A systematic literature review aims to comprehensively identify, select, and analyze the available scientific data on a specific topic, using a predefined and reproducible method. It helps document the state of the art, support study design, and contribute to the assessment of the safety, performance, and clinical benefit of a medical device. At Statitec, a biometrics company within the MultiHealth Group, we conduct these reviews, ranging from literature searches to meta-analyses.
What is a systematic literature review?
The difference from a narrative review can be summed up in one word: the method. A narrative review summarizes the publications that an author deems relevant. A systematic review applies search and selection criteria established in advance, documents each step, and assesses the quality of each included study. A predefined, documented, and transparent methodology should enable another team to replicate the process and arrive at a comparable selection of data.
It is this reproducibility that gives it its regulatory value. Authorities and notified bodies expect a traceable process, not a selection of favorable findings.
| Journal Type | Objective | Level of Evidence |
| Narrative Review | Provide a general overview of a topic | Low, risk of selection |
| Scoping review | Mapping the Available Literature | Exploratory |
| Systematic Review | Answer a specific question | High |
| Meta-analysis | Statistically combine the results | The highest if the studies are homogeneous |
Systematic reviews address several regulatory requirements
- Clinical evaluation of a medical device: Regulation (EU) 2017/745 (MDR) requires a clinical evaluation based, in particular, on the literature, especially when equivalence to an existing device is claimed.
- Postmarketing surveillance: Monitoring and regularly reviewing the literature help keep the clinical evaluation up to date and inform PMS/PMCF activities as well as, where applicable, the PSUR.
- Trial design: It justifies the choice of comparator, the primary endpoint, and the assumptions used to calculate the sample size.
- Regulatory documents for the drug: It serves as the basis for the introduction to the investigator’s brochure and the discussion section of the clinical study report.
For medical devices, the MDR serves as the primary regulatory framework. The methodological recommendations in MEDDEV 2.7/1 Rev. 4 remain relevant to several aspects of research and literature review, particularly Section A5, and supplement the applicable MDCG guidelines, especially regarding clinical evaluation and the demonstration of equivalence. For the clinical evaluation report (CER), the literature review distinguishes between two components: the state of the art, which describes therapeutic alternatives and known risks in the field, and data specific to the device or its equivalent.
Claiming equivalence has become more demanding under the MDR. It is necessary to demonstrate similarity in technical, biological, and clinical characteristics and to justify access to data on the equivalent device. The literature review therefore cannot be limited to favorable publications; it must also document reported incidents, recalls, and limitations.
A systematic review consists of 6 steps
- Formulate the question. The PICO format (Population, Intervention, Comparison, and Outcomes) transforms a general inquiry into a testable question.
- Write the protocol. It specifies the databases to be searched, the search terms, the inclusion and exclusion criteria, and the analysis method. It can be registered in a public registry such as PROSPERO.
- Review the sources. PubMed/MEDLINE, Embase, and the Cochrane Library are the primary databases, supplemented by trial registries (ClinicalTrials.gov, CTIS) and gray literature. Each search is dated, and the search terms are recorded.Generative AI tools can help with sorting, but they do not replace this traceability.
- Select studies. Studies are selected according to a predefined procedure, ideally involving independent evaluation by two reviewers and a process for resolving disagreements. The PRISMA flow diagram summarizes the number of articles at each stage.
- Assessing quality. The risk of bias in each study is analyzed using validated tools, such as RoB 2 for randomized trials or ROBINS-I for non-randomized studies. The overall level of confidence can be graded using the GRADE method.
- Synthesize and write. The extracted data are presented in tables and then discussed. The review is reported transparently in accordance with the PRISMA 2020 guidelines, specifically using its 27-item checklist and flow diagram.
The meta-analysis provides a quantitative synthesis
When the selected studies measure the same outcome in a comparable way, their results can be combined statistically. A meta-analysis yields a more precise overall effect than any single study taken in isolation, as shown in a forest plot. It can also reveal an effect that individual studies—which were too small—could not detect on their own, or, conversely, show that a reported benefit does not hold up to a pooled analysis.
It requires technical decisions that our biostatisticians document right from the protocol stage: fixed-effects or random-effects models, measures of heterogeneity (I² statistic), sensitivity analyses, and assessment of publication bias. If the studies are too heterogeneous, a qualitative synthesis is preferable to a misleading overall figure.
Mistakes That Weaken a Literature Review
- Search criteria are too narrow: relevant studies are overlooked, which the reviewer quickly notices.
- Selection criteria changed during the study: Any deviation from the protocol must be justified and documented.
- Lack of bias assessment: A low-quality study does not carry as much weight as a well-conducted randomized trial.
- Outdated documentation: For a medical device dossier, the review must be updated regularly throughout the product's lifecycle.
Comprehensive expertise, from literature review to statistical analysis
Our scientific writers and biostatisticians work together throughout the entire process: defining the research question, developing a literature search strategy, selecting and critically evaluating publications, extracting data, synthesizing results, and, when appropriate, conducting meta-analyses. This approach allows us to support a wide range of projects, including systematic reviews for the clinical evaluation of medical devices, clinical research projects, scientific publications, and regulatory submissions. Learn more about our expertise in medical and scientific writing, or contact our teams to discuss your project.
Frequently Asked Questions About the Systematic Literature Review
Need assistance?
Are you preparing a clinical evaluation or research project that requires a systematic literature review? Our Statitec experts will guide you every step of the way, from the PICO question to the final report: literature search strategy, selection and critical appraisal of studies, synthesis of results, and meta-analysis, if applicable.
