CRFWEB INFO SHEET

Post Market Clinical Follow up (PMCF) studies – an Overview

The purpose of PMCF studies is to examine how your device performs when used as intended so that clinical performance can be verified and to collect safety information.

Under the MDR, manufacturers will be expected to perform more PMCF studies throughout the device lifecycle than was deemed acceptable under MDD.

Notified Bodies can require a PMCF study at any time if, in their judgement, data provided to support the safety and performance of a product are inadequate or incomplete. This may also be requested by an EU expert panel if they disagree with the Notified Body’s assessment.

The requirements for PMCF studies will be closely aligned to those of pre-market clinical study plans. The PMCF study must demonstrate the clinical safety and performance of a device throughout its lifetime when used as intended. However, a PMCF study does NOT need to include randomization, excessive patient selection criteria, or control groups.

When are PMCF studies required?

According to the guidance MEDDEV 2.12-2, and which still provides a useful baseline for MDR, the circumstances under which PMCF studies may be necessary are as follows:

Innovation: the device’s design, materials, substances, principles of operation, technology or medical indications are novel
Significant changes to the product or to its intended use for which pre-market clinical evaluation and re-certification has been completed
High product-related risk
High-risk anatomical locations
High-risk target populations (e.g., children, elderly)
Severity of disease/treatment challenges
Questions of ability to generalize clinical investigation results
Unanswered questions of long-term safety and performance
Results from any previous clinical investigation, including adverse events or from post-market surveillance activities
Identification of previously unstudied subpopulations which may exhibit different benefit/risk-ratio (i.e., hip implants in different ethnic populations)
Continued validation in cases of discrepancy between reasonable pre-market follow-up time scales and the expected life of the product
Risks identified from the literature for other data sources or similar marketed devices
Interaction with other medical products or treatments
Verification of safety and performance of device when exposed to a larger and more varied population of clinical users
Emergence of new information on safety or performance
Where CE Marking was based on equivalence

PMCF studies may NOT be required when the medium/long-term safety and clinical performance are already known from previous use of the device or where other appropriate post-market surveillance activities would provide sufficient data to address the risks. As the Notified Body will assess the justification provided by the manufacturer for the non-performance of PMCF studies, so it’s extremely important that your submission is as detailed and robust as possible.

PMCF surveys

Whilst randomised clinical trials and registries provide higher levels of evidence, they may not always be realistic or even possible to carry out for certain devices. For yet other devices – particularly in the lower risk classes – they may not be strictly required under MDR PMCF regulations.

Patient or subject surveys may well be an acceptable alternative for such devices, and these are attractive as they are generally less expensive, onerous and time consuming when compared with the alternatives. Below, we explain the basics principles underpinning the design of a PMCF survey.

Developing your PMCF Survey questionnaire

It’s been a common misconception that surveys are not usually an acceptable data collection method for PMCF. Historically, though, surveys have been rejected not as method per se, but because question quality fails to pass muster with the Notified Body.

Surveys are most useful when they are used to gather safety and performance data from patients or subjects – providing of course that the sample size has been correctly calculated and that in the event, the survey is successfully administered to a viable number of respondents.

First things first; it’s important to undertake a thorough literature review to establish whether a previously validated questionnaire exists that could be repurposed for use with your device.

A “validated” questionnaire is a questionnaire (or scale) that has previously been developed and administered to a group of patients or subjects exploring comparable safety and efficacy metrics across a cohort representative of your intended respondents. There are standard and well understood processes for validation that need to have been conducted and completed on a correctly calculated representative sample. If a questionnaire exists in a different language, the translated questionnaire will need to be validated in the new language. Validated, standardized questionnaires deliver reliable, well-structured datasets that are easily analysed. The need for data cleansing and other processing activities such as merging text answers will also be minimised. All in all, using a previously validated questionnaire – if possible – will save an awful lot of time and money.

Survey Design – Key Considerations

If no appropriate validated questionnaires can be discovered, then there will be a need to develop and validate a new questionnaire. Naturally you’ll need to bring together a team of people with the relevant clinical experience of your product’s use in the patient population together with clinical research experts to drive the development of your questionnaire.

Experience suggests that it’s surprisingly easy to be drawn into developing overly complicated questionnaires. It might appear to be stating the obvious, but the whole process should be driven by first establishing exactly what it is that you need to know to fulfil the needs of PMCF. The questionnaire should be designed with this in mind, first and foremost. Secondary purposes may be considered – surveying is a potentially expensive and time-consuming activity and therefore there is some merit in optimising the opportunity. However, any questions that don’t contribute directly to the primary objective should be subject to fierce scrutiny; the more complex – and lengthy – the questionnaire, the more likely there are to be issues with subject compliance and validation. Important design issues that need to be considered are:

How long should the questionnaire be? At the risk of repetition, the guiding principle should be “keep it simple”. Ask only what you really need to know
Who will be answering the questions? Will it be physicians, or users?
What do I need to know about respondents? This is particularly important if your survey is focused on gathering information from patients. Vital considerations might include:
1. Literacy and language skills
2. Age and health status
3. Race, gender, and sexual orientation

Information about protected characteristics – e.g. race, gender, sexual orientation, disability, and age – is often included to identify where these might be potential factors in the quality of the patient experience you are seeking to explore. Many people prefer not to disclose some or all elements of personal information, so it’s important to explain why it’s being sought and to provide reassurances about anonymity. Usually, it’s best to position such questions towards the end of the questionnaire since people are less likely to complete the other questions if they are put off at the start.

Keep language as simple as possible. Why would you do otherwise?
Avoid using jargon unless relevant and useful when asking questions of professionals
Always be specific; for example, when asking about factors such as frequency avoid using vague, subjective terms like “regularly” instead, use prices times like “daily”, “weekly”, “monthly” etc.
Avoid negative phrasing. This is particularly relevant when seeking to understand the extent to which someone agrees or disagrees with a proposition. DO ask “To what extent do you agree that people should pay for their prescriptions?” DO NOT ask “To what extent do you agree that people should not have to pay for prescriptions?”
Avoid “double-barrelled” questions Don’t create ambiguity in how to accurately answer a question by asking more than one thing in that question e.g. Don’t ask “Do you have less pain and less anxiety after …” Which issue do you really need to know about? Focus on that, and if you need to know about both issues, ask two separate questions.
Avoid “leading” questions that might suggest that one answer is any way preferable than any other. Do NOT ask “Over 80% of users report significant pain reduction one month after beginning treatment. How would you describe your pain reduction? Significant, slight, none at all?”
Check and check again that questions DO NOT implicitly assume prior knowledge or are structured in a way to create bias. For example, the question “Do you support the idea that pensioners should not pay for their TV license?” implicitly assumes the respondent knows this and secondly, “do you support…should not” forces the respondent into a binary decision – and people, by and large do not like to be seen supporting harsh or negative positions.
Make sure that the question types selected will deliver the data you need. Open or closed questions, unipolar and bipolar scales can all have a useful role to play but you need to decide which type of question is most effective for delivering the data you need.
Test, test, test. First, seek the views from a panel of people not involved in the project to test for clarity – is the language as clear and as straightforward as possible? Have any ambiguities slipped through the net? This can sometimes be brutal, but always worth doing. When the wrinkles have been ironed out, then you are ready to undertake a Preliminary Pilot Test with a small patient or user group that will be fully representative of your sample. Do the questions make sense to the small test group? What suggestions do they have? You can get an idea of the response distribution to each question which can help decide if there is enough variation in the responses to justify moving forward with a large-scale pilot test. Feasibility and the presence of floor (almost all respondents scored near the bottom) or ceiling effects (almost all respondents scored near the top) will help you decide what questions need to be reworked or removed at this stage.
Pilot Testing and Validation. If you’ve iterated your preliminary testing process to your satisfaction, this is next stage, to thoroughly test will a significant sample size of your target cohort to fully validate the questionnaire. Validation is a term that covers a range of statistical activities aimed at ensuring that the questionnaire really does do what it’s intended to. Statistical validation methods include tests for:
1. Reliability (internal consistency, test-retest reliability, inter-rate reliability)
2. Validity (content validity, construct validity)

It’s beyond the scope of this article to explain the technicalities of these methods, but all these processes are integral to the development of a gold-standard clinical questionnaire. For many devices studies in the past, though, it would be fair to say that these activities have been honoured more in the breach rather than the observance. However, as the quality of clinical data expected from PMCF is more stringently defined and monitored, serious consideration should be given to ensuring that the process is rigorously followed.

For further information concerning PMS & PMCF follow the links below:

How can we help you?

With the advent of MDR, it’s harder than ever to argue that NOT investing in EDC makes sense.

CRFweb is a full-service EDC system optimized for medical devices. CRFweb will support your journey to compliance providing a single, integrated platform for:

Clinical trial design
Data capture by site or subject
Trial process management and troubleshooting
Data management, analysis and reporting

The CRFweb software suite delivers all the key functionality you will need to ensure an MDR-compliant clinical trial in an agile, simple-to-use package optimised for the needs of medical devices at a fraction of the cost of the industry big names.

Contact us today to book your demo

5 Star Reviewed

What our clients say

Maria Galligan, PMD Solutions

" The Clindox team have been an incredible support and are always available to help in any way they can."

Dr Salvi, CRF Health, India

" The time and cost saving is really significant. We're very happy with the performance or CRFweb and I wouldn't hesitate to recommend it."

Nigel McLean, Cytosystems, UK

" CRFWEB eCRF, plus their helpful and accommodating staff, helped us quickly set up and personalize, with quick turn-around times of modifications, exactly what we, the clinical staff, and statistical consultants demanded, all at an affordable price "

Kelly Seamans, Atlantia

" The Clindox (CRFWEB) team are dedicated and enthusiastic about accommodating the unique requirements of each individual study and are great value for money."

Dr Satish Kumar, G7 Synergon

"Our international CRO business has utilized Clindox's integrated randomization module and Kit Management. The functionality provides everything we need, and having key trial components integrated into one system is a huge advantage to us in terms of trial efficiencies. Having both randomization and kit management capabilities, aligned with the EDC really simplifies the whole process and significantly reduces the time and effort required to manage our studies and keep control of our processes and inventory. "

Next Steps

For a quick system overview, click here

Seeing is believing. Book a web-based demo now.

Book a demo

Useful Links

Latest News

June 20, 2022

Meeting the IVDR challenge: digitizing the collection of study data

March 14, 2022

Covid-19: What has been the impact of the pandemic on India and how have Indian Clinical Trials been affected?

January 18, 2022