Importance of Shrinkage in Empirical Bayes Estimates for Diagnostics: Problems and Solutions

Journal Title: The AAPS Journal - Year 2009, Vol 11, Issue 3

Abstract

Empirical Bayes (“post hoc”) estimates (EBEs) of ηs provide modelers with diagnostics: the EBEs themselves, individual prediction (IPRED), and residual errors (individual weighted residual (IWRES)). When data are uninformative at the individual level, the EBE distribution will shrink towards zero (η-shrinkage, quantified as 1-SD(ηEBE)/ω), IPREDs towards the corresponding observations, and IWRES towards zero (ε-shrinkage, quantified as 1-SD(IWRES)). These diagnostics are widely used in pharmacokinetic (PK) pharmacodynamic (PD) modeling; we investigate here their usefulness in the presence of shrinkage. Datasets were simulated from a range of PK PD models, EBEs estimated in non-linear mixed effects modeling based on the true or a misspecified model, and desired diagnostics evaluated both qualitatively and quantitatively. Identified consequences of η-shrinkage on EBE-based model diagnostics include non-normal and/or asymmetric distribution of EBEs with their mean values (“ETABAR”) significantly different from zero, even for a correctly specified model; EBE–EBE correlations and covariate relationships may be masked, falsely induced, or the shape of the true relationship distorted. Consequences of ε-shrinkage included low power of IPRED and IWRES to diagnose structural and residual error model misspecification, respectively. EBE-based diagnostics should be interpreted with caution whenever substantial η- or ε-shrinkage exists (usually greater than 20% to 30%). Reporting the magnitude of η- and ε-shrinkage will facilitate the informed use and interpretation of EBE-based diagnostics.

Authors and Affiliations

Radojka M. Savic, Mats O. Karlsson

Keywords

Related Articles

Quantification of Swelling and Erosion in the Controlled Release of a Poorly Water-Soluble Drug Using Synchrotron X-ray Computed Microtomography

The hydration layer plays a key role in the controlled drug release of gel-forming matrix tablets. For poorly water-soluble drugs, matrix erosion is considered as the rate limiting step for drug release. However, few inv...

Study of crystallization of endogenous surfactant in Eudragit NE30D-Free films and its influence on drug-release properties of controlled-release diphenhydramine HCI pellets coated with Eudragit NE30D

This study investigates the crystallization of the endogenous surfactant nonoxynol 100 in Eudragit NE30D-free films during storage and the influences of nonoxynol 100 on the dissolution of diphenhydramine hydrochloric ac...

Characterising Lipid Lipolysis and Its Implication in Lipid-Based Formulation Development

Facing the increasing number of poorly water-soluble drugs, pharmaceutical scientists are required to break new grounds for the delivery of these pharmaceutically problematic drugs. Lipid-based drug delivery systems (LBD...

Completeness Assessment of Type II Active Pharmaceutical Ingredient Drug Master Files under Generic Drug User Fee Amendment: Review Metrics and Common Incomplete Items

Under the Generic Drug User Fee Amendments (GDUFA) of 2012, Type II active pharmaceutical ingredient (API) drug master files (DMFs) must pay a user fee and pass a Completeness Assessment (CA) before they can be reference...

Lipid-Based Drug Carriers for Prodrugs to Enhance Drug Delivery

The combination of lipid drug delivery systems with prodrugs offers several advantages including improved pharmacokinetics, increased absorption, and facilitated targeting. Lipidization and use of lipid carriers can incr...

Download PDF file
  • EP ID EP681490
  • DOI  10.1208/s12248-009-9133-0
  • Views 130
  • Downloads 0

How To Cite

Radojka M. Savic, Mats O. Karlsson (2009). Importance of Shrinkage in Empirical Bayes Estimates for Diagnostics: Problems and Solutions. The AAPS Journal, 11(3), -. https://europub.co.uk/articles/-A-681490