Model-Based Recursive Partitioning for Subgroup Analyses

Seibold, Heidi; Zeileis, Achim; Hothorn, Torsten

doi:10.1515/ijb-2015-0032

Cited by 142 publications

(177 citation statements)

References 42 publications

Supporting

Mentioning

177

Contrasting

Order By: Relevance

“…Multiple regression analysis was used to model a numeric dependent outcome variable with respect to several independent predictor variables, and the results were displayed in analysis of variance (ANOVA) tables with type two F tests. 23,24 Ordinal response variables were analyzed by cumulative link models. UCD individuals belonging to group 1 have lower cSDS than individuals from group 2 (p < 0.001, t test).…”

Section: Discussionmentioning

confidence: 99%

“…23,24 Compared to the cytosolic (distal) UCDs (ie, ASS1-D, ASL-D, and ARG1-D, hereafter referred to as group 1), height of the initial NH 4 + max was associated with impaired neurocognitive outcome in individuals with mitochondrial . 23,24 Compared to the cytosolic (distal) UCDs (ie, ASS1-D, ASL-D, and ARG1-D, hereafter referred to as group 1), height of the initial NH 4 + max was associated with impaired neurocognitive outcome in individuals with mitochondrial .…”

Section: Impact Of Non-interventional Variables On Neurocognitive Funmentioning

confidence: 99%

See 1 more Smart Citation

Impact of Diagnosis and Therapy on Cognitive Function in Urea Cycle Disorders

Posset

Gropman

Nagamani

et al. 2019

Annals of Neurology

View full text Add to dashboard Cite

Objective: Individuals with urea cycle disorders (UCDs) often present with intellectual and developmental disabilities. The major aim of this study was to evaluate the impact of diagnostic and therapeutic interventions on cognitive outcomes in UCDs. Methods: This prospective, observational, multicenter study includes data from 503 individuals with UCDs who had comprehensive neurocognitive testing with a cumulative follow-up of 702 patient-years. Results: The mean cognitive standard deviation score (cSDS) was lower in symptomatic than in asymptomatic (p < 0.001, t test) individuals with UCDs. Intellectual disability (intellectual quotient < 70, cSDS < −2.0) was associated with the respective subtype of UCD and early disease onset, whereas height of the initial peak plasma ammonium concentration was inversely associated with neurocognitive outcomes in mitochondrial (proximal) rather than cytosolic (distal) UCDs. In ornithine transcarbamylase and argininosuccinate synthetase 1 deficiencies, we did not find evidence that monoscavenger therapy with sodium or glycerol phenylbutyrate was superior to sodium benzoate in providing cognitive protection. Early liver transplantation appears to be beneficial for UCDs. It is noteworthy that individuals with argininosuccinate synthetase 1 and argininosuccinate lyase deficiencies identified by newborn screening had better neurocognitive outcomes than those diagnosed after the manifestation of first symptoms. Interpretation: Cognitive function is related to interventional and non-interventional variables. Early detection by newborn screening and early liver transplantation appear to offer greater cognitive protection, but none of the currently used nitrogen scavengers was superior with regard to long-term neurocognitive outcome. Further confirmation could determine these variables as important clinical indicators of neuroprotection for individuals with UCDs.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Impact Of Non-interventional Variables On Neurocognitive Funmentioning

confidence: 99%

Impact of Diagnosis and Therapy on Cognitive Function in Urea Cycle Disorders

Posset

Gropman

Nagamani

et al. 2019

Annals of Neurology

View full text Add to dashboard Cite

show abstract

“…As such, they are preeminently suited to the detection of treatment-subgroup interactions. Several tree-based algorithms for the detection of treatment-subgroup interactions have already been developed (Dusseldorp, Doove, & Van Mechelen, 2016;Dusseldorp & Meulman, 2004;Su, Tsai, Wang, Nickerson, & Li, 2009;Foster, Taylor, & Ruberg, 2011;Lipkovich, Dmitrienko, Denne, & Enas, 2011;Zeileis, Hothorn, & Hornik, 2008;Seibold, Zeileis, & Hothorn, 2016;Athey & Imbens 2016). Also, Zhang, Tsiatis, Laber, and Davidian (2012b) and Zhang, Tsiatis, Davidian, Zhang, and Laber (2012a) have developed a flexible classification-based approach, allowing users to select from a range of statistical methods, including trees.…”

Section: Introductionmentioning

confidence: 99%

“…The GLMM tree algorithm builds on model-based recursive partitioning (MOB, Zeileis et al, 2008), which offers a flexible framework for subgroup detection. For example, GLM-based MOB has been applied to detect treatment-subgroup interactions for the treatment of depression (Driessen et al, 2016) and amyotrophic lateral sclerosis (Seibold et al, 2016). In contrast to other purely tree-based methods (e.g., Zeileis et al, 2008;Su et al, 2009;Dusseldorp et al, 2016), GLMM trees allow for taking into account the clustered structure of datasets.…”

Section: Introductionmentioning

confidence: 99%

Detecting treatment-subgroup interactions in clustered data with generalized linear mixed-effects model trees

Fokkema¹,

et al. 2017

View full text Add to dashboard Cite

Identification of subgroups of patients for whom treatment A is more effective than treatment B, and vice versa, is of key importance to the development of personalized medicine. Tree-based algorithms are helpful tools for the detection of such interactions, but none of the available algorithms allow for taking into account clustered or nested dataset structures, which are particularly common in psychological research. Therefore, we propose the generalized linear mixed-effects model tree (GLMM tree) algorithm, which allows for the detection of treatment-subgroup interactions, while accounting for the clustered structure of a dataset. The algorithm uses model-based recursive partitioning to detect treatment-subgroup interactions, and a GLMM to estimate the random-effects parameters. In a simulation study, GLMM trees show higher accuracy in recovering treatment-subgroup interactions, higher predictive accuracy, and lower type II error rates than linear-model-based recursive partitioning and mixed-effects regression trees. Also, GLMM trees show somewhat higher predictive accuracy than linear mixed-effects models with pre-specified interaction effects, on average. We illustrate the application of GLMM trees on an individual patient-level data metaanalysis on treatments for depression. We conclude that GLMM trees are a promising exploratory tool for the detection of treatment-subgroup interactions in clustered datasets.

show abstract

“…Generally, tree‐based methods conduct an exhaustive search over all possible cutoffs for each nonbinary factor, and thereby increase the chance of selecting a variable just by chance if it has a larger set of candidate cutoffs. In order to avoid this, the selection of covariates and the choice of the optimal cutoff need to be disentangled, as, for example, described by Loh and Loh et al An alternative idea to prevent variable selection bias is the model‐based recursive partitioning (MOB) approach . The essential idea is to test whether the parameters in a univariate model for treatment would be estimated differently conditional on the values of a further covariate.…”

Section: Introductionmentioning

confidence: 99%

Exploratory identification of predictive biomarkers in randomized trials with normal endpoints

Krzykalla

Benner

Kopp‐Schneider

2019

Statistics in Medicine

View full text Add to dashboard Cite

One of the main endeavours in present‐day medicine, especially in oncological research, is to provide evidence for individual treatment decisions (“stratified medicine”). In the pursuit of optimal treatment decision rules, the identification of predictive biomarkers that modify the treatment effect is essential. Proposed methods have often been based on recursive partitioning since a wide variety of interaction patterns can be captured automatically and the results are easily interpretable. Furthermore, these methods are readily extendable to high‐dimensional settings by means of ensemble learning. In this article, we present predMOB, an adaptation of the model‐based recursive partitioning (MOB) for subgroup analysis approach specifically tailored to the identification of predictive factors. In a simulation study, predMOB outperforms the original MOB with respect to the number of false detections and shows to be more robust in moderately complex settings. Furthermore, we compare the results of predMOB for the application to a public data base of amyotrophic lateral sclerosis patients to those obtained from the original MOB and are able to elucidate the nature of the biomarkers' effects.

show abstract

Model-Based Recursive Partitioning for Subgroup Analyses

Cited by 142 publications

References 42 publications

Impact of Diagnosis and Therapy on Cognitive Function in Urea Cycle Disorders

Impact of Diagnosis and Therapy on Cognitive Function in Urea Cycle Disorders

Detecting treatment-subgroup interactions in clustered data with generalized linear mixed-effects model trees

Exploratory identification of predictive biomarkers in randomized trials with normal endpoints

Contact Info

Product

Resources

About