Aldin, Angela ORCID: 0000-0001-6186-9621 (2023). Evidence-based oncology: the use of methodologically complex systematic reviews to inform cancer research and clinical practice. PhD thesis, Universität zu Köln.
PDF
AAldin_Dissertation_merged_26062023.pdf Download (19MB) |
Abstract
Background: Systematic reviews are produced to inform health research and clinical practice, e.g., by identifying research gaps and by formulating recommendations in clinical practice guidelines. Standardised methodology exists for the conduct of systematic reviews of interventions. To answer clinically diverse research questions, new methods are constantly being developed for the systematic synthesis of results from different types of studies. Moreover, constant monitoring of newly available evidence, particularly in clinical areas that are rapidly evolving, is important to ensure the currency of systematic reviews. Objective: The primary objective of this cumulative dissertation was to conduct systematic reviews using new and complex systematic review methods, and to contribute to the further development and refinement of these methods. Secondary objective was to conduct clinically relevant systematic reviews to provide meaningful evidence that can inform clinical practice and health care in oncology. Methods: Two clinically relevant systematic reviews using novel and complex methodological approaches were conducted: Systematic review I: A systematic review with network meta analysis and an adapted living approach to evaluate and compare the benefits and harms of first-line therapies for adults with advanced renal cell carcinoma. Systematic review II: A systematic review with meta-analysis of prognostic factor studies to explore the interim positron emission tomography (PET) scan result as a prognostic factor in adults with newly diagnosed Hodgkin lymphoma. Results: Methodological results Systematic review I: The evidence for the currently recommended treatments and important comparisons in this review stem from direct evidence from one trial per comparison only. This is due to the great lack of head-to-head comparisons of the many treatment options available. Statistical validation of the homogeneity and consistency assumptions was not possible for every network meta-analysis, so the validity of estimates is largely based on the transitivity assumption. When a strong evidence base is missing, the results of a network meta-analysis, including the ranking of treatments, should be interpreted with caution. The adapted living approach, where monthly update searches were conducted during the conduct of the review, was an appropriate method to maintain the currency of the evidence in such a rapidly evolving treatment landscape. Systematic review II: The greatest methodological challenges identified in synthesising evidence from prognostic factor studies were that, firstly, searching for prognosis studies is challenging due to insufficient indexing and missing search filters that are specific and sensitive enough to identify prognostic factor studies. Secondly, extracting and analysing outcome results was particularly difficult due to incomplete reporting of important data in the, usually retrospective, studies. Thirdly, available methods for the quality assessments had to be adapted to fit to the review question. Lastly, methods for the certainty assessment of the evidence from prognosis studies had to be developed during the conduct of the review as there was no official guidance at that time. The challenges encountered during the conduct of both reviews were discussed and resolved through the involvement of methodological and clinical experts as coauthors. Clinical results: Systematic Review I: Combinations of novel therapies (e.g., a checkpoint inhibitor with a tyrosine kinase inhibitor) appear to be superior to monotherapy with sunitinib (a tyrosine kinase inhibitor) as first-line therapy in terms of survival for adults with advanced renal cell carcinoma. However, these novel treatments may cause more (serious) side effects. Moreover, the question on the potential impact of these novel treatments on the quality of life of affected individuals remains unanswered. Systematic Review II: Evidence was found on the prognostic ability of the interim PET-scan result to predict survival in adults with Hodgkin lymphoma. It successfully distinguishes between PET-negative people, who have a better outcome prognosis, and PET-positive people, who have a worse outcome prognosis. Conclusion:Future methodological research needs to further address these different challenges, for example the challenges one encounters when trying to search for and identify prognostic factor studies, or the limitations one encounters when underlying assumptions of a network meta analysis cannot be verified. When evidence from such methodologically complex systematic reviews shall be used to inform clinical practice guidelines and, thereby, health care decision making, all involved stakeholders need to be aware of the methodological complexity and limitations behind the evidence produced.
Item Type: | Thesis (PhD thesis) | ||||||||||||
Translated title: |
|
||||||||||||
Translated abstract: |
|
||||||||||||
Creators: |
|
||||||||||||
URN: | urn:nbn:de:hbz:38-703618 | ||||||||||||
Date: | 2023 | ||||||||||||
Language: | English | ||||||||||||
Faculty: | Faculty of Medicine | ||||||||||||
Divisions: | Faculty of Medicine > Innere Medizin > Klinik I für Innere Medizin - Hämatologie und Onkologie | ||||||||||||
Subjects: | Generalities, Science Medical sciences Medicine |
||||||||||||
Uncontrolled Keywords: |
|
||||||||||||
Date of oral exam: | 22 June 2023 | ||||||||||||
Referee: |
|
||||||||||||
Refereed: | Yes | ||||||||||||
URI: | http://kups.ub.uni-koeln.de/id/eprint/70361 |
Downloads
Downloads per month over past year
Export
Actions (login required)
View Item |