Abstract
dc:descriptionThe current thesis aims to address two common challenges encountered in network meta-analysis (NMA), the evaluation of consistency and the handling of multi-component interventions. Consistency assessment is vital as the validity of the NMA findings is primary affected by the satisfaction of this assumption. Significant discrepancies between direct and indirect evidence may lead to biased NMA estimates. In particular, a brief overview of (N)MA and the standard method used to evaluate consistency is presented. Additionally, a novel method for identifying inconsistencies is proposed that evaluates network consistency both globally and locally. This was accomplished by integrating the Stochastic Search Variable Selection method into the NMA framework and treating inconsistency factors as variables in a generalized linear model. Historical evidence and differences between direct and indirect evidence that are considered of practical significant, can be also incorporated into the inconsistency detection process. The performance of the proposed method was evaluated in two published networks under various scenarios. Furthermore, to facilitate the implementation of the method, the user-friendly R-package ssifs was developed and published on CRAN. Concerning the handling of multi-component interventions, novel ways of visualizing NMA results are proposed that allow for the easy identification of the most efficacious components, exploration of the network’s geometry and examination of the components’ behaviour. To implement these tools, the user-friendly R-package viscomp was developed and published on CRAN. Lastly, some concluding remarks and suggestions for future research are provided.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Seitidis, Georgios
- Σεϊτίδης, Γεώργιος
Subjects
dc:subject × 27- Μετα-ανάλυση
- Μετα-ανάλυση δικτύου
- Μεταβατικότητα
- Συνέπεια
- Επιλογή μεταβλητών
- Παρεμβάσεις πολλών συνιστωσών
- Οπτικοποίηση
- Meta-analysis
- Network meta-analysis
- Transitivity
- Consistency
- Variable selection
- SSVS
- Multicomponent
- Visualization
- Μαθηματικά
- Φυσικές Επιστήμες
- Στατιστική και Πιθανότητες
- Επιστήμες Υγείας
- Ιατρική και Επιστήμες Υγείας
- Επιδημιολογία
- Mathematics
- Natural Sciences
- Statistics and Probability
- Health Sciences
- Medical and Health Sciences
- Epidemiology
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/54280