Check out our Interactive Tools
Body mass index (BMI) is used in clinical settings and research studies as a surrogate measure of adiposity. BMI is calculated as weight in kilograms divided by height in meters squared. BMI categories are commonly used as a classification system for adult men and women over age 20 years. In older adults, weight change, change in body composition, height loss, or, most likely, a combination of these factors can all influence BMI values and resulting inferences made using BMI. The use of standard BMI categories across the lifespan does not reflect our current understanding of changes in adiposity that occur with the aging process. The BMI-for-age percentile curves that we have developed for older adults are a descriptive tool to facilitate comparison among individuals in the population and track individual-level change over time relative to the population. With this interactive tool, you can create and modify BMI-for-age percentile curves using data from the U.S. National Health and Nutrition Examination Survey.
Quantitative bias analysis allows to estimate nonrandom errors in epidemiologic studies, assessing the magnitude and direction of biases, and quantifying their uncertainties. Every study has some random error due to its limited sample size, and is susceptible to systematic errors as well, from selection bias to the presence of (un)known confounders or information bias (measurement error, including misclassification). Bias analysis methods were compiled by Lash et al. in their book Applying Quantitative Bias Analysis to Epidemiologic Data. This Shiny app implements bias analyses from the book, as well as others (e.g. by S. Greenland), as computed by the R package episensr . More can be found in the episensr package available for download on R CRAN . The four tabs allow to perform (1) a simple analysis (for bias analysis requiring a 2-by-2 table as data input), (2) an analysis for covariate misclassification (requiring two 2-by-2 tables as data input), (3) a simple analysis with no observed data (for bias analysis that does not have as input an observed 2-by-2 table), and (4) a probabilistic bias analysis.
