Search

Recurrent events—outcomes that an individual can experience repeatedly over the course of follow-up—are common in epidemiologic and health services research. Studies involving recurrent events often focus on time to first occurrence or on event rates, which assume constant hazards over time. In this paper, we contextualize recurrent event parameters of interest using counterfactual theory in a causal inference framework and describe an approach for estimating a target parameter referred to as the mean cumulative count. This approach leverages inverse probability weights to control measured confounding with an existing (and underutilized) nonparametric estimator of recurrent event burden first proposed by Dong et al. in 2015. We use simulations to demonstrate the unbiased estimation of the mean cumulative count using the weighted Dong-Yasui estimator in a variety of scenarios. The weighted Dong-Yasui estimator for the mean cumulative count allows researchers to use observational data to flexibly estimate and contrast the expected number of cumulative events experienced per individual by a given time point under different exposure regimens. We provide code to ease application of this method. [ABSTRACT FROM AUTHOR]

In diagnostic medicine, the true disease status of a patient is often represented on an ordinal scale—for example, cancer stage (0, I, II, III, or IV) or coronary artery disease severity measured using the Coronary Artery Disease Reporting and Data System (CAD-RADS) scale (none, minimal, mild, moderate, severe, or occluded). With advances in quantitation of diagnostic images and in artificial intelligence (AI), both supervised and unsupervised algorithms are being developed to help physicians correctly grade disease. Most of the diagnostic accuracy literature deals with binary disease status (disease present or absent); however, tests diagnosing ordinal-scaled diseases should not be reduced to a binary status just to simplify diagnostic accuracy testing. In this paper, we propose different characterizations of ordinal-scale accuracy for different clinical use scenarios, along with methods for comparing tests. In the simplest scenario, just the proportion of correct grades is considered; other scenarios address the magnitude and direction of misgrading; and at the other extreme, a weighted accuracy measure with weights based on the relative costs of different types of misgrading is presented. The various scenarios are illustrated using a coronary artery disease example where the accuracy of AI algorithms in providing patients with the correct CAD-RADS grade is assessed. [ABSTRACT FROM AUTHOR]

Epidemiologists commonly use an adjusted hazard ratio or incidence density ratio, or a standardized mortality ratio, to measure a difference in all-cause mortality rates. They seldom translate it into an age-, time-, or probability-based measure that would be easier to communicate and to relate to. Several articles have shown how to translate from a standardized mortality ratio or hazard ratio to a longevity difference, a difference in actuarial ages, or a probability of being outlived. In this paper, we describe the settings where these translations are and are not appropriate and provide some of the heuristics behind the formulae. The tools that yield differences in "effective age" and in longevity are applicable when both 1) the mortality rate ratio (hazard ratio) is constant over age and 2) the rates themselves are log-linear in age. The "probability/odds of being outlived" metric is applicable whenever the first condition holds, and thus it provides no direct information on the magnitude of the effective age/longevity difference. [ABSTRACT FROM AUTHOR]

Principled methods with which to appropriately analyze missing data have long existed; however, broad implementation of these methods remains challenging. In this and 2 companion papers (Am J Epidemiol. 2018;187(3):576-584 and Am J Epidemiol. 2018;187(3):585-591), we discuss issues pertaining to missing data in the epidemiologic literature. We provide details regarding missing-data mechanisms and nomenclature and encourage the conduct of principled analyses through a detailed comparison of multiple imputation and inverse probability weighting. Data from the Collaborative Perinatal Project, a multisite US study conducted from 1959 to 1974, are used to create a masked data-analytical challenge with missing data induced by known mechanisms. We illustrate the deleterious effects of missing data with naive methods and show how principled methods can sometimes mitigate such effects. For example, when data were missing at random, naive methods showed a spurious protective effect of smoking on the risk of spontaneous abortion (odds ratio (OR) = 0.43, 95% confidence interval (CI): 0.19, 0.93),while implementation of principled methods multiple imputation (OR = 1.30, 95% CI: 0.95, 1.77) or augmented inverse probability weighting (OR = 1.40, 95% CI: 1.00, 1.97) provided estimates closer to the "true" full-data effect (OR = 1.31, 95% CI: 1.05, 1.64). We call for greater acknowledgement of and attention to missing data and for the broad use of principled missing-data methods in epidemiologic research. [ABSTRACT FROM AUTHOR]

The authors respond to a commentary on a study which evaluated properties of machine learning methods for estimating causal effects. They point out the need for correct parametric regression model specification when using machine learning methods. They describe two possible scenarios when using stacked regression to estimate causal effects. They agree with conducting study-specific simulations to evaluate potential issues with the use of different algorithms in the stacked regression library.

A conventional Mendelian randomization analysis assesses the causal effect of a risk factor on an outcome by using genetic variants that are solely associated with the risk factor of interest as instrumental variables. However, in some cases, such as the case of triglyceride level as a risk factor for cardiovascular disease, it may be difficult to find a relevant genetic variant that is not also associated with related risk factors, such as other lipid fractions. Such a variant is known as pleiotropic. In this paper, we propose an extension of Mendelian randomization that uses multiple genetic variants associated with several measured risk factors to simultaneously estimate the causal effect of each of the risk factors on the outcome. This “multivariable Mendelian randomization” approach is similar to the simultaneous assessment of several treatments in a factorial randomized trial. In this paper, methods for estimating the causal effects are presented and compared using real and simulated data, and the assumptions necessary for a valid multivariable Mendelian randomization analysis are discussed. Subject to these assumptions, we demonstrate that triglyceride-related pathways have a causal effect on the risk of coronary heart disease independent of the effects of low-density lipoprotein cholesterol and high-density lipoprotein cholesterol. [ABSTRACT FROM PUBLISHER]

Estimation of causal effects of time-varying exposures using longitudinal data is a common problem in epidemiology. When there are time-varying confounders, which may include past outcomes, affected by prior exposure, standard regression methods can lead to bias. Methods such as inverse probability weighted estimation of marginal structural models have been developed to address this problem. However, in this paper we show how standard regression methods can be used, even in the presence of time-dependent confounding, to estimate the total effect of an exposure on a subsequent outcome by controlling appropriately for prior exposures, outcomes, and time-varying covariates. We refer to the resulting estimation approach as sequential conditional mean models (SCMMs), which can be fitted using generalized estimating equations. We outline this approach and describe how including propensity score adjustment is advantageous. We compare the causal effects being estimated using SCMMs and marginal structural models, and we compare the two approaches using simulations. SCMMs enable more precise inferences, with greater robustness against model misspecification via propensity score adjustment, and easily accommodate continuous exposures and interactions. A new test for direct effects of past exposures on a subsequent outcome is described. [ABSTRACT FROM AUTHOR]

Missing data is a common occurrence in epidemiologic research. In this paper, 3 data sets with induced missing values from the Collaborative Perinatal Project, a multisite US study conducted from 1959 to 1974, are provided as examples of prototypical epidemiologic studies with missing data. Our goal was to estimate the association of maternal smoking behavior with spontaneous abortion while adjusting for numerous confounders. At the same time, we did not necessarily wish to evaluate the joint distribution among potentially unobserved covariates, which is seldom the subject of substantive scientific interest. The inverse probability weighting (IPW) approach preserves the semiparametric structure of the underlying model of substantive interest and clearly separates the model of substantive interest from the model used to account for the missing data. However, IPW often will not result in valid inference if the missing-data pattern is nonmonotone, even if the data are missing at random. We describe a recently proposed approach to modeling nonmonotone missing-data mechanisms under missingness at random to use in constructing the weights in IPW complete-case estimation, and we illustrate the approach using 3 data sets described in a companion article (AmJ Epidemiol. 2018;187(3):568-575). [ABSTRACT FROM AUTHOR]

Propensity score methods are a popular tool with which to control for confounding in observational data, but their bias-reduction properties--as well as internal validity, generally--are threatened by covariate measurement error. There are few easy-to-implement methods of correcting for such bias. In this paper, we describe and demonstrate how existing sensitivity analyses for unobserved confounding--propensity score calibration, VanderWeele and Arah's bias formulas, and Rosenbaum's sensitivity analysis--can be adapted to address this problem. In a simulation study, we examine the extent to which these sensitivity analyses can correct for several measurement error structures: classical, systematic differential, and heteroscedastic covariate measurement error. We then apply these approaches to address covariate measurement error in estimating the association between depression and weight gain in a cohort of adults in Baltimore, Maryland. We recommend the use of VanderWeele and Arah's bias formulas and propensity score calibration (assuming it is adapted appropriately for the measurement error structure), as both approaches perform well for a variety of propensity score estimators and measurement error structures. [ABSTRACT FROM AUTHOR]

There have been recent proposals advocating the use of additive gene-environment interaction instead of the widely used multiplicative scale, as a more relevant public health measure. Using gene-environment independence enhances statistical power for testing multiplicative interaction in case-control studies. However, under departure from this assumption, substantial bias in the estimates and inflated type I error in the corresponding tests can occur. In this paper, we extend the empirical Bayes (EB) approach previously developed for multiplicative interaction, which trades off between bias and efficiency in a data-adaptive way, to the additive scale. An EB estimator of the relative excess risk due to interaction is derived, and the corresponding Wald test is proposed with a general regression setting under a retrospective likelihood framework. We study the impact of gene-environment association on the resultant test with case-control data. Our simulation studies suggest that the EB approach uses the gene-environment independence assumption in a data-adaptive way and provides a gain in power compared with the standard logistic regression analysis and better control of type I error when compared with the analysis assuming gene-environment independence. We illustrate themethods with data from the Ovarian Cancer Association Consortium. [ABSTRACT FROM AUTHOR]

Longitudinal data sources offer new opportunities for the evaluation of sequential interventions. To adjust for time-dependent confounding in these settings, longitudinal targeted maximum likelihood based estimation (TMLE), a doubly robust method that can be coupled with machine learning, has been proposed. This paper provides a tutorial in applying longitudinal TMLE, in contrast to inverse probability of treatment weighting and g-computation based on iterative conditional expectations. We apply these methods to estimate the causal effect of nutritional interventions on clinical outcomes among critically ill children in a United Kingdom study (Control of Hyperglycemia in Paediatric Intensive Care, 2008-2011). We estimate the probability of a child's being discharged alive from the pediatric intensive care unit by a given day, under a range of static and dynamic feeding regimes. We find that before adjustment, patients who follow the static regime "never feed" are discharged by the end of the fifth day with a probability of 0.88 (95% confidence interval: 0.87, 0.90), while for the patients who follow the regime "feed from day 3," the probability of discharge is 0.64 (95% confidence interval: 0.62, 0.66). After adjustment for time-dependent confounding, most of this difference disappears, and the statistical methods produce similar results. TMLE offers a flexible estimation approach; hence, we provide practical guidance on implementation to encourage its wider use. [ABSTRACT FROM AUTHOR]

Chain event graphs (CEGs) are a graphical representation of a statistical model derived from event trees. They have previously been applied to cohort studies but not to case-control studies. In this paper, we apply the CEG framework to a Yorkshire, United Kingdom, case-control study of childhood type 1 diabetes (1993-1994) in order to examine 4 exposure variables associated with the mother, 3 of which are fully observed (her school-leaving-age, amniocenteses during pregnancy, and delivery type) and 1 with missing values (her rhesus factor), while incorporating previous type 1 diabetes knowledge. We conclude that the unknown rhesus factor values were likely to be missing not at random and were mainly rhesus-positive. The mother's school-leaving-age and rhesus factor were not associated with the diabetes status of the child, whereas having at least 1 amniocentesis procedure and, to a lesser extent, birth by cesarean delivery were associated; the combination of both procedures further increased the probability of diabetes. This application of CEGs to case-control data allows for the inclusion of missing data and prior knowledge, while investigating associations in the data. Communication of the analysis with the clinical expert is more straightforward than with traditional modeling, and this approach can be applied retrospectively or when assumptions for traditional analyses are not held. [ABSTRACT FROM AUTHOR]

The number of operations that surgeons have previously performed is associated with their patients' outcomes. However, this association may not be causal, because previous studies have often been cross-sectional and their analyses have not considered time-varying confounding or positivity violations. In this paper, using the example of surgeons who perform coronary artery bypass grafting, we describe (hypothetical) target trials for estimation of the causal effect of the surgeons' operative volumes on patient mortality. We then demonstrate how to emulate these target trials using data from US Medicare claims and provide effect estimates. Our target trial emulations suggest that interventions on physicians' volume of coronary artery bypass grafting operations have little effect on patient mortality. The target trial framework highlights key assumptions and draws attention to areas of bias in previous observational analyses that deviated from their implicit target trials. The principles of the presented methodology may be adapted to other scenarios of substantive interest in health services research. [ABSTRACT FROM AUTHOR]

There is no complete solution for the problem of abuse of statistics, but methodological training needs to cover cognitive biases and other psychosocial factors affecting inferences. The present paper discusses 3 common cognitive distortions: 1) dichotomania, the compulsion to perceive quantities as dichotomous even when dichotomization is unnecessary and misleading, as in inferences based on whether a P value is "statistically significant"; 2) nullism, the tendency to privilege the hypothesis of no difference or no effect when there is no scientific basis for doing so, as when testing only the null hypothesis; and 3) statistical reification, treating hypothetical data distributions and statistical models as if they reflect known physical laws rather than speculative assumptions for thought experiments. As commonly misused, null-hypothesis significance testing combines these cognitive problems to produce highly distorted interpretation and reporting of study results. Interval estimation has so far proven to be an inadequate solution because it involves dichotomization, an avenue for nullism. Sensitivity and bias analyses have been proposed to address reproducibility problems (Am J Epidemiol. 2017;186(6):646--647); these methods can indeed address reification, but they can also introduce new distortions via misleading specifications for bias parameters. P values can be reframed to lessen distortions by presenting them without reference to a cutoff, providing them for relevant alternatives to the null, and recognizing their dependence on all assumptions used in their computation; they nonetheless require rescaling for measuring evidence. I conclude that methodological development and training should go beyond coverage of mechanistic biases (e.g., confounding, selection bias, measurement error) to cover distortions of conclusions produced by statistical methods and psychosocial forces. [ABSTRACT FROM AUTHOR]

Distributed networks of health-care data sources are increasingly being utilized to conduct pharmacoepidemio-logic database studies. Such networks may contain data that are not physically pooled but instead are distributed horizontally (separate patients within each data source) or vertically (separate measures within each data source) in order to preserve patient privacy. While multivariable methods for the analysis of horizontally distributed data are frequently employed, few practical approaches have been put forth to deal with vertically distributed healthcare databases. In this paper, we propose 2 propensity score-based approaches to vertically distributed data analysis and test their performance using 5 example studies. We found that these approaches produced point estimates close to what could be achieved without partitioning. We further found a performance benefit (i.e., lower mean squared error) for sequentially passing a propensity score through each data domain (called the "sequential approach") as compared with fitting separate domain-specific propensity scores (called the "parallel approach"). These results were validated in a small simulation study. This proof-of-concept study suggests a new multivariable analysis approach to vertically distributed health-care databases that is practical, preserves patient privacy, and warrants further investigation for use in clinical research applications that rely on health-care databases. [ABSTRACT FROM AUTHOR]

Bayesian methods have seen an increase in popularity in a wide variety of scientific fields, including epidemiology. One of the main reasons for their widespread application is the power of the Markov chain Monte Carlo (MCMC) techniques generally used to fit these models. As a result, researchers often implicitly associate Bayesian models with MCMC estimation procedures. However, Bayesian models do not always require Markov-chain-based methods for parameter estimation. This is important, as MCMC estimation methods, while generally quite powerful, are complex and computationally expensive and suffer from convergence problems related to the manner in which they generate correlated samples used to estimate probability distributions for parameters of interest. In this issue of the Journal, Cole et al. (Am J Epidemiol. 2012;175(5):368–375) present an interesting paper that discusses non-Markov-chain-based approaches to fitting Bayesian models. These methods, though limited, can overcome some of the problems associated with MCMC techniques and promise to provide simpler approaches to fitting Bayesian models. Applied researchers will find these estimation approaches intuitively appealing and will gain a deeper understanding of Bayesian models through their use. However, readers should be aware that other non-Markov-chain-based methods are currently in active development and have been widely published in other fields. [ABSTRACT FROM AUTHOR]

In several studies of antiretroviral treatment (ART) programs for persons with human immunodeficiency virus infection, investigators have reported that there has been a higher rate of loss to follow-up (LTFU) among patients initiating ART in recent years than among patients who initiated ART during earlier time periods. This finding is frequently interpreted as reflecting deterioration of patient retention in the face of increasing patient loads. However, in this paper we demonstrate by simulation that transient gaps in follow-up could lead to bias when standard survival analysis techniques are applied. We created a simulated cohort of patients with different dates of ART initiation. Rates of ART interruption, ART resumption, and mortality were assumed to remain constant over time, but when we applied a standard definition of LTFU, the simulated probability of being classified LTFU at a particular ART duration was substantially higher in recently enrolled cohorts. This suggests that much of the apparent trend towards increased LTFU may be attributed to bias caused by transient interruptions in care. Alternative statistical techniques need to be used when analyzing predictors of LTFU—for example, using “prospective” definitions of LTFU in place of “retrospective” definitions. Similar considerations may apply when analyzing predictors of LTFU from treatment programs for other chronic diseases. [ABSTRACT FROM PUBLISHER]

The very insightful and clear paper by VanderWeele and Vansteelandt in this issue of the Journal (Am J Epidemiol. 2010;172(12):1339–1348) bridges the gap between biostatistics methodologists focusing on causal methods for mediation analyses and the practitioners of mediational analyses to the benefit of both groups. In an effort to continue the bridging of this gap, this invited commentary relates the important issue of “natural direct effects” to the well-known epidemiologic method of direct standardization. Additionally, attention is paid to the importance of temporal sequencing to help substantiate the mediation relations among the exposure, mediation, and outcome. A crucial mathematical distortion under the logistics model, called “absence of collapsibility,” is noted in motivating VanderWeele and Vansteelandt's use of the log-linear model for comparing the effect of exposure adjusted for the mediator with the effect of exposure unadjusted for the mediator. It is also noted that this issue applies to one approach to assessing confounding. Finally, some issues are raised for consideration when testing the interaction between the exposure and mediator before assessing mediation. [ABSTRACT FROM PUBLISHER]

Secondary trait genetic association provides insight into the genetic architecture of disease etiology but requires caution in estimation. Ignoring case-control sampling may introduce bias into secondary trait association. In this paper, we compare the efficiency and robustness of various inverse probability weighted (IPW) estimators and maximum likelihood (ML) estimators. ML methods have been proposed but require correct modeling of both the secondary and the primary trait associations for valid inference. We show that ML methods using a misspecified primary trait model can severely inflate the type I error. IPW estimators are typically less efficient than ML estimators but are robust against model misspecification. When the secondary trait is available for the entire cohort, the IPW estimator with selection probabilities estimated nonparametrically and the augmented IPW estimator improve efficiency over the simple IPW estimator. We conclude that in large genetic association studies with complex sampling schemes, IPW-based estimators offer flexibility and robustness, and therefore are a viable option for analysis. [ABSTRACT FROM PUBLISHER]

In this paper, we propose new methods for analyzing cases of vaccine adverse events spontaneously reported to a surveillance database. The methods use the self-controlled case series approach, extended in several ways with parametric and nonparametric assumptions to account for the specific features of the data (large amount of underreporting and variation of reporting with time since vaccination). This work was motivated by the documented risk of intussusception after RotaShield vaccination (Wyeth-Lederle Vaccines, Radnor, Pennsylvania) and used worldwide spontaneous reports of intussusception occurring after Rotarix vaccination (GlaxoSmithKline Biologics, Research Triangle Park, North Carolina) collected between January 2004 and February 2010. The estimated risk during the 3- to 7-day period after vaccination was approximately 5 times higher after dose 1 of Rotarix than after dose 2, which is similar to published findings on the same topic. We undertook a large simulation study to evaluate the performance of the method in different scenarios, including its robustness to different sample sizes and time-dependent reporting functions. The bias was generally small, the type I error rate was correctly controlled, and the power to detect a risk ratio of 4 was satisfactory, provided that the sample size was over 100. The proposed methods are an effective way to explore and quantify vaccine safety signals from spontaneous reports. [ABSTRACT FROM PUBLISHER]

Consider a study in which the effect of a binary exposure on an outcome operates partly through a binary mediator but measurement of the mediator is nondifferentially misclassified. Suppose that an investigator wishes to estimate the direct and indirect effects of the exposure on the outcome. In this paper, the authors describe a mathematical correspondence between the empirical expressions for the natural direct effect and the effect of exposure among the unexposed standardized by a binary confounder. They then exploit this correspondence to prove that the direction of the bias due to nondifferential measurement error in estimating the natural direct and indirect effects is to overestimate the natural direct effect and underestimate the natural indirect effect. [ABSTRACT FROM AUTHOR]

The Life Journeys of Young Women Project is the first population-based study to examine the role of economic uncertainty throughout early adulthood on age at first childbirth. A retrospective cross-sectional component was added to an existing cohort study that is based on a birth cohort of women born during 1973–1975 in Adelaide, South Australia (n ˜ 1,000). An event history calendar instrument was used to obtain data regarding a range of life domains including partnering, educational attainment, home ownership, higher education debt, employment, and pregnancies over a 20-year period (sometimes as detailed as at monthly intervals). Interviews were conducted between 2007 and 2009. An analysis framework applying time-varying and time-constant survival analysis techniques within a life-course framework was developed that will guide analyses to examine the role of duration and life-course timing of economic uncertainty on age at first childbirth. This paper discusses study objectives and design, fieldwork procedures, planned statistical analyses, and recruitment outcomes, focusing on novel features that would facilitate analogous epidemiologic research. [ABSTRACT FROM AUTHOR]

In this paper, the authors describe different instrumental variable (IV) estimators of causal risk ratios and odds ratios with particular attention to methods that can handle continuously measured exposures. The authors present this discussion in the context of a Mendelian randomization analysis of the effect of body mass index (BMI; weight (kg)/height (m)2) on the risk of asthma at age 7 years (Avon Longitudinal Study of Parents and Children, 1991–1992). The authors show that the multiplicative structural mean model (MSMM) and the multiplicative generalized method of moments (MGMM) estimator produce identical estimates of the causal risk ratio. In the example, MSMM and MGMM estimates suggested an inverse relation between BMI and asthma but other IV estimates suggested a positive relation, although all estimates had wide confidence intervals. An interaction between the associations of BMI and fat mass and obesity-associated (FTO) genotype with asthma explained the different directions of the different estimates, and a simulation study supported the observation that MSMM/MGMM estimators are negatively correlated with the other estimators when such an interaction is present. The authors conclude that point estimates from various IV methods can differ in practical applications. Based on the theoretical properties of the estimators, structural mean models make weaker assumptions than other IV estimators and can therefore be expected to be consistent in a wider range of situations. [ABSTRACT FROM PUBLISHER]

Doubly robust estimation combines a form of outcome regression with a model for the exposure (i.e., the propensity score) to estimate the causal effect of an exposure on an outcome. When used individually to estimate a causal effect, both outcome regression and propensity score methods are unbiased only if the statistical model is correctly specified. The doubly robust estimator combines these 2 approaches such that only 1 of the 2 models need be correctly specified to obtain an unbiased effect estimator. In this introduction to doubly robust estimators, the authors present a conceptual overview of doubly robust estimation, a simple worked example, results from a simulation study examining performance of estimated and bootstrapped standard errors, and a discussion of the potential advantages and limitations of this method. The supplementary material for this paper, which is posted on the Journal's Web site (http://aje.oupjournals.org/), includes a demonstration of the doubly robust property (Web Appendix 1) and a description of a SAS macro (SAS Institute, Inc., Cary, North Carolina) for doubly robust estimation, available for download at http://www.unc.edu/∼mfunk/dr/. [ABSTRACT FROM PUBLISHER]

In observational studies on cancer patients, progression of performance status over time can be described by using a multistate model in which state-to-state transitions represent changes in a patient's health condition. Although a patient experiences transitions in continuous time, assessments on the patient are often made at irregularly spaced time points. In this paper, the authors formulate a Markov 4-state model for examining longitudinal data on performance status collected under intermittent observation. The cohort consisted of 11,342 patients diagnosed with cancer in Ontario, Canada, from 2007 to 2009. The authors extend the model to estimate the predicted probability of reaching the absorbing state, death, over various time intervals. The authors also illustrate what happens to the estimated transition intensities if the true observational scheme is overlooked. Methods for multistate analysis should be used by epidemiologists, since they prove particularly useful for examining the complexities of disease processes. [ABSTRACT FROM PUBLISHER]

The etiology, particularly the genetic basis, of multifactorial late-onset diseases is the subject of many genetic epidemiologic studies. The authors' aim in this paper was to investigate the circumstances under which competing risks can lead to bias in studies of genetic susceptibility to late-onset diseases. The authors used a model built in an epidemiologic framework to show that when a genetic risk factor and an environmental risk factor interact to cause a frequent competing risk of death, the measure of the association between the disease under investigation and the genetic risk factor will be biased if the environmental risk factor is also associated with the latter disease and is omitted from the analysis. This is an example of confounding bias, and it is the consequence of an association between the genetic risk factor and the environmental risk factor that appears over time. Numerical examples show that under certain conditions this bias can be substantial. The authors present several patterns of association in favor of such a bias. Because competing risks of death are likely to be present in older subjects, researchers studying the etiology of late-onset diseases should be aware of the possibility of this bias. [ABSTRACT FROM PUBLISHER]

In case-control studies, the crude odds ratio derived from a 2 × 2 table and the common odds ratio adjusted for stratification variables are staple measures of exposure-disease association. While missing exposure data are encountered in the majority of such studies, formal attempts to deal with them are rare, and a complete-case analysis is the norm. Furthermore, the probability that exposure is missing may depend on true exposure status, so the missing-at-random assumption is often unreasonable. In this paper, the authors present an adjustment to the usual product binomial likelihood to properly account for missing data. Estimation of model parameters without restrictive assumptions requires an additional data collection effort akin to a validation study. Closed-form results are provided to facilitate point and confidence interval estimation of crude and common odds ratios after properly accounting for informatively missing data. Simulations assess performance of the likelihood-based estimates and inferences, and they display the potential for bias in complete-case analyses. An example is presented to illustrate the approach. [ABSTRACT FROM PUBLISHER]

This paper introduces the projection methods for describing and testing the differences between pairs of continuous distributions. These methods include the projection plot, the projection spline, and the iter-1 test. The projection plot displays the difference between corresponding quantiles against the average of the corresponding quantiles. It is analogous to an empirical quantile-quantile plot that has been rotated 45 degrees. The projection spline is a knotted linear spline iteratively fit to the projection plot so that all knots are associated with significant changes in slope. It summarizes nonrandom deviations from linearity on the projection plot, allowing classification of the highest level of difference between two distributions as a difference in shape, in spread, or in location. The iter-1 test compares the first iteration of the projection spline with the line y = 0, providing a global test of difference between two distributions that is more powerful in simulations than either the chi-square test of independence or the Kolmogorov-Smirnov test. These methods will enhance epidemiologic practice by making the comparison of full distributions an accessible tool for routine data analysis. Am J Epidemiol 1997;146:1056–66. [ABSTRACT FROM AUTHOR]

This paper introduces the concept of vector diagnostics. In contrast to the conventional approach where one diagnosis takes precedence, the authors propose an alternative strategy that addresses the clinical reality of comorbidity and multiple diagnoses for an individual. Based on a Bayesian approach, the probability distribution for the etiologically heterogeneous dementia diagnoses is estimated from the Canadian Study of Health and Aging database. These data were collected between February 1991 and May 1992. This method facilitates the establishment of a probability for more than one diagnosis within a given individual. By analyzing the correspondence between diagnostic groups, it is demonstrated that some clinical diagnoses are not reliably distinguished on the basis of the considered subset of symptoms and signs. As a consequence, the conventional diagnostic categories might require revision. The resulting probabilistic algorithm allows for the mining of existing epidemiologic databases for patterns of signs and symptoms that characterize emerging diagnostic categories which might better account for the heterogeneity of the dementia subtypes and individual variability. [ABSTRACT FROM AUTHOR]

Most polio virus infections are silent. Vaccination reduces the incidence of infection, and the period between clinical cases of poliomyelitis becomes longer. As the point of eradication is approached, it becomes increasingly difficult to use the case-free period to determine whether silent infections have ceased. In this paper, the authors use stochastic computer simulations to relate case-free periods to the presence or absence of silent infections. After 2 years without paralytic cases in a population of 200, 000 inhabitants, the probability for the presence of silent infections can still be as high as 38%. The case-free period must exceed 3 years before one can be 95% certain that there has been local extinction of the wild polio virus infection. Even after 5 years without cases, the probability of silent polio virus transmission can still be in the range of 0.1–1.0%. 1996; 143: 816–22. [ABSTRACT FROM AUTHOR]

Recurrent events can occur more than once in the same individual; such events may be of different types, known as multitype recurrent events. They are very common in longitudinal studies. Often there is a terminating event, after which no further events can occur. The risk of any event, including terminating events such as death or cure, is typically affected by prior events. We propose a flexible joint multitype recurrent-events model that explicitly provides estimates of the change in risk for each event due to subject characteristics, including number and type of prior events and the absolute risk for every event type (terminating and nonterminating), and predicts event-free survival probability over a desired time period. The model is fully parametric, and therefore a standard likelihood function and robust standard errors can be constructed. We illustrate the model with applications to the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (1994–2002) and provide discussion of the results and model features. [ABSTRACT FROM AUTHOR]

Interest in using internet search data, such as that from the Google Health Trends Application Programming Interface (GHT-API), to measure epidemiologically relevant exposures or health outcomes is growing due to their accessibility and timeliness. Researchers enter search term(s), geography, and time period, and the GHT-API returns a scaled probability of that search term, given all searches within the specified geographic-time period. In this study, we detailed a method for using these data to measure a construct of interest in 5 iterative steps: first, identify phrases the target population may use to search for the construct of interest; second, refine candidate search phrases with incognito Google searches to improve sensitivity and specificity; third, craft the GHT-API search term(s) by combining the refined phrases; fourth, test search volume and choose geographic and temporal scales; and fifth, retrieve and average multiple samples to stabilize estimates and address missingness. An optional sixth step involves accounting for changes in total search volume by normalizing. We present a case study examining weekly state-level child abuse searches in the United States during the coronavirus disease 2019 pandemic (January 2018 to August 2020) as an application of this method and describe limitations. [ABSTRACT FROM AUTHOR]

Colorectal cancer (CRC) incidence rates have decreased among adults aged 50 years or older while increasing in adults under age 50 years. Understanding these trends is challenging because of the multiple related time scales of age, diagnosis period, and birth cohort. We analyzed incidence rates of rectal, distal colon, and proximal colon cancer for individuals aged 20 years or more from the Surveillance, Epidemiology, and End Results Program for diagnosis years 1978–2017. We used a 2-stage generalized linear model to determine age, period, and cohort effects for CRC incidence. We first estimated birth cohort effects among people under age 45 years. We used these results to specify prior distributions for cohort effects in a Bayesian model to estimate period effects among people aged 45 years or more. There was no evidence of period effects for people under age 45 years. Risks of rectal and distal colon cancer increased for later birth cohorts. Compared with the 1943–1952 birth cohort, the 1983–1992 birth cohort had 2.2 times the risk of rectal cancer, 1.9 times the risk of distal colon cancer, and 1.3 times the risk of proximal colon cancer. For people aged ≥45 years, period effects showed declines in CRC risk that were attributable to screening. [ABSTRACT FROM AUTHOR]

A covariate-adjusted estimate of an exposure-outcome association may be biased if the exposure variable suffers measurement error. We propose an approach to correct for exposure measurement error in a covariate-adjusted estimate of the association between a continuous exposure variable and outcome of interest. Our proposed approach requires data for a reference population in which the exposure was a priori set to some known level (e.g. 0, and is therefore unexposed); however, our approach does not require an exposure validation study or replicate measures of exposure, which are typically needed when addressing bias due to exposure measurement error. A key condition for this method, which we refer to as "partial population exchangeability," requires that the association between a measured covariate and outcome in the reference population equals the association between that covariate and outcome in the target population in the absence of exposure. We illustrate the approach using simulations and an example. [ABSTRACT FROM AUTHOR]

Inverse probability weighting (IPW) and targeted maximum likelihood estimation (TMLE) are methodologies that can adjust for confounding and selection bias and are often used for causal inference. Both estimators rely on the positivity assumption that within strata of confounders there is a positive probability of receiving treatment at all levels under consideration. Practical applications of IPW require finite inverse probability (IP) weights. TMLE requires that propensity scores (PS) be bounded away from 0 and 1. Although truncation can improve variance and finite sample bias, this artificial distortion of the IP weights and PS distribution introduces asymptotic bias. As sample size grows, truncation-induced bias eventually swamps variance, rendering nominal confidence interval coverage and hypothesis tests invalid. We present a simple truncation strategy based on the sample size, n , that sets the upper bound on IP weights at |$\sqrt{\textit{n}}$| ln  n /5. For TMLE, the lower bound on the PS should be set to 5/(⁠|$\sqrt{\textit{n}}$| ln  n /5). Our strategy was designed to optimize the mean squared error of the parameter estimate. It naturally extends to data structures with missing outcomes. Simulation studies and a data analysis demonstrate our strategy's ability to minimize both bias and mean squared error in comparison with other common strategies, including the popular but flawed quantile-based heuristic. [ABSTRACT FROM AUTHOR]

In the analysis of observational studies, inverse probability weighting (IPW) is commonly used to consistently estimate the average treatment effect (ATE) or the average treatment effect in the treated (ATT). The variance of the IPW ATE estimator is often estimated by assuming that the weights are known and then using the so-called "robust" (Huber-White) sandwich estimator, which results in conservative standard errors (SEs). Here we show that using such an approach when estimating the variance of the IPW ATT estimator does not necessarily result in conservative SE estimates. That is, assuming the weights are known, the robust sandwich estimator may be either conservative or anticonservative. Thus, confidence intervals for the ATT using the robust SE estimate will not be valid, in general. Instead, stacked estimating equations which account for the weight estimation can be used to compute a consistent, closed-form variance estimator for the IPW ATT estimator. The 2 variance estimators are compared via simulation studies and in a data analysis of the association between smoking and gene expression. [ABSTRACT FROM AUTHOR]

The inverse probability of treatment weighting (IPTW) approach is popular for evaluating causal effects in observational studies, but extreme propensity scores could bias the estimator and induce excessive variance. Recently, the overlap weighting approach has been proposed to alleviate this problem, which smoothly down-weights the subjects with extreme propensity scores. Although advantages of overlap weighting have been extensively demonstrated in literature with continuous and binary outcomes, research on its performance with time-to-event or survival outcomes is limited. In this article, we propose estimators that combine propensity score weighting and inverse probability of censoring weighting to estimate the counterfactual survival functions. These estimators are applicable to the general class of balancing weights, which includes IPTW, trimming, and overlap weighting as special cases. We conduct simulations to examine the empirical performance of these estimators with different propensity score weighting schemes in terms of bias, variance, and 95% confidence interval coverage, under various degrees of covariate overlap between treatment groups and censoring rates. We demonstrate that overlap weighting consistently outperforms IPTW and associated trimming methods in bias, variance, and coverage for time-to-event outcomes, and the advantages increase as the degree of covariate overlap between the treatment groups decreases. [ABSTRACT FROM AUTHOR]

Instrumental variable (IV) and risk adjustment (RA) estimators, including propensity score adjustments, are both used to alleviate confounding problems in nonexperimental studies on treatment effects, but it is not clear how estimates based on these 2 approaches compare. Methodological considerations have shown that IV and RA estimators yield estimates of distinct types of causal treatment effects regardless of confounding problems. Many investigators have neglected these distinctions. In this paper, the authors use 3 schematic models to explain visually the relations between IV and RA estimates of intended treatment effects as demonstrated in the methodological studies. When treatment effects are homogeneous across a study population or when treatment effects are heterogeneous across the study population but treatment decisions are unrelated to the treatment effects, RA and IV estimates should be equivalent when the respective assumptions are met. In contrast, when treatment effects are heterogeneous and treatment decisions are related to the treatment effects, RA estimates of treatment effect can asymptotically differ from IV estimates, but both are correct even when the respective assumptions are met. Appropriate interpretations of IV or RA estimates can be facilitated by developing conceptual models related to treatment choice and treatment effect heterogeneity prior to analyses. [ABSTRACT FROM AUTHOR]

Effect measure modification is often evaluated using parametric models. These models, although efficient when correctly specified, make strong parametric assumptions. While nonparametric models avoid important functional form assumptions, they often require larger samples to achieve a given accuracy. We conducted a simulation study to evaluate performance tradeoffs between correctly specified parametric and nonparametric models to detect effect modification of a binary exposure by both binary and continuous modifiers. We evaluated generalized linear models and doubly robust (DR) estimators, with and without sample splitting. Continuous modifiers were modeled with cubic splines, fractional polynomials, and nonparametric DR-learner. For binary modifiers, generalized linear models showed the greatest power to detect effect modification, ranging from 0.42 to 1.00 in the worst and best scenario, respectively. Augmented inverse probability weighting had the lowest power, with an increase of 23% when using sample splitting. For continuous modifiers, the DR-learner was comparable to flexible parametric models in capturing quadratic and nonlinear monotonic functions. However, for nonlinear, nonmonotonic functions, the DR-learner had lower integrated bias than splines and fractional polynomials, with values of 141.3, 251.7, and 209.0, respectively. Our findings suggest comparable performance between nonparametric and correctly specified parametric models in evaluating effect modification. [ABSTRACT FROM AUTHOR]

Ambitious World Health Organization targets for disease elimination require monitoring of epidemics using routine health data in settings of decreasing and low incidence. We evaluated 2 methods commonly applied to routine testing results to estimate incidence rates that assume a uniform probability of infection between consecutive negative and positive tests based on 1) the midpoint of this interval and 2) a randomly selected point in this interval. We compared these with an approximation of the Poisson binomial distribution, which assigns partial incidence to time periods based on the uniform probability of occurrence in these intervals. We assessed bias, variance, and convergence of estimates using simulations of Weibull-distributed failure times with systematically varied baseline incidence and varying trend. We considered results for quarterly, half-yearly, and yearly incidence estimation frequencies. We applied the methods to assess human immunodeficiency virus (HIV) incidence in HIV-negative patients from the Treatment With Antiretrovirals and Their Impact on Positive and Negative Men (TAIPAN) Study, an Australian study of HIV incidence in men who have sex with men, between 2012 and 2018. The Poisson binomial method had reduced bias and variance at low levels of incidence and for increased estimation frequency, with increased consistency of estimation. Application of methods to real-world assessment of HIV incidence found decreased variance in Poisson binomial model estimates, with observed incidence declining to levels where simulation results had indicated bias in midpoint and random-point methods. [ABSTRACT FROM AUTHOR]

To investigate temporal patterns, sociodemographic gradients, and structural breaks in adolescent marijuana use in the United States from 1991 to 2018, we used hierarchical age-period-cohort logistic regression models to distinguish temporal effects of marijuana use among 8th, 10th, and 12th graders from 28 waves of the Monitoring the Future survey (1991–2018). Structural breaks in period effects were further detected via a dynamic-programing–based method. Net of other effects, we found a clear age-related increase in the probability of marijuana use (10.46%, 23.17%, and 31.19% for 8th, 10th, and 12th graders, respectively). Period effects showed a substantial increase over time (from 16.23% in 2006 to 26.38% in 2018), while cohort effects remained stable throughout the study period. Risk of adolescent marijuana use varied by sex, racial group, family status, and parental education. Significant structural breaks during 1995–1996, 2006–2008, and 2011–2013 were identified in different subpopulations. A steady increase in marijuana use among adolescents during the latter years of this time period was identified. Adolescents who were male, were non-Black, lived in nonintact families, and had less educated parents were especially at risk of marijuana usage. Trends in adolescent marijuana use changed significantly during times of economic crisis. [ABSTRACT FROM AUTHOR]

Here we describe methods for assessing heterogeneity of treatment effects over prespecified subgroups in observational studies, using outcome-model–based (g-formula), inverse probability weighting, doubly robust, and matching estimators of subgroup-specific potential outcome means, conditional average treatment effects, and measures of heterogeneity of treatment effects. We compare the finite-sample performance of different estimators in simulation studies where we vary the total sample size, the relative frequency of each subgroup, the magnitude of treatment effect in each subgroup, and the distribution of baseline covariates, for both continuous and binary outcomes. We find that the estimators' bias and variance vary substantially in finite samples, even when there is no unobserved confounding and no model misspecification. As an illustration, we apply the methods to data from the Coronary Artery Surgery Study (August 1975–December 1996) to compare the effect of surgery plus medical therapy with that of medical therapy alone for chronic coronary artery disease in subgroups defined by previous myocardial infarction or left ventricular ejection fraction. [ABSTRACT FROM AUTHOR]

Multilevel regression and poststratification (MRP) is a model-based approach for estimating a population parameter of interest, generally from large-scale surveys. It has been shown to be effective in highly selected samples, which is particularly relevant to investigators of large-scale population health and epidemiologic surveys facing increasing difficulties in recruiting representative samples of participants. We aimed to further examine the accuracy and precision of MRP in a context where census data provided reasonable proxies for true population quantities of interest. We considered 2 outcomes from the baseline wave of the Ten to Men study (Australia, 2013–2014) and obtained relevant population data from the 2011 Australian Census. MRP was found to achieve generally superior performance relative to conventional survey weighting methods for the population as a whole and for population subsets of varying sizes. MRP resulted in less variability among estimates across population subsets relative to sample weighting, and there was some evidence of small gains in precision when using MRP, particularly for smaller population subsets. These findings offer further support for MRP as a promising analytical approach for addressing participation bias in the estimation of population descriptive quantities from large-scale health surveys and cohort studies. [ABSTRACT FROM AUTHOR]

The author offers a response to commentaries made on his article "The Harm Done to Reproducibility by the Culture of Null Hypothesis Significant Testing." Topics discussed include the author's main point regarding null hypothesis significant testing, influences that give the appearance of poor reproducibility, and alternative prescription to improve reproducibility.

The emergence and spread of Lyme disease and other infections associated with black-legged ticks is causing a public health crisis. No human vaccines are currently available, and both diagnosis and treatment are sometimes ineffectual, leading to advocacy for self-directed preventative measures. These recommendations are widely communicated to the public, but there is limited evidence for their efficacy. We undertook a systematic review and mixed-effects meta-regression analysis of factors purported to increase or decrease risk of black-legged tick bites and tick-borne disease. Published articles used in the study spanned the years 1984–2018. Variables associated with increased probability of tick-borne disease, with odds ratios significantly greater than 1, included deer abundance, high density of nymph-stage black-legged ticks, landscapes with interspersed herbaceous and forested habitat, low human population density, gardens, cat ownership, and race. Contrary to recommendations, use of landscape-related tick control measures, such as clearing brush, trimming branches, and having a dry barrier between lawn and woods, tended to increase risk. Pet ownership increased bite risk. Bite risk was highest for children aged 5 years or less, with a secondary peak in persons aged 50–70 years. Although some widely disseminated recommendations are supported by the research analyzed, others require further evaluation. Additional research is also needed to understand the mechanisms underlying significant relationships. [ABSTRACT FROM AUTHOR]

Matching methods are assumed to reduce the likelihood of a biased inference compared with ordinary least squares (OLS) regression. Using simulations, we compared inferences from propensity score matching, coarsened exact matching, and unmatched covariate-adjusted OLS regression to identify which methods, in which scenarios, produced unbiased inferences at the expected type I error rate of 5%. We simulated multiple data sets and systematically varied common support, discontinuities in the exposure and/or outcome, exposure prevalence, and analytical model misspecification. Matching inferences were often biased in comparison with OLS, particularly when common support was poor; when analysis models were correctly specified and common support was poor, the type I error rate was 1.6% for propensity score matching (statistically inefficient), 18.2% for coarsened exact matching (high), and 4.8% for OLS (expected). Our results suggest that when estimates from matching and OLS are similar (i.e. confidence intervals overlap), OLS inferences are unbiased more often than matching inferences; however, when estimates from matching and OLS are dissimilar (i.e. confidence intervals do not overlap), matching inferences are unbiased more often than OLS inferences. This empirical "rule of thumb" may help applied researchers identify situations in which OLS inferences may be unbiased as compared with matching inferences. [ABSTRACT FROM AUTHOR]

We evaluated the role of seasonality in self-reported diet quality among postmenopausal women participating in the Women's Health Initiative (WHI). A total of 156,911 women completed a food frequency questionnaire (FFQ) at enrollment (1993–1998). FFQ responses reflected intake over the prior 3-month period, and seasons were defined as spring (March–May), summer (June–August), fall (September–November), and winter (December–February). FFQ data were used to calculate the Alternate Healthy Eating Index (AHEI), a measure of diet quality that has a score range of 2.5–87.5, with higher scores representing better diet quality. In multivariable linear regression models using winter as the reference season, AHEI scores were higher in spring, summer, and fall (all P values < 0.05); although significant, the variance was minimal (mean AHEI score: winter, 41.7 (standard deviation, 11.3); summer, 42.2 (standard deviation, 11.3)). Applying these findings to hypothesis-driven association analysis of diet quality and its relationship with chronic disease risk (cardiovascular disease) showed that controlling for season had no effect on the estimated hazard ratios. Although significant differences in diet quality across seasons can be detected in this population of US postmenopausal women, these differences are not substantial enough to warrant consideration in association studies of diet quality. [ABSTRACT FROM AUTHOR]

The main aim of mediation analysis is to study the direct and indirect effects of an exposure on an outcome. To date, the literature on mediation analysis with multiple mediators has mainly focused on continuous and dichotomous outcomes. However, the development of methods for multiple mediation analysis of survival outcomes is still limited. Here we extend to survival outcomes a method for multiple mediation analysis based on the computation of appropriate weights. The approach considered has the advantages of not requiring specific models for mediators, allowing nonindependent mediators of any nature, and not relying on the assumption of rare outcomes. Simulation studies show good performance of the proposed estimator in terms of bias and coverage probability. The method is further applied to an example from a published study on prostate cancer mortality aimed at understanding the extent to which the effect of DNA methyltransferase 3b (DNMT3b) genotype on mortality was explained by DNA methylation and tumor aggressiveness. This approach can be used to quantify the marginal time-dependent direct and indirect effects carried by multiple indirect pathways, and software code is provided to facilitate its application. [ABSTRACT FROM AUTHOR]

The popular inverse probability weighting method in causal inference is often hampered by extreme propensity scores, resulting in biased estimates and excessive variance. A common remedy is to trim patients with extreme scores (i.e. remove them from the weighted analysis). However, such methods are often sensitive to the choice of cutoff points and discard a large proportion of the sample. The implications for bias and the precision of the treatment effect estimate are unclear. These problems are mitigated by a newly developed method, the overlap weighting method. Overlap weights emphasize the target population with the most overlap in observed characteristics between treatments, by continuously down-weighting the units in the tails of the propensity score distribution. Here we use simulations to compare overlap weights to standard inverse probability weighting with trimming, in terms of bias, variance, and 95% confidence interval coverage. A range of propensity score distributions are considered, including settings with substantial nonoverlap and extreme values. To facilitate practical implementation, we further provide a consistent estimator for the standard error of the treatment effect estimated using overlap weighting. [ABSTRACT FROM AUTHOR]

Time-series studies are routinely used to estimate associations between adverse health outcomes and short-term exposures to ambient air pollutants. Use of the Poisson log-linear model with the assumption of constant overdispersion is the most common approach, particularly when estimating associations between daily air pollution concentrations and aggregated counts of adverse health events throughout a geographical region. We examined how the assumption of constant overdispersion plays a role in estimation of air pollution effects by comparing estimates derived from the standard approach with those estimated from covariate-dependent Bayesian generalized Poisson and negative binomial models that accounted for potential time-varying overdispersion. Through simulation studies, we found that while there was negligible bias in effect estimates, the standard quasi-Poisson approach can result in a larger standard error when the constant overdispersion assumption is violated. This was also observed in a time-series study of daily emergency department visits for respiratory diseases and ozone concentration in Atlanta, Georgia (1999–2009). Allowing for covariate-dependent overdispersion resulted in a reduction in the ozone effect standard error, while the ozone-associated relative risk remained robust to different model specifications. Our findings suggest that improved characterization of overdispersion in time-series modeling can result in more precise health effect estimates in studies of short-term environmental exposures. [ABSTRACT FROM AUTHOR]