Evidence synthesis refers to methods used to consolidate existing scientific findings into a clear, comprehensible overall conclusion. In pharmaceutical and medical device development, evidence synthesis is employed to assess efficacy, safety, and benefit-risk aspects based on multiple sources rather than relying on a single study. The term encompasses both quantitative methods (e.g., meta-analysis) and qualitative or structured approaches (e.g., narrative synthesis) and forms the foundation of many regulatory and HTA-relevant dossiers.
How is evidence synthesis used in practice?
In clinical development, evidence synthesis supports study planning, interpretation of conflicting results, and derivation of evidence-based decisions. Typical use cases include establishing clinically relevant endpoints, positioning effect size relative to the standard of care, and justifying why certain populations or comparator arms are appropriate. Even in rare diseases (orphan drugs) or small patient populations, systematic consolidation of external evidence can help transparently identify knowledge gaps and justify supplementary data sources.
In Europe, evidence synthesis is also closely linked to requirements from Health Technology Assessment (HTA) and benefit assessment. For marketing authorization and reimbursement decisions, sponsors and manufacturers must frequently demonstrate how the totality of evidence supports the claimed benefit, what uncertainties remain, and how these should be addressed (e.g., through additional studies or post-authorization measures). In early access settings, a robust synthesis can help openly describe the plausibility and limitations of external data.
Main forms: quantitative and qualitative synthesis
Quantitative evidence synthesis encompasses methods such as meta-analyses, in which results from multiple studies are statistically combined. The objective is a more precise estimate of effect size (e.g., hazard ratio or risk difference) and exploration of heterogeneity between studies. In practice, studies must be sufficiently comparable, and data preparation follows predefined rules to minimize bias.
Qualitative or structured synthesis is used when statistical combination is not appropriate, for instance due to substantial differences in design, populations, or endpoints. Here, results are consolidated along a transparent framework, including assessment of study quality, consistency of findings, and plausibility. This form should also be documented in a comprehensible manner so that third parties can verify the conclusions.
Methodological components: systematic search, selection, and assessment
A robust evidence synthesis begins with a clear research question (e.g., according to PICO: Population, Intervention, Comparator, Outcome) and a systematic literature and database search. This is followed by predefined inclusion and exclusion criteria, a structured selection process, and extraction of relevant data points. Essential is the assessment of internal and external validity of included studies, for example through risk-of-bias considerations, sensitivity analyses, and discussion of transferability to the target population.
In regulated environments, it is also important that decisions in the process are justified and documented with version control. This applies, for example, to changes in search strategies, handling of multiple publications from the same study, or management of missing data. Clean documentation supports audits and review processes and reduces the risk that the synthesis is assessed as selective or incomplete.
Regulatory and HTA perspective (DE/EU focus)
Authorities and assessment bodies expect the totality of evidence to be presented in a comprehensible manner, particularly when external data are used to justify assumptions. In clinical trials, Good Clinical Practice principles play a role, for instance regarding data integrity and transparent reporting. In the EU, it is also relevant that clinical data and their analysis are prepared consistently, verifiably, and appropriately for the assessment of benefits and risks under EU Regulation 536/2014 (Clinical Trials Regulation).
For medical devices, evidence synthesis can be part of the clinical evaluation when clinical data from literature and experience must be systematically consolidated. A structured process is necessary that justifies the relevance, quality, and validity of the sources used and discloses remaining uncertainties. In practice, it is particularly important that statements on equivalence or clinical performance are based on a comprehensible evidence base and not solely on individual publications.
Common errors and how to avoid them
A common error is an overly broad or unclear research question, making the search unwieldy and the selection susceptible to subjective decisions. Equally problematic is the mixing of non-comparable endpoints or populations, which can result in an apparently precise but substantively misleading overall conclusion. Selective inclusion of positive studies, lack of transparency in exclusions, or inadequate assessment of study quality can also undermine the credibility of the evidence synthesis.
In practice, sponsors, CROs, and medical writing should define roles, responsibilities, and quality checks early: Who formulates the research question? Who reviews search strategy and data extraction? What criteria apply for sensitivity analyses? Clear governance reduces subsequent rework in dossier reviews and improves consistency between clinical reports, publications, and regulatory submissions.
FAQ
Is evidence synthesis the same as a meta-analysis?
No. A meta-analysis is a quantitative method within evidence synthesis. Evidence synthesis additionally encompasses structured qualitative approaches, quality assessments, and transparent positioning of the totality of evidence.
When is quantitative synthesis not appropriate?
When studies are highly heterogeneous (e.g., different endpoints, populations, or designs) or data are missing, purely statistical combination can be misleading. In such cases, a structured qualitative synthesis is often more appropriate.
What role does evidence synthesis play in regulatory submissions?
It supports the benefit-risk argumentation, positioning within the therapeutic context, and justification of assumptions or data gaps. A comprehensible process with clear documentation is essential.
Regulatory References (Selection)
- EU Regulation (EU) No 536/2014 (Clinical Trials Regulation, CTR)
- Regulation (EU) 2017/745 on medical devices (MDR)
- ICH E6(R3): Guideline for Good Clinical Practice