The Importance of Scale for Spatial-confounding Bias and Precision of Spatial Regression Estimators

Increasingly, regression models are used when residuals are spatially correlated. Prominent examples include studies in environmental epidemiology to understand the chronic health effects of pollutants. I consider the effects of residual spatial structure on the bias and precision of regression coefficients, developing a simple framework in which to understand the key issues and derive informative analytic results. When the spatial residual is induced by an unmeasured confounder, regression models with spatial random effects and closely-related models such as kriging and penalized splines are biased, even when the residual variance components are known. Analytic and simulation results show how the bias depends on the spatial scales of the covariate and the residual; bias is reduced only when there is variation in the covariate at a scale smaller than the scale of the unmeasured confounding. I also discuss how the scales of the residual and the covariate affect efficiency and uncertainty estimation when the residuals can be considered independent of the covariate. In an application on the association between black carbon particulate matter air pollution and birth weight, controlling for large-scale spatial variation appears to reduce bias from unmeasured confounders, while increasing uncertainty in the estimated pollution effect.

Analyzing Bivariate Survival Data with Interval Sampling and Application to Cancer Epidemiology

In medical follow-up studies, ordered bivariate survival data are frequently encountered when bivariate failure events are used as the outcomes to identify the progression of a disease. In cancer studies interest could be focused on bivariate failure times, for example, time from birth to cancer onset and time from cancer onset to death. This paper considers a sampling scheme where the first failure event (cancer onset) is identified within a calendar time interval, the time of the initiating event (birth) can be retrospectively confirmed, and the occurrence of the second event (death) is observed sub ject to right censoring. To analyze this type of bivariate failure time data, it is important to recognize the presence of bias arising due to interval sampling. In this paper, nonparametric and semiparametric methods are developed to analyze the bivariate survival data with interval sampling under stationary and semi-stationary conditions. Numerical studies demonstrate the proposed estimating approaches perform well with practical sample sizes in different simulated models. We apply the proposed methods to SEER ovarian cancer registry data for illustration of the methods and theory.

ASSESSING THE UNRELIABILITY OF THE MEDICAL LITERATURE: A RESPONSE TO "WHY MOST PUBLISHED RESEARCH FINDINGS ARE FALSE"

A recent article in this journal (Ioannidis JP (2005) Why most published research findings are false. PLoS Med 2: e124) argued that more than half of published research findings in the medical literature are false. In this commentary, we examine the structure of that argument, and show that it has three basic components: 1)An assumption that the prior probability of most hypotheses explored in medical research is below 50%. 2)Dichotomization of P-values at the 0.05 level and introduction of a “bias” factor (produced by significance-seeking), the combination of which severely weakens the evidence provided by every design. 3)Use of Bayes theorem to show that, in the face of weak evidence, hypotheses with low prior probabilities cannot have posterior probabilities over 50%. Thus, the claim is based on a priori assumptions that most tested hypotheses are likely to be false, and then the inferential model used makes it impossible for evidence from any study to overcome this handicap. We focus largely on step (2), explaining how the combination of dichotomization and “bias” dilutes experimental evidence, and showing how this dilution leads inevitably to the stated conclusion. We also demonstrate a fallacy in another important component of the argument –that papers in “hot” fields are more likely to produce false findings. We agree with the paper’s conclusions and recommendations that many medical research findings are less definitive than readers suspect, that P-values are widely misinterpreted, that bias of various forms is widespread, that multiple approaches are needed to prevent the literature from being systematically biased and the need for more data on the prevalence of false claims. But calculating the unreliability of the medical research literature, in whole or in part, requires more empirical evidence and different inferential models than were used. The claim that “most research findings are false for most research designs and for most fields” must be considered as yet unproven.

TRENDS IN PARTICULATE MATTER AND MORTALITY: AN APPROACH TO THE ASSESSMENT OF UNMEASURED CONFOUNDING

We propose a method for diagnosing confounding bias under a model which links a spatially and temporally varying exposure and health outcome. We decompose the association into orthogonal components, corresponding to distinct spatial and temporal scales of variation. If the model fully controls for confounding, the exposure effect estimates should be equal at the different temporal and spatial scales. We show that the overall exposure effect estimate is a weighted average of the scale-specific exposure effect estimates. We use this approach to estimate the association between monthly averages of fine particles (PM2.5) over the preceding 12 months and monthly mortality rates in 113 U.S. counties from 2000-2002. We decompose the association between PM2.5 and mortality into two components: 1) the association between “national trends” in PM2.5 and mortality; and 2) the association between “local trends,” defined as county-specificdeviations from national trends. This second component provides evidence as to whether counties having steeper declines in PM2.5 also have steeper declines in mortality relative to their national trends. We find that the exposure effect estimates are different at these two spatio-temporalscales, which raises concerns about confounding bias. We believe that the association between trends in PM2.5 and mortality at the national scale is more likely to be confounded than is the association between trends in PM2.5 and mortality at the local scale. If the association at the national scale is set aside, there is little evidence of an association between 12-month exposure to PM2.5 and mortality.