistical significance testing, given the observation that systematic error is often greater than random error. His compelling and intuitive examples point out the need to consider costs and benefits when making decisions about health-promoting interventions. These lessons, which offer ways to dramatically increase the contribution of health science to decision making, are as needed today as they were when Hill presented them.

As a supplement to our lead editorial, the editors of the new journal, Epidemiologic Perspectives & Innovations, provide a partial list of specific analyses and topic areas they would like to see submitted to the journal.

This editorial introduces the new online, open-access, peer-reviewed journal, Epidemiologic Perspectives & Innovations. Epidemiology (which we define broadly, to include clinical research and various approaches to studying the health of populations) is a critically important field in informing decisions about the health of individuals and populations. But the desire for new information means that the health science literature is overwhelmingly devoted to reporting new findings, leaving little opportunity to improve the quality of the science. By creating a journal dedicated to all topics of and about epidemiology, except standard research reports, we hope to encourage authors to write more on the neglected aspects of the field. The journal will publish articles that analyze policy implications of health research, present new research methods and better communicate existing methods, reassess previous results and dogma, and provide other innovations in and perspectives on the field. Online publishing will permit articles of whatever length is required for the work, speed the time to publication and allow free access to the full content.