Literature Reviews for EU MDR

Want to know what the most commonly criticized document in an MDR submission is?

Your clinical literature review.

Reviewing the literature is essential to the clinical evaluation process for medical devices.

The clinical evaluation report assesses a medical device’s benefit-risk profile to ensure its continued safety, efficiency, and compliance. A critical part of the clinical evaluation report is the literature search and review, which determines the current state-of-the-art, and the behavior of a medical device in the market, while summarizing the clinical data.

The clinical evaluation report (CER) demonstrates compliance with the European Medical Device Regulation (MDR; 2017/745) for medical devices and MEDDEV 2.7 rev. 4.

MEDDEV 2.7 rev 4 describes a medical device’s clinical evaluation process for medical devices and the requirements for the literature search, i.e., searching for relevant literature, compiling research, and explaining how the gathered clinical evidence pertains to the safety and efficiency of your medical device.

Note: This article uses the terms literature search and literature review interchangeably.

For many medical device manufacturers, the literature search represents the majority, if not all, of the clinical data they collect on their devices during device development and subsequent clinical evaluation updates. With it, it is possible to demonstrate and critically evaluate the safety and efficacy of their device during clinical evaluation.

Even for companies that have performed clinical trials, a literature search still ensures a summary of the valuable clinical data regarding the different aspects of their device and its performance.

Lastly, and maybe most importantly, the literature search is mandatory for clinical evaluation reports under the EU MDR and will help determine if your medical device receives a CE mark.

A good literature search provides the following:

• The context for a problem.
• Its proposed solution.
• A disease/condition or medical process problem.
• The intervention is designed to treat or solve it.

It should quickly summarize the maximum amount of published literature.

A well-organized and thorough literature search provides manufacturers with critical information on who else has attempted to solve this problem, which approaches were taken (similar devices on the market), how they were received in the marketplace, and if they were effective and safe.

It can also reveal to the manufacturer whether other available products may be better suited to solve the problem than the device they’re attempting to put on the market.

Understanding how other devices have been received can help ensure that the new solution will be well received.

For example, suppose a device is designed to address a specific concern shared by many physicians but a few patients. In that case, it might not be as attractive as a similar device that addresses concerns held by both patient groups and doctors.

A thorough literature search will also help you create an excellent systematic process that will set you up for future documentation updates, reducing future workload.

A literature review serves many purposes in the medical device development process, from informing the R&D team about what has been done before to understanding how a new device fits into the market and helping determine whether a project is worth pursuing.

An example of a good literature review is one that also helps manufacturers communicate their ideas to partners, investors, and other stakeholders.

Reading a good literature review can better understand a device’s value and benefits/risks more than other methods, like reading articles describing the device or watching a video.

Although the literature review for clinical evaluations is more about pre-market surveillance, the clinical evidence gathered can significantly help you design your medical device’s post-market surveillance program.

A literature review can help convince regulatory bodies that a new device’s benefit-risk ratio is beneficial enough to approve it.

It also helps support claims made by manufacturers about their products, including their intended use, safety, and effectiveness. Furthermore, it can be used as clinical data in marketing materials, such as brochures and videos.

This is particularly important when introducing an innovative product to the market because it can help healthcare professionals understand how it differs from existing devices.

A document showing how the theoretical device will work in the market or exactly which deficiencies current medical device designs have – which the new device is covering – is beneficial.

Even when everything else goes right, you might not get the relevant results you expected from your literature review.

This can be due to incorrect framing of the research question, needing to understand exactly how your search terms interact with the database, or even problems with choosing suitable databases to search.

So, how can you get the right results in your literature reviews that will support your clinical evaluation and demonstrate the acceptability of your benefit-risk profile?

Not all databases are equal, and they will not necessarily yield the same results for your literature review. Therefore, more is needed to search just one – to find articles and research examples relevant to your device, you need to search multiple databases.

For example, using the (unfiltered) search term “Silk fibroin AND (dermal OR filler OR “dermal filler” OR “soft tissue augmentation”) will yield three results in PubMed, 16,700 results in GoogleScholar, and 181 results in LIVIVO.

Some of these differences are due to how the search engines work, while others are due to different publications available in each database.

It is essential to understand how each database works and how to phrase your literature review terms to get valuable results in each database.

If you are going after a CE Mark under EU MDR, then you’ll want to select databases that give you a comprehensive and global (with a EU focus) result set.

Some literature reviews will give you a plethora of information, primarily if you work with devices in larger fields, such as cardiovascular and orthopedic surgery. However, it is essential to evaluate all that information critically.

You must set up a straightforward process for conducting your literature search before you start, i.e., you have to write a literature search protocol. You need to have each step in place before you do it – otherwise, it is easy to get lost when you’re halfway through or, even worse, when you’re almost done.

You should always do a preliminary search to get an overview of the information and publications and test the waters of your search terms. Get a feel for the task.

Once you have determined what you are looking for and how to get there, write a detailed plan for the search. Include search terms, inclusion/exclusion criteria, key concepts, the databases you will search and what tools you will use for your search.

You are writing a recipe for your search you can follow when you start. This recipe becomes the literature search protocol.

Expanding on the previous step, generating bibliographies and search result files, and storing references, let alone organizing them, is a pain in the butt.

There is a lot of meticulous detail work and moving parts to the literature search, and there is no need to complicate matters for yourself – use as many digital tools to make your life easier as possible. Download your search results, use a citation manager, and ensure you’re using the right programs when reviewing your data results.

The team at CiteMed has spent years refining and perfecting our Systematic Literature Review Software platform so that it’s easy to use, and can save you massive amounts of hours per review project. If you’d like to give it a test drive, contact us today.

Boolean operators are always highlighted when people advise on literature reviews, and for a good reason. The right boolean operator can make or break your literature search (and sanity).

Operators allow you to narrow your literature review to focus on only the publications relevant to you and save you hours. Likewise, your search can be widened if you are not getting enough results.

The most well-known operators are AND, OR, and NOT. Use them wisely.

Most literature reviews have some search restrictions in place. It can be a time limit, i.e., only doing literature searches for the last five years. It can be a language (no use getting results in Japanese if you can’t understand them), a medical discipline, or the type of publication you are looking for.

Every database will have filters you can use to refine and narrow down your search – make sure to use them as much as possible. It is your most straightforward tool to obtain the results you need.

A complete Protocol and search strategy are defined by our team when working on Literature Search projects. You can read more about them here.

More than 3 million scientific articles are published in English annually, and publication rates are growing by almost 10% annually.

This information explosion makes it increasingly difficult for researchers to keep up with the latest field developments. In addition to the apparent challenge of keeping up with new content, there is another problem that many researchers face: understanding the context, methods, and goal of the research being presented. Those can often only be understood by reading papers in detail. Hence, the regulatory team members must read, organize, and report on many articles.

The increasing addition of papers to the scientific community has created information overload. For many regulatory people and medical writers, it is simply impossible to read through all the search results carefully to find relevant information for clinical evaluation reports (CER).

Artificial Intelligence (AI) is not novel, and AI allows for an excellent collaboration between humans and machines, the very best of which is yet to come. Even so, currently available software can significantly reduce the workload for medical device manufacturers preparing literature reviews and clinical evaluation reports.

Manufacturers have more control over the data screening process and data extraction in an automated process than in a manual process.

Several automated programs with integrated artificial intelligence and machine learning allow users to create labels, filters for the data extraction process, inclusion criteria & research questions customization, and indexing to balance precision and recall.

Some come with summarisation and reporting tools highlighting hidden connections in the textual data. In addition, automated bias assessment and study identification ensure the automated approach only needs a little human intervention beyond the search strategy.

Interactive machine learning systems are more straightforward than one would assume. For example, some software offers data visualization, meta-analysis, evidence synthesis, and text mining, which can all be highly useful during literature screening and clinical data appraisal.

Despite advanced with Artificial Intelligence, we advise our clients to avoid it when performing SLRs for an EU MDR submission. The reason? Because they invite further scrutiny form your Notified Body auditor, and the ever-challenging question of “how can you validate your results?”.