The presented thesis describes the historical evolution of growth and laboratory reference values and the methods for their creation – leading finally to a family of methods applied in the WHO Multicentre Growth Reference Study (MGRS). The following part describes these methods, their assumptions, and model diagnostics. The original article at the beginning of part III combines these methods with resampling to be able to use LMS -type methods on data containing different dependencies like follow-up measures and family relationships. This method has been applied in the estimation of reference values of several laboratory values in the context of the LIFE child study. Three papers are already published and are also presented.
In part III. The next section concentrates on the accompanying
R package childsds. A short summary and outlook conclude the work.:I Introduction
1 Introduction
1.1 Standardization and Reference Values
1.2 The Scope of this Thesis
2 Historical Aspects
2.1 Development of Basic Concepts
2.2 Growth References and Growth Charts
2.3 Laboratory Reference Values
II Methods
3 Method Selection
3.1 Nomenclature
3.2 The Choice of Statistical Method
3.3 GAMLSS
3.4 LMS-type Methods
4 Model Diagnostics
4.1 Normalized Quantile Residuals
4.2 Diagnostic Plots
4.3 Numerical Statistics
III Original Articles & R Pakackage
5 Original Article
6 Articles Using the Proposed Method
7 R Package childsds
7.1 The Example Data
7.2 The Fitting Functions
7.3 The Classes
7.4 The Back-transformation Function
7.5 Convenience Functions
7.6 Collection of Reference Values
8 Summary and Outlook
8.1 Summary
8.2 SDS Values and Quality Control
8.3 The R Package / Die vorliegende Arbeit soll die Entwicklung und Bedeutung von Referenzwerten im pädiatrischen Kontext beschreiben. Der zugehörige Artikel beschreibt die Kombination der von der WHO empfohlene Methode und Resampling, um die Herleitung von Referenzwerten auch im Fall von Messwiederholungen und anderen Abhängigkeiten zu ermöglichen. Drei weitere Artikel, in denen die Methode angewendet wurde und eine kurze Vorstellung des zugehörigen, begleitend entwickelten R Pakets (statistische Software) bilden Teil III der Arbeit.:I Introduction
1 Introduction
1.1 Standardization and Reference Values
1.2 The Scope of this Thesis
2 Historical Aspects
2.1 Development of Basic Concepts
2.2 Growth References and Growth Charts
2.3 Laboratory Reference Values
II Methods
3 Method Selection
3.1 Nomenclature
3.2 The Choice of Statistical Method
3.3 GAMLSS
3.4 LMS-type Methods
4 Model Diagnostics
4.1 Normalized Quantile Residuals
4.2 Diagnostic Plots
4.3 Numerical Statistics
III Original Articles & R Pakackage
5 Original Article
6 Articles Using the Proposed Method
7 R Package childsds
7.1 The Example Data
7.2 The Fitting Functions
7.3 The Classes
7.4 The Back-transformation Function
7.5 Convenience Functions
7.6 Collection of Reference Values
8 Summary and Outlook
8.1 Summary
8.2 SDS Values and Quality Control
8.3 The R Package
| Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:21168 |
| Date | 27 April 2018 |
| Creators | Vogel, Mandy |
| Contributors | Universität Leipzig |
| Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
| Language | English |
| Detected Language | English |
| Type | info:eu-repo/semantics/acceptedVersion, doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.002 seconds