Increasingly, the statistical and epidemiologic literature is focusing beyond issues of internal validity and turning its attention to questions of external validity. Here, we discuss some of the ...challenges of transporting a causal effect from a randomized trial to a specific target population. We present an inverse odds weighting approach that can easily operationalize transportability. We derive these weights in closed form and illustrate their use with a simple numerical example. We discuss how the conditions required for the identification of internally valid causal effects are translated to apply to the identification of externally valid causal effects. Estimating effects in target populations is an important goal, especially for policy or clinical decisions. Researchers and policy-makers should therefore consider use of statistical techniques such as inverse odds of sampling weights, which under careful assumptions can transport effect estimates from study samples to target populations.
A Framework for Descriptive Epidemiology Lesko, Catherine R; Fox, Matthew P; Edwards, Jessie K
American journal of epidemiology,
11/2022, Letnik:
191, Številka:
12
Journal Article
Recenzirano
Odprti dostop
Abstract
In this paper, we propose a framework for thinking through the design and conduct of descriptive epidemiologic studies. A well-defined descriptive question aims to quantify and characterize ...some feature of the health of a population and must clearly state: 1) the target population, characterized by person and place, and anchored in time; 2) the outcome, event, or health state or characteristic; and 3) the measure of occurrence that will be used to summarize the outcome (e.g., incidence, prevalence, average time to event, etc.). Additionally, 4) any auxiliary variables will be prespecified and their roles as stratification factors (to characterize the outcome distribution) or nuisance variables (to be standardized over) will be stated. We illustrate application of this framework to describe the prevalence of viral suppression on December 31, 2019, among people living with human immunodeficiency virus (HIV) who had been linked to HIV care in the United States. Application of this framework highlights biases that may arise from missing data, especially 1) differences between the target population and the analytical sample; 2) measurement error; 3) competing events, late entries, loss to follow-up, and inappropriate interpretation of the chosen measure of outcome occurrence; and 4) inappropriate adjustment.
Healthy worker survivor bias may occur in occupational studies due to the tendency for unhealthy individuals to leave work earlier, and consequently accrue less exposure, compared with their ...healthier counterparts. If occupational data are not analyzed using appropriate methods, this bias can result in attenuation or even reversal of the estimated effects of exposures on health outcomes. Recent advances in computing power, coupled with state-of-the-art statistical methods, have greatly increased the ability of analysts to control healthy worker survivor bias. However, these methods have not been widely adopted by occupational epidemiologists. We update the seminal review by Arrighi and Hertz-Picciotto (Epidemiology.1994; 5: 186–196) of the sources and methods to control healthy worker survivor bias. In our update, we discuss methodologic advances since the publication of that review, notably with a consideration of how directed acyclic graphs can inform the choice of appropriate analytic methods. We summarize and discuss methods for addressing this bias, including recent work applying g-methods to account for employment status as a time-varying covariate affected by prior exposure. In the presence of healthy worker survivor bias, g-methods have advantages for estimating less biased parameters that have direct policy implications and are clearly communicated to decision-makers.
Abstract
The coronavirus disease 2019 (COVID-19) pandemic and the coming transition to a postpandemic world where COVID-19 will likely remain as an endemic disease present a host of challenges and ...opportunities in epidemiologic research. The scale and universality of this disruption to life and health provide unique opportunities to study phenomena and health challenges in all branches of epidemiology, from the obvious infectious disease and social consequences to less clear impacts on chronic disease and cancer. If we are to both take advantage of the largest natural experiment of our lifetimes and provide evidence to inform the numerous public health and clinical decisions being made every day, we must act quickly to ask critical questions and develop new methods for answering them. In doing so, we should build on each of our strengths and expertise and try to provide new insights rather than become yet another voice commenting on the same set of questions with limited evidence.
Surprise Cole, Stephen R; Edwards, Jessie K; Greenland, Sander
American journal of epidemiology,
02/2021, Letnik:
190, Številka:
2
Journal Article
Recenzirano
Odprti dostop
Abstract
Measures of information and surprise, such as the Shannon information value (S value), quantify the signal present in a stream of noisy data. We illustrate the use of such information ...measures in the context of interpreting P values as compatibility indices. S values help communicate the limited information supplied by conventional statistics and cast a critical light on cutoffs used to judge and construct those statistics. Misinterpretations of statistics may be reduced by interpreting P values and interval estimates using compatibility concepts and S values instead of “significance” and “confidence.”
Abstract
In recent years, increasing attention has been paid to problems of external validity, specifically to methodological approaches for both quantitative generalizability and transportability of ...study results. However, most approaches to these issues have considered external validity separately from internal validity. Here we argue that considering either internal or external validity in isolation may be problematic. Further, we argue that a joint measure of the validity of an effect estimate with respect to a specific population of interest may be more useful: We call this proposed measure target validity. In this work, we introduce and formally define target bias as the total difference between the true causal effect in the target population and the estimated causal effect in the study sample, and target validity as target bias = 0. We illustrate this measure with a series of examples and show how this measure may help us to think more clearly about comparisons between experimental and nonexperimental research results. Specifically, we show that even perfect internal validity does not ensure that a causal effect will be unbiased in a specific target population.
Great care is taken in epidemiologic studies to ensure the internal validity of causal effect estimates; however, external validity has received considerably less attention. When the study sample is ...not a random sample of the target population, the sample average treatment effect, even if internally valid, cannot usually be expected to equal the average treatment effect in the target population. The utility of an effect estimate for planning purposes and decision making will depend on the degree of departure from the true causal effect in the target population due to problems with both internal and external validity. Herein, we review concepts from recent literature on generalizability, one facet of external validity, using the potential outcomes framework. Identification conditions sufficient for external validity closely parallel identification conditions for internal validity, namely conditional exchangeability; positivity; the same distributions of the versions of treatment; no interference; and no measurement error. We also require correct model specification. Under these conditions, we discuss how a version of direct standardization (the g-formula, adjustment formula, or transport formula) or inverse probability weighting can be used to generalize a causal effect from a study sample to a well-defined target population, and demonstrate their application in an illustrative example.
Background Patients with end-stage renal disease (ESRD) receiving dialysis have been reported to have increased risk of cancer. However, contemporary cancer burden estimates in this population are ...sparse and do not account for the high competing risk of death characteristic of dialysis patients. Study Design Retrospective cohort study. Setting & Participants US adult patients enrolled in Medicare’s ESRD program who received in-center hemodialysis. Factors Demographic/clinical characteristics. Outcomes For overall and site-specific cancers identified using claims-based definitions, we calculated annual incidence rates (1996-2009). We estimated 5-year cumulative incidence since dialysis therapy initiation using competing-risk methods. Results We observed a constant rate of incident cancers for all sites combined, from 3,923 to 3,860 cases per 100,000 person-years (annual percentage change, 0.1; 95% CI, −0.4 to 0.6). Rates for some common site-specific cancers increased (ie, kidney/renal pelvis) and decreased (ie, colon/rectum, lung/bronchus, pancreas, and other sites). Of 482,510 incident hemodialysis patients, cancer was diagnosed in 37,128 within 5 years after dialysis therapy initiation. The 5-year cumulative incidence of any cancer was 9.48% (95% CI, 9.39%-9.57%) and was higher for certain subgroups: older age, males, nonwhites, non-Hispanics, nondiabetes primary ESRD cause, recent dialysis therapy initiation, and history of transplantation evaluation. Among blacks and whites, we observed 35,767 cases compared with 25,194 expected cases if the study population had experienced rates observed in the US general population (standardized incidence ratio SIR, 1.42; 95% CI, 1.41-1.43). Risk was most elevated for cancers of the kidney/renal pelvis (SIR, 4.03; 95% CI, 3.88-4.19) and bladder (SIR, 1.57; 95% CI, 1.51-1.64). Limitations Claims-based cancer definitions have not been validated in the ESRD population. Information for cancer risk factors was not available in our data source. Conclusions These results suggest a high burden of cancer in the dialysis population compared to the US general population, with varying patterns of cancer incidence in subgroups.