The Clinical Evaluation Report (CER) stands as a cornerstone document in the lifecycle management of medical devices, including Software as a Medical Device (SaMD). It evidences the medical device’s safety, performance, and clinical benefits. This article delves into the literature review process within the CER, focusing on its alignment with the Medical Device Regulation (MDR) and relevant Medical Device Coordination Group (MDCG) guidances, particularly emphasizing SaMD.
Understanding the Clinical Evaluation and Literature Review
Clinical evaluation is a methodologically sound ongoing procedure to collect, appraise, and analyze clinical data pertaining to a medical device. The literature review is a crucial component of this evaluation, serving to establish and maintain the conformity of the device in question. You can find more details in the following articles of the MDR:
Article 61 – Clinical Evaluation
Annex XIV – Clinical Evaluation (Plan and Report)
Given the unique nature and rapid development cycle of SaMD, the literature review process must be adept at identifying clinically relevant data that reflect the most current understanding and utilization of the technology. This includes, but is not limited to, clinical trial results, post-market surveillance data, and real-world evidence.
Key Steps in the Literature Review Process for the clinical evaluation
1. Defining the Scope
The scope encompasses the objectives, the device under review, its intended medical purpose, and the context of use. For SaMD, this might include specific functions such as diagnostic algorithms, therapeutic suggestions, or patient monitoring capabilities.
2. Planning the Search Strategy
This involves selecting databases and search terms, defining inclusion and exclusion criteria, and setting time frames for the search. The strategy must be comprehensive enough to capture all clinically relevant literature.
3. Executing the Search
Public databases like PubMed are invaluable resources when you write the clinical evaluation report. The search can be conducted using a combination of keywords related to the SaMD, its medical purpose, and clinical outcomes of interest.
Searching in PubMed: A Brief Guide
- Step 1: Develop a list of relevant keywords and MeSH (Medical Subject Headings) terms related to your SaMD.
- Step 2: Combine these terms using AND/OR operators to refine your search.
- Step 3: Utilize filters to narrow down the results by publication date, study type, and language, if necessary.
- Step 4: Review the search results, and select articles that meet your predefined criteria.
4. Screening and Selection
Screen titles and abstracts to eliminate irrelevant studies, followed by a full-text review of the remaining articles to determine their suitability for inclusion. Focus on the state of the art, the intended purpose of your device and risks associated with the device. Screen the abstracts first to identify irrelevant publications.
5. Data Extraction and Analysis
Extract pertinent data from the selected literature, including study design, population characteristics, outcomes, and limitations. Analyze this data in the context of the SaMD’s intended use and clinical claims.
6. Synthesizing Evidence and Drafting the Report
Synthesize the extracted data to form a coherent narrative regarding the SaMD’s safety, performance, and clinical benefits. This synthesis forms the basis of the literature review section of the clinical evaluation report.
Evaluate the results in the clinical evaluation
The quality and relevance of the selected literature are critical. Utilize established appraisal tools and checklists (e.g., CONSORT for clinical trials, STROBE for observational studies) to assess the methodological quality of studies. The evaluation should also consider the relevance of the evidence to the specific context of the SaMD’s application.
Examples to illustrate concepts
Example 1: Diabetic Retinopathy Detection Software
Background: Software designed for the early detection of diabetic retinopathy through analysis of retinal images.
Literature Search and Review Process:
- Objective: To evaluate the software’s diagnostic accuracy and usability in clinical settings.
- Search Strategy: Utilize PubMed to search for clinical trials, observational studies, and reviews focusing on diabetic retinopathy detection using image analysis software. Keywords might include “diabetic retinopathy detection,” “retinal image analysis software,” and “SaMD.”
- Inclusion Criteria: Peer-reviewed articles published in the last five years, studies that specifically evaluate software accuracy against gold-standard diagnosis, and studies reporting on software usability by healthcare professionals.
- Exclusion Criteria: Articles not in English, studies focusing on hardware devices for retinal imaging without software analysis, and case reports.
Evaluation of Results:
- Analysis would focus on the specificity and sensitivity of the software reported in the studies, comparing it to traditional diagnostic methods.
- Assessment of any reported limitations or adverse events related to the software’s use in a clinical setting.
- Keep the overview! When searching for publications you will quickly identify over 100 potential publications. Use a basic Excel spreadsheet to track the publications and to document the search results.
Findings:
- Several high-quality studies might demonstrate the software’s accuracy is comparable to or exceeds that of manual retinal image analysis by experienced clinicians.
- User feedback from healthcare professionals may highlight the software’s integration into clinical workflows and its ease of use, with suggestions for further improvements in user interface design.
Example 2: Mobile App for Asthma Management
Background: A mobile application that helps patients manage their asthma by tracking symptoms, medication use, and providing personalized advice based on entered data.
Literature Search and Review Process:
- Objective: To assess the effectiveness of the mobile app in improving asthma control and patient adherence to medication.
- Search Strategy: Searches in PubMed for randomized controlled trials (RCTs), quasi-experimental studies, and systematic reviews related to mobile apps for asthma management. Keywords could include “asthma management app,” “mobile health,” and “patient adherence.”
- Inclusion Criteria: Studies published within the last ten years, articles that measure clinical outcomes such as asthma control test (ACT) scores, emergency department visits, and medication adherence rates.
- Exclusion Criteria: Non-peer-reviewed articles, studies focusing on non-asthmatic conditions, and articles on apps without interactive patient management features.
Evaluation of Results:
- Application of appraisal tools like STROBE for observational studies to assess methodological quality.
- Synthesis of evidence to determine the impact of the app on asthma control measures and whether it leads to improved medication adherence.
Findings:
- Evidence from multiple RCTs may indicate that patients using the app show significant improvement in ACT scores and reduced emergency department visits compared to those receiving standard care.
- Data might also suggest high user engagement with the app contributes to better adherence to prescribed asthma medications.
Conclusion
The literature review process in the clinical evaluation report for SaMD under the MDR and MDCG guidances is a crucial step in the certification process that ensures the safety, performance, and clinical utility of your software. By systematically searching, selecting, and evaluating clinically relevant literature, manufacturers can substantiate their clinical claims and demonstrate conformity with regulatory requirements. The process not only supports regulatory compliance but also fosters trust among users and stakeholders, ultimately contributing to the advancement of healthcare through innovative medical software solutions.
