Tonika Chester, Chris Webb, Martha Vakalopoulou

Since the implementation of the Medical Device Regulation (2017/745) (MDR) in Europe, which mandates manufacturers to proactively monitor the safety and the performance of their medical devices throughout its entire lifecycle, Post-Market Clinical Follow-Up (PMCF) surveys have become widely accepted as an appropriate and essential component of PMCF data collection.

According to the Medical Device Co-ordination Group (MDCG) 2020-6 guidance document, Appendix III (as shown below) ranks data collection methods based on the level of evidence they can provide, with 1 being the highest, and 12 being the lowest. Between clinical investigations, registries, and other non-clinical sources of data, PMCF surveys fall into two categories, with high-quality user surveys assigned an evidence level of 4, and general usability surveys assigned a lower evidence level of 8.\

When to implement a high-quality (level 4) PMCF survey

Fundamentally, the level of evidence required, and therefore the type of survey to implement, will depend on several factors, namely, the risk classification of the device, time spent on the market, and existing data, as outlined below:

Medical devices with higher risk classifications that have been on the market for a short time, and likely to have limited clinical data, will require more extensive evidence. Although clinical investigations would provide the highest level of evidence, it is worthwhile to note that high-quality (level 4) PMCF surveys can also be used. Furthermore, legacy Class III (non- implantable) devices, without any clinical data, also require a minimum of level 4 clinical data to be certified under the MDR.

High-quality (level 4) PMCF surveys can serve to supplement clinical findings, especially in cases where clinical data gaps (ie, certain age groups or device variants) have been identified, by providing real-world evidence on the safety and performance of a medical device over its entire lifecycle. For this reason, high-quality (level 4) PMCF surveys allow for the evaluation of potential systematic misuse and off-label usage, which may not be possible to identify due to the strict inclusion/exclusion criteria that are associated with clinical investigations.

How do high-quality (level 4) PMCF surveys provide a higher level of clinical evidence than general usability (level 8) PMCF surveys?

The prospective nature of high-quality (Level 4) PMCF surveys allow for more reliable and accurate data to be obtained compared to their level 8 counterparts, which tend to be recollective in nature. This is facilitated by fundamental differences in their survey designs, as shown below:

Typically, a well-designed high-quality (level 4) PMCF survey will allow clinicians to complete a survey on a case-by-case basis using patient records, which reduces, if not eliminates, the effects of the recall bias we often see in data collected from general usability (level 8) PMCF surveys. 

However, it is important to note that as a single high-quality (level 4) PMCF survey focuses on a specific medical device, you will likely need a larger, statistically justified sample size in order to address your safety and performance objectives. For this reason, implementing high-quality (level 4) PMCF surveys can be more expensive, and require a longer period of time to collect data.

Shift towards high-quality (level 4) PMCF surveys

Over the last year or so, there has been a greater push by notified bodies and medical device manufacturers to implement high-quality (level 4) PMCF surveys, compared to their level 8 counterparts. Despite this, it is important to consider the following statement from the MDR Article 61(1):

“The manufacturer shall specify and justify the level of clinical evidence necessary to demonstrate conformity with the relevant general safety and performance requirements. That level of clinical evidence shall be appropriate in view of the characteristics of the device and its intended purpose.”

Therefore, while high-quality (level 4) PMCF surveys appear to be an attractive, go-to alternative to general usability (level 8) surveys, there are some cases, such as low-risk classification devices that belong to a group of well-established technologies, in which the latter approach is both justified and sufficient.