Search PubMed⌕ Search

Biomedical subjects

S Greenland

Publications and source records attributed to S Greenland.

At least 19 recordsLinked to original sources
PubMed1992

Divergent biases in ecologic and individual-level studies.

Several authors have shown that ecologic estimates can be biased by effect modification and misclassification in a different fashion from individual-level estimates. This paper reviews and discusses ecologic biases induced by model misspecification; confounding; non-additivity of exposure and covariate effects (effect modification); exposure misclassification; and non-comparable standardization. Ecologic estimates can be more sensitive to these sources of bias than individual-level estimates, primarily because ecologic estimates are based on extrapolations to an unobserved conditional (individual-level) distribution. Because of this sensitivity, one should not rely on a single regression model for an ecologic analysis. Valid ecologic estimates are most feasible when one can obtain accurate estimates of exposure and covariate means in regions with internal exposure homogeneity and mutual covariate comparability; thus, investigators should seek out such regions in the design and analysis of ecologic studies.

Bias↗
PubMed1992

Methods for trend estimation from summarized dose-response data, with applications to meta-analysis.

Meta-analysis often requires pooling of correlated estimates to compute regression slopes (trends) across different exposure or treatment levels. The authors propose two methods that account for the correlations but require only the summary estimates and marginal data from the studies. These methods provide more efficient estimates of regression slope, more accurate variance estimates, and more valid heterogeneity tests than those previously available. One method also allows estimation of nonlinear trend components, such as quadratic effects. The authors illustrate these methods in a meta-analysis of alcohol use and breast cancer.

Alcohol Drinking↗
PubMed1992

A semi-Bayes approach to the analysis of correlated multiple associations, with an application to an occupational cancer-mortality study.

Thomas et al. presented the application of empirical-Bayes methods to the problem of multiple inference in epidemiologic studies. One limitation of their approach, which they noted, was the need to assume exchangeable log relative-risk parameters, and independent relative-risk estimates. Numerical integration was also required. Here I generalize their approach to allow for non-exchangeable parameters and non-independent estimates. The resulting method is Bayesian in so far as some feature of the prior distribution are specified from prior information, but is empirical Bayes in so far as some explicit parameters in the prior distribution are estimated from the data. Estimation is based on approximations to the posterior distribution; this allows one to implement the approach with standard software packages for matrix algebra. The method is illustrated in an occupational mortality study of 84 exposure-cancer associations.

Bayes Theorem↗
PubMed1992

Effects of nondifferential exposure misclassification in ecologic studies.

Although many authors have argued against inferring individual-level exposure-disease relations from ecologic data because of the potential "ecological fallacy." the availability of data from diverse populations promotes the continued use of this rapid and inexpensive study design. In ecologic studies, the exposure status of groups is often defined by the proportion of individuals exposed. In these studies, nondifferential exposure misclassification of individuals is shown to produce overestimation of exposure-disease associations that may be extreme when the ecologically derived rate ratios are applied to individuals. This overestimation contrasts with the bias toward the null resulting from nondifferential misclassification of a binary exposure in epidemiologic studies conducted at the individual level. Given the magnitude of the potential bias from nondifferential exposure misclassification and other sources, quantitative estimates of individual-level rate ratios from ecologic data should be interpreted with extreme caution.

Environmental Health↗
PubMed1992

Tests for trend and dose response: misinterpretations and alternatives.

Tests for overall trend, such as the Mantel extension test, are not tests for monotonic dose response. A survey of epidemiologic articles shows widespread misinterpretation of the Mantel extension test and overstatement of evidence for monotonic dose response when there are few exposed subjects. To properly evaluate the hypothesis that risk continues to increase with further increases in exposure, one must examine several statistics and estimates. Given sufficient data, nonparametric or polynomial regression analyses can provide more detailed dose-response information.

Epidemiologic Methods↗
PubMed1992

Identifiability and exchangeability for direct and indirect effects.

We consider the problem of separating the direct effects of an exposure from effects relayed through an intermediate variable (indirect effects). We show that adjustment for the intermediate variable, which is the most common method of estimating direct effects, can be biased. We also show that even in a randomized crossover trial of exposure, direct and indirect effects cannot be separated without special assumptions; in other words, direct and indirect effects are not separately identifiable when only exposure is randomized. If the exposure and intermediate never interact to cause disease and if intermediate effects can be controlled, that is, blocked by a suitable intervention, then a trial randomizing both exposure and the intervention can separate direct from indirect effects. Nonetheless, the estimation must be carried out using the G-computation algorithm. Conventional adjustment methods remain biased. When exposure and the intermediate interact to cause disease, direct and indirect effects will not be separable even in a trial in which both the exposure and the intervention blocking intermediate effects are randomly assigned. Nonetheless, in such a trial, one can still estimate the fraction of exposure-induced disease that could be prevented by control of the intermediate. Even in the absence of an intervention blocking the intermediate effect, the fraction of exposure-induced disease that could be prevented by control of the intermediate can be estimated with the G-computation algorithm if data are obtained on additional confounding variables.

Algorithms↗
PubMed1992

The effects of nondifferential confounder misclassification in ecologic studies.

In ecologic studies, covariate levels of groups are often quantified as the prevalence of a dichotomous covariate. We show that, under certain conditions, nondifferential misclassification of such a binary covariate does not reduce the ability to control confounding by the covariate in ecologic studies. Thus, any remaining exposure-disease association in an adjusted ecologic analysis cannot be ascribed to incomplete control for confounding due to nondifferential misclassification of the dichotomy under those conditions, although residual confounding by the underlying covariate may still be present. This point is illustrated by ecologic analyses of the association between population density and mortality from lung cancer in women in 30 administrative districts of the Federal Republic of Germany, in which control for cigarette smoking is required.

Confounding Factors, Epidemiologic↗
PubMed1991

Estimability and estimation of expected years of life lost due to a hazardous exposure.

Expected years of life lost is an important concept in public-health and legal issues. We describe conditions under which the expected years of life lost due to hazardous exposure is estimable (identifiable) from epidemiologic data. We show that, in general, the average years of life lost among exposed subjects dying at a given age (the age-specific expected years of life lost) is not identifiable, although the average years of life lost among all exposed subjects (the unconditional expected years of life lost) is identifiable from an unbiased epidemiologic study. We also show that the average years of life lost among all exposed subjects dying of a specific cause (the cause-specific expected years of life lost) is not identifiable. We discuss the implications of these results for compensation schemes based on years of life lost, and compare such schemes with those based on the probability of causation.

Accidents, Traffic↗
PubMed1991

Analytic methods for two-stage case-control studies and other stratified designs.

Nested case-control studies, or case-control studies within a cohort, combine the advantages of cohort studies with the efficiency of case-control studies. Case-control studies can often be viewed as having two stages; the first stage consists of vital status, disease, and basic covariate ascertainment, and the second stage consists of detailed covariate and exposure ascertainment. Breslow and Cain (1988) and Breslow and Zhao (1988) recently showed that conventional analyses of such two-stage studies may ignore some of the available information. In this paper, we show how one can adapt the pseudo-likelihood analyses developed by Kalbfleisch and Lawless (1988) to the analysis of data from two-stage case-control studies.

Case-Control Studies↗
PubMed1991

Estimating standardized parameters from generalized linear models.

Although the traditional unrestricted ('non-parametric') estimators of directly standardized rates and rate differences remain unbiased in sparse data, they tend to suffer from instability (low precision). As a result, many authors have proposed more precise estimators based on parametric models for the rates. This paper provides a general approach for constructing estimators of standardized parameters using generalized linear models, and shows that, in some common special cases, these model-based ('smoothed') estimators can have an exceptionally simple form.

Adult↗
PubMed1991

A mathematic analysis of the "epidemiologic necropsy".

There has been much debate regarding the validity and implications of the "epidemiologic necropsy." Two opposing mathematic analyses have previously appeared--one by Flanders and O'Brien and one by Feinstein and colleagues. Because these analyses omitted discussion of certain key parameters and assumptions, I provide a more detailed analysis. I show that under the assumptions used by earlier authors, the observed ("registry") disease rates will equal the true disease rates, even if there is differential disease detection and large pool of undetected cases. Under realistic violations of assumptions, both the observed rates and the necropsy estimates will be biased. In any case, the data offered thus far are insufficient to accurately assess the impact of undetected disease on epidemiologic studies.

Bias↗