Weighted mixed-effects dose–response models for tables of correlated contrasts

• Autores: Nicola Orsini
• Localización: The Stata journal, ISSN 1536-867X, Vol. 21, Nº. 2, 2021, págs. 320-347
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Recognizing a dose–response pattern based on heterogeneous tables of contrasts is hard. Specification of a statistical model that can consider the possible dose–response data-generating mechanism, including its variation across studies, is crucial for statistical inference. The aim of this article is to increase the un- derstanding of mixed-effects dose–response models suitable for tables of correlated estimates. One can use the command drmeta with additive (mean difference) and multiplicative (odds ratios, hazard ratios) measures of association. The postes- timation command drmeta graph greatly facilitates the visualization of predicted average and study-specific dose–response relationships. I illustrate applications of the drmeta command with regression splines in experimental and observational data based on nonlinear and random-effects data-generation mechanisms that can be encountered in health-related sciences