University of Thessaly (UTH)
Δίκτυα θεραπειών από τυχαιοποιημένες κλινικές μελέτες: μεθοδολογική προσέγγιση
Abstract
dc:descriptionObjective: The aim of this thesis was to explore methodological aspects related to network meta-analysis, especially in oncology field. For the purposes of this analysis, Small Cell Lung Cancer (SCLC) was used as an example, where the combination of Cisplatin plus Etoposide (EP) is currently the standard treatment.Methods PubMed, EMBASE and Cochrane Central Register of Controlled Trials weresystematically searched to identify all RCTs that compared treatments for SCLC. Then,effectiveness of the treatments relative to the combination of Cisplatin plus Etoposide,reference treatment) was estimated by performing a network of treatments analysis, using both Bayesian and frequentist approaches.Results We identified 71 articles eligible for inclusion, involving 91 different treatments. In total, 16,026 patients were included in the analysis. Frequentist analysis (direct) revealed combination of Cisplatin plus Cyclophosphamide plus Etoposide plus Epirubicin showed better response than EP for the ORR outcome, but with worse tolerability. Indirect analysis revealed that the combination of Cisplatin plus Doxorubicin plus Etoposide (plus Vincrisitine) showed better response that EP for the ORR outcome. Bayesian analysis revealed that the combination of carboplatin or cisplatin plus etoposide with granulocytecolony stimulating factor (GCSF) provides higher probability of achieving ORR compared to other treatments.Conclusions The results should be interpreted with caution because the network wasdominated by indirect comparisons. Large scale head-to-head RCTs are needed to confirm the present findings. Bayesian and frequentist approaches should be considered complementary tools in the clinical evaluator's toolkit.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Bakalos, Georgios
- Μπάκαλος, Γεώργιος
Subjects
dc:subject × 6Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/39901