Three Clocks, One Scan: Reading NHS Diagnostic Imaging Queues in 2026

A waiting list is an unusually persuasive number. It is precise, it is published monthly, and it appears to answer a simple question: how many people are waiting for a scan? But the figure that dominates headlines about NHS diagnostic imaging is better understood as the residue of several different processes. It counts requests, not symptoms; tests, not patients; and moments in a queue, not the full interval between a GP’s concern and a consultant’s decision. Reading it well requires separating at least three clocks that run at once.

The first clock: the stock of outstanding requests

NHS England’s monthly diagnostic activity statistics—commonly known by the name of the return, DM01—record the number of diagnostic tests and procedures waiting at a point in time. The total is often described as a waiting list, but it behaves more like a stock than a flow. A fall in the stock can mean patients were seen faster. It can also mean referrals were redirected, capacity was moved, or a provider reclassified how it records a test. A rise can reflect new demand, new capacity that has attracted referrals, or simply better recording.

The aggregate figure also mixes modalities that are not clinically interchangeable. MRI, CT, ultrasound, endoscopy, DEXA scanning and nuclear medicine have different indications, different equipment constraints and different staffing models. A CT scan requested to investigate an acute symptom and a DEXA scan requested to assess bone density may both sit in the same national total, but they do not carry the same clinical urgency or the same consequence of delay. When a headline says the imaging queue has grown, the first useful question is: which modalities, and in which settings?

The count is also of tests rather than people. One patient may have several requests open at once; another may have a single request that is counted, cancelled, rebooked and counted again. The unit of measurement matters because it determines whether a change is a change in demand, a change in workflow, or a change in bookkeeping.

The second clock: the referral route

The route by which a scan is requested shapes both its priority and its visibility in the data. A GP referral, a consultant-to-consultant request, an urgent suspected cancer pathway referral and a request generated inside an emergency department do not enter the same queue in the same way. Some are triaged against standardised urgency criteria; some are booked directly into slots; some are routed to a community diagnostic centre rather than a hospital imaging department.

Community diagnostic centres (CDCs) are a useful case study in why the headline figure can mislead. CDCs were designed to separate elective diagnostics from acute hospital pressures, often offering MRI, CT, ultrasound and endoscopy closer to where patients live. When a test moves from a hospital scanner to a CDC, the total number of tests may not change, but the recorded waiting time, the provider code and the apparent pressure on the hospital can all shift. A shorter queue at one site may therefore reflect redistribution rather than a faster diagnosis for the population as a whole.

Referral route also affects what the data can say about urgency. The DM01 return does not, on its own, tell a reader whether the patients waiting are on a cancer pathway, whether they are symptomatic, or whether they are being monitored for a known condition. Those distinctions sit in other datasets—cancer waiting times, referral-to-treatment pathways, and local triage records—and they do not always join neatly to the imaging count.

The third clock: the interval that matters

The most consequential clock is not the length of the queue but the interval between referral and result. That interval includes the time taken to triage the request, the wait for an appointment, the wait for the scan itself, the reporting step, and the communication of the result to the patient and the referring clinician. Only some of these stages appear in the published waiting-list figure.

A queue of 1,000 patients that is cleared in six weeks may be less clinically troubling than a queue of 300 in which a subset has waited many months for a report that has been generated but not communicated. Median waits can conceal a long tail; averages can be pulled by a small number of complex cases. Where data are published as proportions seen within a standard—such as the operational standard that 99% of diagnostic patients should wait no more than six weeks—the percentage is only as informative as the denominator and the exclusions behind it.

The interval also matters because delay is not neutral. For some conditions, a scan is a rule-out step that relieves anxiety; for others, it is a staging step that determines treatment. The same number of weeks can mean different things on different pathways. That is why the gap between referral and result is a better candidate for public attention than the raw count, even though it is harder to measure and harder to headline.

What the data cannot see

Imaging data are also silent on capacity in the round. A scanner is necessary but not sufficient: radiographers, radiologists, reporting software and appointment slots all constrain throughput. A trust may have spare scanner hours and no one to report the images; another may have reporting capacity but ageing equipment. The waiting list records the consequence, not the cause.

Nor does it capture the patients who never enter the formal queue—those managed in primary care without referral, those who decline a test, or those whose symptoms are investigated through other routes. The digitised record is a powerful instrument, but it is an instrument with a field of view.

Questions to ask of any headline

When a 2026 news report says the NHS imaging waiting list has risen or fallen, four questions are worth keeping in mind. Which modalities and providers are included? Is the figure a count of tests or patients? What has happened to referral routes and community diagnostic centre activity over the same period? And is there any accompanying measure of the time from referral to result, rather than the size of the queue at a single moment?

None of these questions diminishes the value of the data. They simply place it in its proper frame. A waiting list is not a diagnosis, and a shorter queue is not automatically a faster answer. The three clocks—stock, route and interval—run together, and only reading them together can turn a monthly statistic into a meaningful account of how quickly, and for whom, the NHS is turning a clinical question into a result.