Currently, information sharing about medical products in the health system is inefficient and leads to poor quality data. The Department for health and social care (DHSC) asked us to explore how we might make sharing medical device data between suppliers and consumers easier, more efficient and more consistent.
There are currently over 3,000,000 different medical devices on the UK market available for use by 215 NHS Trusts and many more healthcare organisation. These devices are registered and regulated by the Medicines and Healthcare products Regulatory Agency (MHRA). As part of their registration process, they collect a range of data and product characteristics about each device, some of which is currently published on their Public Access Registration Database (PARD).
The Department for Health and Social Care (DHSC) is responsible, more broadly, for ensuring high standards and effective delivery of these health services. In 2023, Their Medical Technology Strategy identified a gap in the DHSCs ability to understand products and their characteristics due to a lack of standardised information. It also highlighted an inability to join datasets together, limiting the potential for more advanced analysis and evidence gathering.
As such, The MedTech Data and Analysis Team within DHSC conducted a discovery phase to understand how they could solve these issues. The discovery recommended that they develop a national Product Information Management (PIM) system for ‘or collecting, managing, and exporting product information’ and to act ‘as a shared reference system for information’. The objective was to take a set of basic device data and make it available to users in a standardised format, with the goal of the PIM system to become a single shared reference point for basic medical device product information.
There are over 24,000+ different categories of medical devices - ranging from syringes and tweezers tests to MRI machines and pacemakers. These are produced by approximately 10,000 manufacturers based around the globe. There is, therefore a huge amount of data related to these devices, and the data needs can vary wildly depending on the time of device, or what role in the NHS the user is performing (e.g. procurement, device management, clinical work, etc.)
The discovery report had highlighted how, information sharing about medical products in the health system is inefficient and leads to poor quality data. We needed to explore how we might make sharing medical device data between suppliers and consumers easier, more efficient and more consistent.
Some of the challenges included:
Having identified and validated the concept of using a PIM to solve these issues, the DHSC asked us to help explore how this could be developed, and develop a plan, roadmap and business case for delivering the system.
We conducted a 12-week Alpha following the Government Service Standard, using Agile and user-centred design principles and processes.
Having analysed the Discovery report and Alpha brief, we identified our riskiest assumptions, that we wanted to validate or disprove, so we could build confidence in our final recommendations. These assumptions included:
To ensure we’d comprehensively identified the best solutions to problems from discovery we explored and tested a range of concepts, wireframes and prototypes with users with both data providers and data consumers.
These explorations Focused on key user journeys, such as ‘how data providers could get their data into the system and maintain it, and how data consumers can find device information, how they consume that information, and what that information actually needs to be – latterly, how they can help request or flag data that may be incorrect.In order to test our ideas. We produced a range of prototypes and technical proof of concepts. Over the course of the Alpha these were tested 3 iterations of a dynamic prototype, with a realistic database of device data and numerous examples of workflows, data flows and user journeys, so we could validate how well our ideas would meet users needs and fit into their existing workflows.
This research was conducted with 49 user research participants, covering a representative sample of both data providers and data consumers. We engaged them through collaborative workshops, user interviews and prototype testing. This was supported by two additional user surveys, to gather more quantitative data, and engagement with 26 further stakeholders from across similar projects, programmes and other organisation and teams with a vested interest in the service.
This work was supported by a comprehensive analysis of what other data, and data sources, we might be able to integrate with to bolster the PIM’s database. Our analysis included products such as MHRAs Public Access Registration Database (PARD), AccessGUDID, European Database on Medical Devices (EUDAMED), and many more. While there were lots of different catalogues providing similar, but slightly different, services. We wanted to understand if we could match and compile data across these datasets, to build out a more comprehensive overall dataset. This involved working closely with these product teams to better understand how we could work together to share data, as well as how the PIM might support their goals and objectives.
We found a number of challenges that we needed to overcome with our DHSC colleagues.
An initial workshop to gather some early feedback from data consumers found that the Initial data points that the discovery had highlighted for inclusion in the PIM would not be enough to meet users needs. Users reported that if they couldn’t get what I need in PIM, then I’ll still need to go somewhere else, and if that was the case, why would they come to PIM in the first place? We therefore needed to pivot our assumption, to identify what would would be a good minimum viable dataset. Having worked with DHSC and a range of stakeholders to explore a range of options, we settled on using a dataset currently being collected manually by a PDF form called the Pre-Acquisition template. This option validated that this was a dataset being actively used across the NHS, and could be completed by manufacturers details – this approach had a further benefit that, as it was currently being manually completed and shared via email, PIM could potentially provide additional value by at some point, incorporating that process into the system.
We also quickly identified an issue with the quality of data. The assumption that we could use our initial dataset as the basis for the PIM proved to be not correct, because although it had entries for all 3m devices, the actual data was missing or incorrect, due to it’s collection stage being largely optional for manufacturers. As such we explored a number of different datasources we could compile with that initial dataset, to link up additional fields. We also explored a range of methods for users (both manufacturers and NHS users) to either flag, request or update any data that was missing or incorrect.
Because there were so many different manufacturers (e.g. in terms of size, structure, number and types of devices), we found their ability and means of providing and maintaining their data was challenging. The types of data they collected about their device, how it was stored and how it could be shared, varied massively across different manufacturers, and even within a manufacturer's organisational structure, on many instances. In order to overcome this, we identified the need to provide a range of different options for how manufacturers could maintain their data, ranging from simple webforms or spreadsheet uploads, to API integrations – depending on the organisation's digital maturity.
Another challenge was to identify how manufacturers would be motivated to maintain their data in the service. This was because engagement with the service wouldn’t be able to be mandated, and due to the vast number of devices and manufacturers, some kind of central curation of the database by DHSC, wouldn’t be practical or viable. We therefore engaged with a range of manufacturers in collaborative workshops to better understand their current pain points, processes and motivations to use the system. This enabled so to identify what kinds of things to avoid, as well as what the most promising motivating factors could be. As they’re all running businesses, the service would need to clearly show how it would reduce time taken/effort/cost in responding to queries.
As well as a broad range of types of manufacturers, we also found a broad range of different users and use cases within NHS trusts. With so many different possibilities for what problems the service could solve and for whom, we needed to rigorously track who we’ve engaged with, to ensure we had a representative sample. We also sought to support our qualitative insights with more quantitative data via a couple of feedback surveys – with only 12 weeks, we could only speak to so many users, so we wanted to be able to cast a wider net to ensure we could cover everyone.
Despite these challenges, we tweaked our approach and identified an approach to take into development for Beta that met all of our success criteria. We developed a clear roadmap for delivery of our 8 sets of recommendations for delivering a successful. Our recommendations for Beta spanned the scope and flexibility required for PIM data, what integrations with other products and services would be required, how the service could be maintained and how to drive user adoption.
These recommendations were supported by a comprehensive business case, outlining:
Over the course of the project we also identified other workstreams in the medical device space that would benefit from our work on PIM, either as possible future integrations with the service, or reusing PIMs data – ensuring that PIMs development could provide an even broader set of benefits, than identified within the discovery.
Having confirmed the need for, value of, and approach to delivering a PIM for medical device data, the DHSC are currently going through the process of signing off the business case and procuring the Beta phase delivery.
Over the next year, the service will be developed. An early-stage version will be released to a limited number of manufacturers and NHS Trusts in a ‘Private beta’ – this will enable real users to use the real service, albeit in a controlled environment. This will enable the team to iron out any creases. The service will then be opened up to all 10,000 manufacturers and 250+ healthcare organisations into its ‘Public beta’ phase.
Testamonial
"TODO” Someone, Department for Health and Social Care
Whether you’re ready to start your project now or you just want to talk things through, we’d love to hear from you.
© 2025 Marvell Consulting • Privacy Policy • Social Value Statement • Registered Company Number: 06223331
