Postdoc Bayesian Methodology Predicting of Implantable Medical Devices

Unleashing the potential of Bayesian statistics to predict the performance of new implantable medical devices based on existing data while accounting for uncertainties

Innovative medical devices can offer important benefits to patients and are approved based on sufficient clinical evidence showing the safety and beneficial effects. Manufacturers currently face challenges in determining whether new clinical evidence is needed for a small change in the material or design, or that evidence is sufficient. In addition, regulators and notified bodies need to decide whether this requires a full (new) assessment of the evidence or can be considered an equivalent device.

The key question for both stakeholders is whether we may expect that the performance is similar to that of the parent design, or that performance is uncertain and we need further evidence, but are lacking tools to inform their decision making for such device iterations.

This postdoc position will delve into these complexitieis by developing a methodology for bridging studies, using Bayesian methodology to integrate prior knowledge (historical data, benchmark tests, and literature) about exisiting meical devices with real-world evidence to predict the impact of iterative design changes on the device performance, accounting for the inherent uncertainties around such predictions.

A major challenge in assessing the performance of a medical device variants is lack of data for the new device variants. The postdoc will develop a mechanism to make use of existing data, translate past success into current confidence while satisfying medical device regulation requirements. It will transform the current static models into a "Digital Twin" for safety of medical devices, specifically supporting decisions regarding iterative small changes.

Finally, the postdoc will evaluate the quality of evidence required after a device is on the market, and provides a roadmap for regulators to demand the right data at the right time, rather than merely more data.

The position is part of the European COMPASS4MD project consisitng of an international consortium and will involve close collaboration with research institutes, regulators, clinical experts, patient representatives, and other partners across Europe. You will contribute to joint project activities, exchange findings with partners from different countries and disciplines. The position therefore suits someone who communicates easily, enjoys working with people from different countries and disciplines, and is willing to take an active role in an international consortium.

You should be comfortable working independently on your own research while also contributing to shared activities, meetings, workshops, and deliverables with European project partners. Some travel within Europe may be part of the position.

English proficiency is required. Proficiency in Dutch is considered an advantage, but is not a requirement. Experience with advanced statistical methods, econometric methods, and Monte Carlo simulation is an advantage. We do not expect candidates to already possess expertise in these areas; we particularly value the ability and motivation to work across disciplinary boundaries.

A methodologically oriented PhD degree e.g. in Epidemiology, Health Sciences, (Technical) Medicine, Medical Statistics, Data Science, Safety Science, Biomedical Engineering or a related field. As the project combines quantitative analysis of clinical evidence with empirical research into regulatory decision-making, we particularly welcome candidates who demonstrate expertise in advanced quantitative methods to answer biomedical or clinical research questions or to tackle methodological problems…