Clinical evaluation is a continuous process used to evaluate the conformity of medical devices to safety and performance requirements, when used as intended by the manufacturer. The role of literature search in clinical evaluation is immense. Literature search is performed to identify clinical data which is not available with the manufacturer, but is critical to the completion of the clinical evaluation. The clinical data identified undergoes a series of stringent checks to assess its suitability before being included in the clinical evaluation. The data obtained from various sources are appraised to evaluate their relevance and contribution, to evaluation of clinical safety and clinical performance. The appraised dataset is then analysed to check if they can be used collectively to demonstrate the compliance of the device to essential requirements.
The first step towards conducting a thorough literature search is to prepare a search plan considering the scope of clinical evaluation set by the manufacturer. The review questions based on which the search will be conducted are mentioned in the plan. The questions are formed taking into consideration aspects like safety, performance, side-effects and risks. Another important aspect that needs to be established with the clinical data is clinical background and state of the art. The most commonly used search method is PICO (P- Population, I- Intervention, C- Control or Comparison, O- Outcome).
Depending on the device the PICO framework can be adapted to include T-time (PICOT), S- Study design (PICOS) or C- Context (PICOS). Other search methods like PRISMA statement, MOOSE proposal, SPICE, PEO and ECLIPSE are also available. Medical Device Clinical Evaluation The quality of the data is the determined by the source. Different sources like scientific databases, internet searches, non-published data, and citations are used to obtain the required clinical data. Some of the scientific databases used are Pubmed, Cochrane central trial register, PDQ evidence, Trip and CINAHL. Internet search is mainly conducted on Google Scholar, manufacturer website, FDA, Cochrane database of systematic review etc. Non-published data include information from instructions for use, implant registries information and data presented in congress. The index of citation in scientific literature is available in Scopus and Web of science. The search is performed across various sources by forming keywords related to each of the review questions. Keywords are words or concepts that are of great significance. Each keyword or phrase is carefully selected so as to get the best possible results while searching. Logical operators like AND, OR, NOT etc. are used to combine different keywords to make the search term more meaningful. Another method to improve the results of search is use of MeSH terms. It is a controlled vocabulary that can be used to build searches that utilises all the synonyms and related words connected to the keyword. Filters are also applied to refine the search results to meet the expected criteria.
A comprehensive search is performed to obtain clinical evidence corresponding to the safety and performance requirements being evaluated, since any lack in clinical data held by the manufacturer is compensated using the data obtained through literature search. A two level screening of the search results is performed -the title or abstract screening based on inclusion and exclusion criteria and - full text screening of the included articles to determine the relevancy. The search results that pass the screening are used to form the reports for each of the review questions. This report is used present the results of the search that address the aspects like safety, performance, benefits to patient, side-effects, adverse reactions, state of the art etc. These steps ensure the quality of the literature included in the reports. what is clinical evaluation report
The appraisal of the included literature is performed based on various criteria that assess the methodological quality, scientific validity, relevance to the device and intended use. The data contribution of the articles is assessed based on factors like data source, measurement of outcome, duration of follow-up, statistical significance and clinical significance. Only clinical data from literatures that meet these criteria are considered to be clinically relevant. Hence, these criteria should be considered while searching for literature. The data found to be relevant after appraisal are then analysed based on requirements on safety, performance, risk-benefit profile and acceptability of side-effects. The aim of this assessment to demonstrate the compliance of the device to essential requirements corresponding to clinical safety and performance, when used as intended.
The search for literature forms an important part of clinical evaluation as it is used during various stages of clinical evaluation explained in MEDDEV 2.7/1 Rev 4. The clinical data obtained from the literature together with the data held by the manufacturer provides the clinical evidence needed to prove the safety and performance of the device. In some cases when the data from manufacturer is insufficient, clinical data from literature will represent a major part of clinical evidence. It can therefore be concluded that literature search is vital for clinical evaluation.
ABBREVIATIONS
PRISMA statement: Preferred Reporting Items for Systematic Review and Meta Analysis
MOOSE proposal: Meta Analysis Of Observational Studies in Epidemiology
SPICE: Setting, Population or Perspective, Intervention, Comparison, Evaluation
PEO: Population, Exposure, Outcomes
ECLIPSE: Expectation, Client group, Location, Impact, Professionals, SErvice
REFERENCE
MEDDEV 2.7/1 Rev.4- Clinical Evaluation: A Guide for Manufacturers and Notified Bodies under Directives 93/42/EEC and 90/385/EEC