Clinical obligations are the underlying monitoring requirements and actions needed to provide safe caree They are clinically meaningful, relevant across large patient populations and often tied to preventable harm. But they sit outside the clean machinery of QOF. They are weakly incentivised, poorly owned and scattered across thousands of practices, each with its own workflows, staff, searches and local workarounds.
No single team clearly owns them, no national system reliably tracks them, no one can easily say which patients are safe, which obligations are overdue and which risks are quietly accumulating in the background.
So the work sits there.
Then it ages.
Then it rots.
That rot is clinical debt and The NHS is full of it.
Clinical debt is the accumulation of open monitoring obligations that should have been completed, reviewed, escalated or consciously deferred, but were not. It is not the same as a waiting list, or a dashboard showing that a patient has not attended, or a monthly search result that someone exports into Excel. It is unresolved clinical responsibility.
A patient taking methotrexate without the right monitoring. A woman with previous gestational diabetes who never receives the follow-up she should. A patient with coeliac disease who falls outside the annual review machinery. A bisphosphonate course that continues because nobody has reliable oversight of treatment duration. A frail patient whose blood pressure and medication risks are never considered together.
None of this is exotic medicine. The NHS knows what good care looks like and the guidance and evidence already exists.
The problem is execution.
Once an obligation is spread across multiple teams, weakly incentivised and difficult to track manually, failure becomes the default. It happens slowly, one patient at a time, until the debt is paid in the form of preventable harm.
The NHS tends to respond to this with more searches, more dashboards, more templates and more activity. But activity is not control. For example, a report can show that work exists, but it can't execute:
This distinction matters because the NHS has spent years confusing evidence of activity with evidence of execution.
Most primary care systems were built around recording events, not maintaining clinical state. They can tell you that a blood test happened. They struggle to tell you whether every relevant obligation has been satisfied, whether the result changed what should happen next and whether the entire pathway has reached a defensible endpoint. That leaves practices relying on memory, spreadsheets, searches and heroic staff.
This is why clinical debt is not mainly a workforce problem. More people may temporarily reduce the backlog, but they do not fix the underlying absence of control. You can hire another LTC administrator. You can create another spreadsheet. You can run the search more often, but this doesn't solve execution
If we continue calling this “recall”, “admin backlog” or “non-attendance”, we will continue building tools that send messages and generate lists. We will keep measuring contact attempts while missing the more important question:
Was the clinical obligation actually closed?
That question changes everything. It forces us to stop treating every patient as a row in a report and start treating each obligation as an object with a current state. Due. Completed. Deferred. Escalated. No longer relevant. Awaiting evidence. Requiring clinical review.
Once that state is maintained reliably, clinical debt becomes visible -> Once it is visible, it becomes measurable -> Once it is measurable, it can be reduced.
This is the part that should make people uncomfortable. Much of the NHS currently has no reliable way of knowing how much clinical debt it holds. Practices know they are busy. They know there are recalls outstanding. They know some patients are difficult to reach. But they often cannot quantify the unresolved obligations sitting across the population, because the underlying systems were never designed to represent them properly.
The debt still exists whether the dashboard can see it or not.
In fact, the most dangerous clinical debt is often the work that is not attached to a payment, a performance framework or an obvious operational owner. Those obligations decay quietly because nobody is punished for missing them today. The harm comes later and this is why clinical debt matters.
Population level health needs oversight and executon infrastructure that can maintain clinical obligations longitudinally, assign ownership, detect failure and drive work to closure. Until then, patients will continue to fall through the gaps that nobody knew even existed