Focus: The Role of Statistics in the Drafting of a Post-Market Clinical Follow-Up (PMCF) Plan for Medical Devices
In the sector of medical devices, ensuring safety and efficacy is not limited to the design and manufacturing phase only.
Post-market surveillance is essential to monitor a device performance under real conditions.
A key element of this surveillance is the Post-Market Clinical Follow-Up (PMCF) Plan, a strategic document describing the activities required to collect and analyse clinical data once a device has been put on the market.
Regulation EU 2017/745 on Medical Devices (MDR) has substantially changed the specific requirements for the PMCF, turning it into a systematic, continuous and essential activity.
Statistics play a crucial role in the PMCF drafting and implementation. From the identification of monitoring targets to result assessment, statistical analysis ensures that any decision is supported by sound and impartial evidence.
PMCF planning
Planning a PMCF study requires a detailed and thorough approach.
First of all, it is essential to determine the sample size required to ensure reliable results.
The sample size determination is one of the most important phases when planning a PMCF study, as the robustness of the conclusions to be deduced from the collected data depends on it.
To calculate an appropriate sample, the first thing is the identification of the target population, that is the group of patients or users representing the device real market.
Then, it is important to consider the prevalence of the events of interest. These data can be derived from pre-market studies, scientific literature or clinical records. Based on this information, the sample size required to obtain a sufficient statistical power is calculated, often fixed at 80% or 90%, minimising the risks of type I (false positive) and type II (false negative) errors.
Finally, it is fundamental to establish a detailed timeline for the data collection. A clear time planning ensures a systematic and continuous monitoring of the device performance.
Data Analysis
During PMCF implementation, data analysis requires advanced expertise and suitable statistical tools. Collected data are processed using software like R, SAS or SPSS, each offering specific functions for different types of analysis.
Device performance is monitored via descriptive and inferential analyses. Parameters like clinical efficacy or user’s satisfaction are assessed using appropriate statistical tests, such as t-tests, ANOVA or regression models.
Moreover, machine learning techniques, implemented with libraries such as scikit-learn or TensorFlow, can be applied to identify data patterns.
A further step is the identification of data trends and anomalies. Exploratory methods like the principal component analysis (PCA) or cluster analysis help in detecting subgroups of patients or specific characteristics associated to adverse events. Data visualisation is supported by tools like Tableau, ggplot2 in R, favouring result interpretation and communication.
Uncertainty Management
The management of uncertainty is a crucial aspect of PMCF, since clinical data collected in real contexts can present significant variability. Statistics offer specific tools to quantify these uncertainties, improving conclusion robustness.
Confidence intervals are used to represent estimation precision, providing a range within which the real value of the parameter is believed to be with a certain probability (for instance, 95%). These intervals help assess estimation reliability and communicate the degree of uncertainty associated with the results.
Hypothesis tests allow us to verify if the differences observed in the data are statistically significant or if they can be ascribed to chance.
A further approach consists in the use of bayesian models, combining observed data with prior information (priors) to update the parameter estimation interactively. This method is particularly useful when available data are limited or highly variable.
The management of uncertainty requires a clear and transparent communication of results. Diagrams showing confidence intervals, sensitivity analyses and probability distributions can help to better understand the degree of reliability of the conclusions and to take informed decisions.
Reporting and result interpretation
The reporting phase requires a special attention to clarity and precision in the presentation of the results. Statistical reports should be organised in such a way as to allow easy understanding, even by people with no specific statistical training. This includes the use of a simple language, the clarification of the methods used and a clear explanation of the result implications.
To improve understanding, data visualisation plays a crucial role. Well-designed diagrams and tables can summarise big volumes of information, highlighting main trends and relations between variables. For instance, a line chart could show the time trend of a clinical parameter, whereas a histogram could point out the distribution of adverse events.
Moreover, it is essential to document in details all steps of the analytical process, from data collection to the final interpretation. This ensures result repeatability.
The integration of statistics in the drafting and execution of a PMCF is fundamental to ensure effective and compliant monitoring of medical devices. With a data-based approach, manufacturers can not only fulfil regulatory requirements, but also improve their products’ safety and quality.
Entrusting experts such as Di Renzo Regulatory Affairs can make the difference, ensuring the drafting of Post-Market Clinical Follow-Up (PMCF) Plan in compliance with the regulatory requirements. Contact us and find out how we can help you.
Di Renzo Regulatory Affairs has a unit made of highly qualified technical personnel, with a long-standing experience in this sector.
The main services we offer include:
- Feasibility studies and Regulatory Intelligence for medical devices
- Advice on Technical Dossiers and CE-marking
- Advice during the marketing and post-market phase
- Quality Management System according to the ISO 13485
- Free Sale Certificates
- Medical Device Promotion
- Clinical trials on medical devices
- Post-market surveillance
- Assumption of the role of Person Responsible for Regulatory Compliance (PRRC)
- Assumption of the role of UK Responsible Person (UK RP)
- Registration into the database of the Ministry of Health and of the main European Competent Authorities
- Registration into the EUDAMED database
- Regulatory updating service
