Over the years, I’ve had countless conversations with researchers and professional services colleagues who thought records retention was simply about deciding when information can be deleted. In reality, those conversations almost always become discussions about trust, accountability, future value, and how organisations explain the decisions they make about data.<\/p>\n
That is why I think the phrase \u201cfrom archive to advantage\u201d<\/strong> is a helpful one. It invites us to think differently about retention. Not as an end-point, and not simply as a compliance requirement, but as part of the infrastructure that allows information to remain meaningful over time.<\/p>\n One of the most common misunderstandings about records retention is that it is mainly about how long something is kept. Good retention practice is not just about keeping information for longer or disposing of it sooner. It is about making careful decisions so that information remains manageable, accountable, and meaningful. A record that has been kept without context may not be very useful. A dataset without clear provenance, governance, or supporting documentation may be difficult to interpret later. Information that is technically retained, but practically impossible to find or understand, may not provide the value people assume it will.<\/p>\n Keeping data without context is a little like finding a box of old photographs in an attic. The photographs may have been preserved perfectly, but if nobody knows who the people are, when the photographs were taken, or why they matter, much of their value has been lost. Information can face the same problem. Retention preserves the record itself, but documentation, metadata, and governance preserve its meaning.<\/p>\n This is particularly important in research and healthcare contexts, where information may sit across different systems, projects, teams, and purposes. Patient data can have immediate operational value, longer-term research value, evidential value, and institutional value. Those values do not always sit neatly in separate boxes.<\/p>\n Policy can make retention look deceptively straightforward. A document may appear to say: keep this type of record for this length of time, then dispose of it. But the lived reality is often more complex. If you\u2019ve been to any of my training sessions you\u2019ll know one of my key catchphrases is \u201crecords management is not an exact science\u201d.<\/p>\n People may not always be sure whether they are dealing with a formal record, working material, research data, metadata, documentation, correspondence, consent information, or something that supports interpretation of the main dataset. They may be unsure whether something is still active, whether it has become part of an archive, whether it falls under research governance, information governance, records management, data protection, or several of these at once.<\/p>\n They may also be working across systems that were not designed with long-term records management in mind. Files may move from one platform to another. Projects may close while data still has potential value. Staff may leave. Naming conventions, folder structures, access permissions, and contextual knowledge may all change over time. Policies are essential. They provide the framework for consistent decision-making. They help organisations explain what they have done and why. They support accountability, transparency, and compliance. A retention schedule can tell us the expected retention period for a category of records. But it won\u2019t tell us everything we need to know about a mixed dataset, a long-running research collaboration, documentation that explains how data was created, or material that has value because it provides evidence of decisions made along the way. When we are talking about patient data, retention is not only an institutional issue. It also carries a human dimension. That does not mean patient data should never be retained. Quite the opposite. There may be strong reasons to retain information, including accountability, reproducibility, audit, future research, public benefit, or the need to understand decisions made in the past.<\/p>\n But retention needs to be deliberate. It should be possible to explain not only that information has been kept, but why it has been kept, how it is governed, who can access it, and what conditions apply to its future use.<\/p>\n If we can\u2019t describe what has been retained, where it is, why it is there, and what it can be used for, then we risk losing trust in the systems that support research and innovation. Conversely, when retention is handled well, it can strengthen confidence. It shows that data is not being treated as an unmanaged by-product, but as something requiring care, context, and responsibility.<\/p>\n Records management is often associated with rules, limits, disposal, and risk, but well-managed records can also be enabling. They help people find what they need when they need it.<\/p>\n In digital healthcare innovation, this really matters. Data may be collected or generated for one purpose, then later become relevant to new questions, new methods, or new collaborations. But future value is not automatic. Data does not remain useful simply because it has been stored somewhere.<\/p>\n Imagine a research project that concludes after several years. The dataset itself has been retained, but the supporting documentation explaining how variables were defined, what exclusions were applied, and why particular methodological decisions were made has not. Future researchers may still be able to access the data, but understanding, validating, or reusing it becomes significantly more challenging. In this sense, retaining context can be just as important as retaining the data itself.<\/p>\n To remain useful, it needs context. It needs documentation. It needs governance. It needs clarity about permissions, rights, responsibilities, and limits. It needs to be managed in a way that allows people to understand what it is, where it came from, what has happened to it, and what can responsibly happen next.<\/p>\n This is where retention and records management connect directly to research value. They are not separate from innovation. They are part of what makes sustainable innovation possible.<\/p>\n The word “archive” can sometimes make information sound static, as though it has reached the end of its useful life. Thinking clearly about those distinctions helps ensure that archived information remains understandable, managed appropriately, and capable of supporting future needs.<\/p>\n One of the less visible parts of records management is that it often starts with conversation rather than instruction. My role is not to simply say \u201ckeep this\u201d or \u201cdelete that\u201d. I try to help people think through what they have, what it represents, what obligations apply, what value it may hold, and what decisions need to be documented. I also try to help connect people with the right colleagues where the question crosses into data protection, research governance, information security, archiving, or data management.<\/p>\n At its best, records management is a bridge between policy and practice. It helps translate organisational requirements into workable decisions. It also helps feed practical experience back into better guidance, clearer processes, and more realistic expectations.<\/p>\n One practical lesson I want to emphasise is that retention questions are much easier to deal with early on in any process, as it\u2019s always harder to reconstruct context after a project has ended, staff have moved on, systems have changed, or decisions were made but not recorded. It is harder to determine the value of information when the people who understand it are no longer available. It is harder to manage risk when no one is quite sure what exists, where it sits, or who is responsible for it. It can also help researchers feel more confident. Rather than seeing retention as something that appears at the end, or only when there is a problem, it becomes part of good project planning and responsible data stewardship.<\/p>\n Institutions need to be able to explain and account for their decisions. Researchers need well-managed information so it\u2019s easier to understand, trust, and reuse. Patient and public trust depends on handling personal data responsibly. Patient data retention sits at the intersection of systems, rules, responsibilities, and real people. That is why it can feel complicated. But it is also why it matters.<\/p>\n If we think about retention only as storage or disposal, we miss the bigger picture. We miss its role in preserving context, supporting accountability, and enabling future value. We also miss the opportunity to build better connections between the people who create, use, govern, preserve, and explain data.<\/p>\n Records management may not always be the most visible part of the healthcare data landscape, but it\u2019s a key part of the structure that helps make that landscape navigable.<\/p>\n If you are unsure where records management fits into your work, or if you are facing a retention question and are not sure where to start, reaching out early is almost always worthwhile: complianceandrisk@cardiff.ac.uk<\/p>\n Good retention is rarely about keeping information forever. It is about ensuring that, when the future asks questions of the past, we still have the evidence, context, and trust needed to answer them.<\/p>\nRetention is not just keeping things<\/h2>\n
\nThat is certainly part of it. Retention schedules, disposal dates, review periods, and legal or regulatory requirements all matter. But in practice, the more important question is often not simply \u201chow long?\u201d, but \u201cwhy, in what form, with what context, and for whose future use?\u201d<\/p>\nWhy it feels more complicated in real life<\/h2>\n
\nThat is where retention becomes less about ticking a box and more about judgement, conversation, and stewardship.<\/p>\nPolicy matters, but it does not remove judgement<\/h2>\n
\nBut policies do not remove the need for human judgement. In fact, the more complex the information environment, the more important that judgement becomes.<\/p>\n
\nThis is not a failure of policy. It is the reality of applying policy to real situations.<\/p>\nPatients need to trust us<\/h2>\n
\nPatient data exists because people have interacted with healthcare systems, taken part in research, agreed to certain uses of information, or had information generated through their care. Even where data is de-identified, anonymised, pseudonymised, or otherwise controlled, there is still a responsibility to manage it carefully and be able to explain how and why decisions are made.<\/p>\nRetention can enable reuse and innovation<\/h2>\n
The archive is not the end of the story<\/h2>\n
\nSometimes that is true. Some information is retained because it provides evidence of what happened, rather than because it will be actively reused. In other cases, however, archiving marks the beginning of a longer period of stewardship.
\nOne challenge is that the word “archive” means different things to different people. For some, it means long-term preservation. For others, it means moving something out of active use. For others, it simply means “put it somewhere safe”. Those different interpretations can create confusion about roles, responsibilities, and future expectations.
\nThat is why it helps to be specific. Are we preserving records for accountability? Maintaining research data for potential reuse? Keeping documentation that explains a dataset? Or storing information that is no longer active but may still be required? Each purpose brings its own considerations and governance requirements.<\/p>\nThe human side of records management<\/h2>\n
\nPeople don\u2019t usually come to me with abstract questions. They come with practical uncertainty. They might be setting up a project, closing one down, moving data between systems, trying to understand what can be reused, or wondering whether something can be deleted. They may already have looked for guidance and found that their situation does not fit neatly into one category.<\/p>\nWorking together earlier makes things easier<\/h2>\n
\nEarly conversations can prevent a lot of that difficulty, but that doesn\u2019t mean every project needs a long records management exercise. Often, a short discussion at the right moment can make a significant difference. It can clarify what needs to be documented, what should be retained, what can be disposed of in due course, and which other teams need to be involved.<\/p>\nTL;DR<\/h2>\n
\nSo the three key takeaways on this topic are:<\/p>\n\n
And finally\u2026<\/h2>\n