Maximum likelihood estimation under the Emax model: existence, geometry and efficiency
Open Access
- 10 June 2025
- journal article
- research article
- Published by Springer Nature in Statistical Papers
- Vol. 66 (5) , 1-28
Abstract
This study focuses on the estimation of the Emax dose–response model, a widely utilized framework in clinical trials, experiments in pharmacology, agriculture, environmental science, and more. Existing challenges in obtaining maximum likelihood estimates (MLE) for model parameters are often ascribed to computational issues but, in reality, stem from the absence of a MLE. Our contribution provides new understanding and control of all the experimental situations that practitioners might face, guiding them in the estimation process. We derive the exact MLE for a three-point experimental design and identify the two scenarios where the MLE fails to exist. To address these challenges, we propose utilizing Firth’s modified score, which we express analytically as a function of the experimental design. Through a simulation study, we demonstrate that the Firth modification yields a finite estimate in one of the problematic scenarios. For the remaining case, we introduce a design-augmentation strategy akin to a hypothesis test.Keywords
Funding Information
- PRIN2022 (2022TRB44L)
- Engineering and Physical Sciences Research Council (EP/T021624/1)
- Università degli Studi del Piemonte Orientale Amedeo Avogrado
This publication has 28 references indexed in Scilit:
- Longitudinal model for a dose-finding study for a rare disease treatmentStatistical Papers, 2023
- The effects of adaptation on maximum likelihood inference for nonlinear models with normal errorsJournal of Statistical Planning and Inference, 2021
- Jeffreys-prior penalty, finiteness and shrinkage in binomial-response generalized linear modelsBiometrika, 2020
- Mean and median bias reduction in generalized linear modelsStatistics and Computing, 2019
- On the efficiency of two‐stage response‐adaptive designsStatistics in Medicine, 2012
- Comparison between cerebral state index and bispectral index as measures of electroencephalographic effects of sevoflurane using combined sigmoidal Emax modelActa Pharmacologica Sinica, 2011
- Practical considerations for optimal designs in clinical dose finding studiesStatistics in Medicine, 2010
- An Adaptive Optimal Design for theEmaxModel and Its Application in Clinical TrialsJournal of Biopharmaceutical Statistics, 2009
- Analysis of Dose–Response Studies—Emax Model2006
- Bias reduction of maximum likelihood estimatesBiometrika, 1993