There is a shopping centre in England where, somewhere between a coffee chain and a phone shop, you can have your blood taken, your heart traced, your lungs measured and an ultrasound done, and be walking back to the car with results already moving toward a decision before your parking runs out. No hospital gown. No hospital car park. No hospital at all. This is a community diagnostic centre, and if you want to see where UK diagnostics is heading, you should go and stand in one for an afternoon.
The scale left the pilot stage a long way behind. Community diagnostic centres are an NHS England programme built to move tests and scans out of the acute hospital and into ordinary high street and retail settings, and by April 2026 around 170 of them were operating, in shopping centres, on university campuses, inside football stadiums. Since the first opened in July 2021 they had delivered 14.7 million tests, checks and scans by the end of March 2025, according to NHS England, which remains the last published cumulative total. In 2024 to 2025 that work included over 96,000 point-of-care tests, run on site while the samples that belong in a laboratory still travelled to one. The commitment did not stop there. In April 2026 the government announced a further £237 million for 36 centres, four of them new, on top of the £600 million of diagnostic capital already committed for 2025 to 2026 as part of £6 billion over five years. Across the whole NHS in England, 2025 brought a record of almost 30 million key diagnostic tests. That is a structural bet on decentralised diagnostics, made with committed capital rather than a press release.
My view is a settled one. Community diagnostic centres are the clearest chance we have to prove that decentralised testing and rigorous quality are not opposites. They sit close to the patient, which is where testing wants to be, and they sit inside pathology governance, which is where testing needs to be. Point-of-care testing has a natural home in them, done at scale, with real quality management, owned by people who understand what a result costs to get wrong. If the programme treats point-of-care testing as a serious, quality-managed modality rather than a convenience bolted onto a waiting room, it can model what good decentralised testing looks like for everyone else. If it treats it as a shortcut, it will teach the opposite lesson just as loudly.
What a community diagnostic centre actually is
Strip away the branding and a community diagnostic centre is a simple, powerful idea: take the diagnostic work that does not need to happen inside an acute hospital, and move it somewhere the public can actually reach. Phlebotomy, imaging, cardiology and respiratory tests, and a defined menu of point-of-care assays, delivered in a setting people pass on their way to the supermarket rather than one they dread and delay. These are not small operations. NHS England's own planning standards expect a standard centre to run at least 50,000 tests a year and a large one 60,000, and access is widening in practice, not just on paper: by April 2026, 108 CDCs were open 12 hours a day, seven days a week.
This is no longer a standalone programme either. The government's 10 Year Health Plan for England, published in July 2025, is built on three shifts, and the first of them, moving care from hospital to community, is exactly what a CDC does. The same plan promises a Neighbourhood Health Service delivered through Neighbourhood Health Centres, 120 of them by 2030 and 250 by 2035, that will pull diagnostics and testing even closer to home. NHS England has committed to map existing CDC capacity and planned growth over the next three years and to review GP direct access to diagnostics as part of that work. In other words, the pattern set inside CDCs now is about to propagate into hundreds of neighbourhood settings, which makes getting it right more urgent, not less.
The purpose is threefold, and each part matters. The first is speed and access: a test close to home, at a convenient time, without the friction that makes people put off the appointment that would have caught the thing early. In one example reported by NHS England, the Oldham centre cut the time to a lung cancer diagnosis from 42 days to under 19. The second is relieving the acute hospital: every straightforward scan or blood test done in the community is a slot freed in a pressured hospital department for the work that genuinely needs to be there. The third is catching disease earlier, when it is cheaper, kinder and more survivable to treat. These are not competing goals. A centre that does all three at once is doing exactly what the programme was designed to do.
The easily overlooked word in that description is pathology, and it carries most of the weight. A community diagnostic centre is not a freestanding kiosk answering to nobody. It connects to central laboratories for sample processing, and it operates within a pathology governance framework. That connection is the reason a CDC is not just another place a test happens. It is a place testing outside the main laboratory can be done with laboratory-grade governance, because the laboratory is in the room, at least in spirit and increasingly in system. Hold that thought, because it is the whole argument, and later I will be honest about where it does not yet hold.
Where point-of-care testing fits, and where it does not
Point-of-care testing earns its place in a community diagnostic centre for one reason above all others: it collapses the loop. A conventional pathway sends a sample away, waits, and calls the patient back, or worse, loses them in the gap. A well-chosen point-of-care test lets a patient be sampled, resulted, assessed and advised in a single visit. For the anxious, the elderly, the working parent who cannot take a second morning off, that single visit is not a convenience. It is the difference between a diagnosis made and a diagnosis missed.
This is not abstract. NHS England's planning guidance names the point-of-care menu a centre should be ready to run: eGFR, pro-BNP as a marker used to triage possible heart failure, anticoagulation dose changes measured by INR, C-reactive protein, an hCG pregnancy test and urinalysis, with anything more complex taken on site but processed through existing pathology routes. It is a deliberately narrow list, and every item on it earns its place by clinical logic rather than by what a bench analyser could technically run.
Take C-reactive protein at the point of care. NICE supports using it to guide antibiotic decisions for a patient with a lower respiratory tract infection when pneumonia is not clearly diagnosed, treating a result above 100 mg/L as a reason to offer antibiotics and a low result as a reason to hold off. That is a result that must change the next few minutes of care, run in modest numbers, on a robust method, and it doubles as a piece of antibiotic stewardship. It is close to a textbook fit for immediate turnaround.
But not everything belongs at the point of care, and pretending otherwise is how good programmes get a bad name. The honest question is not "can we test this here" but "should we". The decision turns on three things: clinical urgency, volume and complexity. A result that must change the next few minutes of care, run in modest numbers, on an assay that is robust in non-expert hands, is a strong candidate for immediate turnaround. A result that can wait a day, runs in high analytical volume, or demands the analytical range, calibration discipline and multiplex breadth of a core analyser, belongs on the sample transport to the central laboratory. A centre that chooses well on this axis gets the best of both worlds. A centre that chooses badly ends up with a bench analyser doing work it was never built for, or a patient waiting three days for a result they could have had in their hand.
There is a subtler line to hold too, and it is one a case-finding setting makes easy to cross. Intended use matters as much as the analyte. Some point-of-care assays are trusted for monitoring a known condition but are not, on their own, a safe way to make a first diagnosis unless the method has been validated to laboratory standard under full point-of-care governance. HbA1c is the classic case. In a centre built to find disease early, a fast and convenient result can quietly become a diagnosis it was never meant to carry, so some point-of-care results should trigger a laboratory-confirmed answer rather than stand as the answer themselves.
The question is never "can we test this at the point of care". It is "should we". Get that decision right and the CDC becomes the best of both worlds. Get it wrong and it becomes neither.
Done this way, point-of-care testing in a CDC is not a lesser version of laboratory testing. It is a deliberate division of labour, where the immediate answer is delivered immediately and the complex answer is delivered properly, and the patient experiences one joined-up service rather than two disconnected ones.
Retiring a false trade-off
For years the argument about point-of-care testing has been framed as a trade-off. You can have testing close to the patient, the story goes, or you can have laboratory-grade quality, but not both, because the moment you leave the accredited laboratory you leave the quality system behind. That framing has done real harm. It has made some laboratories suspicious of decentralisation and made some enthusiasts careless about governance, and it is wrong. This is the same false choice I have written about in Pharmacy First and the POCT quality gap, where testing sits further from pathology and the safety net is thinner.
The community diagnostic centre is where that false choice can finally be retired. Because pathology is involved from the start, a CDC has something a high street pharmacy or a care home running its own device does not automatically have: a direct line to the discipline, the standards and the people who do quality for a living. The current international standard for medical laboratories, ISO 15189:2022, has already absorbed the point-of-care requirements that used to sit in the separate ISO 22870, which has now been withdrawn. As UKAS sets out, accredited point-of-care providers are assessed against ISO 15189:2022, and the transition window for laboratories to move across closed on 6 December 2025, after which accreditation still held under the old 2012 edition is suspended. So this is the operating standard, not a future one. The standard says, in effect, that testing is testing and the place it happens does not change what a good result requires. A CDC is the ideal proving ground for that principle, because it has both the decentralised setting and the laboratory relationship in one place.
There is an honest qualification to make here, and making it strengthens the argument rather than weakening it. That governance is a target with a lag, not a guarantee present on day one. Under the programme's own standards a centre is expected to reach UKAS accreditation to ISO 15189:2022 for each of its modalities within two years of becoming fully operational, which means some sites are running tests, point-of-care included, inside a service that is still working toward the badge. And not every centre has an NHS laboratory literally down the corridor. A share of CDCs are run by independent-sector operators, and there the pathology relationship has to be written into the contract rather than assumed from the estate. Where the laboratory is not in the building, someone accredited still has to own the quality. Neither point is an argument against CDCs. Both are the reason the quality scaffolding, the quality control, the competency and the connectivity, has to be built and running from the first day, well before the assessor arrives.
And there is room to prove it. For all the headlines, point-of-care testing is still a tiny fraction of what a CDC does. Those 96,000 point-of-care tests in 2024 to 2025 sit against roughly 7.2 million CDC tests and scans in a comparable year, about 1.3 percent of the total, or one test in seventy-five. Most of the rest is imaging and physiological measurement, MRI, CT, ultrasound, echocardiography and spirometry, work that will never be point-of-care, so the realistic headroom sits inside the pathology and near-patient slice rather than the whole 7.2 million. Even allowing for that, the centres have barely started to use point-of-care testing, which means the pattern they set now, good or bad, is the one that scales. The pathology profession says the same. The Royal College of Pathologists, in its own 2025 update on the programme, points to significant opportunity to integrate more rapid point-of-care testing into CDCs, backed by direct connections to central laboratories.
If the programme gets this right, the benefit reaches far beyond the centres themselves. A CDC that runs point-of-care testing with genuine quality control, real competency assessment, external quality assessment and clean connectivity becomes a working demonstration that decentralised does not mean degraded. Pharmacies, care homes, ambulances, neighbourhood health centres and clinics watching it can copy the pattern. The CDC becomes not just a place that does testing, but a place that teaches how testing outside the laboratory should be done. That is a bigger prize than any throughput figure, and it is entirely within reach.
What it actually requires
Optimism is not a plan, so let me be concrete about what "done properly" demands. None of it is exotic. Every piece carries weight, and the temptation under throughput pressure will be to treat these as optional. They are not.
Connectivity to the wider record
A result that lands only on a printout, or only on the analyser's own screen, is a result that will be repeated, lost or ignored. I have watched a busy clinic hand-key a glucose off a meter into a system and transpose two digits without anyone noticing until the number made no sense against the patient. In a CDC, where the GP, the referring clinician and the host pathology service all have a stake, a point-of-care result must flow into the wider record automatically, attributed to the right patient, with its reference range and its provenance intact. That flow is not magic. It rests on point-of-care connectivity standards such as CLSI POCT1-A2 and HL7, and in practice on a point-of-care data-management layer sitting between the devices and the record. The genuinely hard part is not sending a number, it is matching it to the right patient through a reliable identity feed, which is the failure I unpack in Same patient, different number. NHS England's pathology guidance is blunt about the bar: the level of IT connectivity should be equal to that of the main laboratory and not dependent on the care setting. Manual re-keying is not a workaround. It is a source of error and a governance failure waiting to be found.
Competency for CDC staff
The people running tests in a community diagnostic centre may not be biomedical scientists, and that is fine, provided the competency system treats it honestly. Every operator should be trained against the instructions for use, formally assessed as competent, and reassessed on a schedule. Sample technique in particular, the part that hides its errors inside a plausible-looking number, must be taught, watched and confirmed rather than assumed. A finger squeezed too hard, a drop taken too soon, a cartridge run past its temperature window: none of it shows up in the figure the device reports, and all of it can move a result across a decision threshold. This is the slow-burn risk I called out in The competency time bomb. A CDC has an advantage here, because the host pathology service already runs competency frameworks for its own staff and can extend them rather than inventing quality from nothing. But there is a limit worth naming. That extension assumes the host laboratory has spare capacity to do it, and biomedical scientist time is one of the scarcer resources in the system. Competency assurance is a recurring staff cost, not a one-off, so it has to be resourced and named, often as a dedicated point-of-care coordinator, rather than quietly absorbed.
Quality control and EQA
Internal quality control at a defined frequency, checked against acceptance rules, with a hard rule that stops patient testing when a control is out, is the non-negotiable spine of trustworthy testing. So is enrolment in external quality assessment through a scheme such as UK NEQAS or WEQAS, which sends blind samples and compares performance against peers, exposing the blind spots a service cannot see in itself. This is now concrete for point-of-care work: UK NEQAS runs a dedicated point-of-care CRP scheme aimed at antibiotic stewardship, alongside glucose, HbA1c, lipids and urinalysis, and both UK NEQAS and WEQAS are accredited proficiency-testing providers. And a good score is not the finish line. As I argued in The EQA blind spot, a passing return can still sit alongside a result that would harm a patient, so EQA has to be read, not just filed. A CDC connected to a laboratory has no excuse to skip any of this, and every reason to run it at full laboratory standard. This is precisely where the pathology relationship pays for itself.
Clear ownership between the CDC and the host pathology service
The failure I would watch for most closely is not technical. It is the accountability gap. When something goes wrong with a point-of-care result, who owns the fix? Who signs off the method? Who investigates a failed EQA return? Who decides an analyte should move from the bench to the central laboratory, or the reverse? And one question that is easy to forget in the analytical detail: who acts on an abnormal result when the patient has already walked out of the building, and who safety-nets them? Analytical ownership decides whether the number is right. Clinical ownership decides whether anyone does the right thing with it. If those questions do not have named answers agreed in advance between the CDC and its host pathology service, they will be answered badly in a crisis, or not at all. The Royal College of Pathologists has warned that under-resourced, single-handed point-of-care services without clear clinical leadership are a patient-safety risk that can turn into a postcode lottery in care. Clear ownership, written down before it is needed, is what turns two organisations into one accountable service.
What this means for your service
If you run, commission or support a community diagnostic centre, or a service that will feed one, this is not a warning to slow down. It is a checklist for building the quality in from the start, while it is cheap, rather than retrofitting it under audit pressure later, when it is not. In the order I would take them:
- Decide the point-of-care menu on clinical logic, not vendor enthusiasm. Start from the nationally expected list, score each candidate test on urgency, volume and complexity, and be willing to send some things to the central laboratory even when a bench device could technically run them. Keep the diagnose-versus-monitor line clear. The right menu is a decision, not an accident.
- Wire results into the record before you go live. Insist that every point-of-care result flows electronically to the wider record, correctly attributed through a reliable identity feed, with reference ranges and provenance. Treat any manual re-keying step as a defect to be designed out, not tolerated.
- Extend the host laboratory's competency framework to CDC staff, and resource it. Do not build a parallel, weaker system. Train against the instructions for use, assess and sign off every operator, put sample technique under direct observation, and set the reassessment date on day one. Fund the coordinator time it takes.
- Run internal QC with real stop rules, and enrol in EQA for every analyte a scheme covers. Where no scheme exists, document an alternative comparison. Record controls, act on failures by halting testing, and treat a poor EQA return as an investigation, not a filing task. This is the cheapest way to discover a problem before a patient does.
- Write the ownership map down, and name the accredited partner. Name, in a shared document, who owns method sign-off, QC failure response, EQA investigation, clinical safety-netting of abnormal results, and the decision to move an analyte between the point of care and the central laboratory. Where an independent-sector operator runs the centre, name the accredited pathology partner responsible for point-of-care governance rather than assuming the estate delivers it. Rehearse the escalation route so it exists in practice, not just on paper.
If you are standing this up from scratch, my longer walk-through in Building a POCT service from the ground up takes these steps in order, and the free POCT Fundamentals course covers exactly this ground, quality control, competency, sample technique and escalation, in plain language for a non-laboratory setting. Our wider training goes deeper for teams taking on more analytes. When you are ready to formalise the arrangement between a CDC and its host pathology service, our consultancy helps you build a proportionate quality system and a clean ownership map between the two organisations. And when you are choosing which assays belong at the point of care, our analyte reference is a practical place to weigh urgency, volume and complexity test by test.
More than moving diagnostics closer to home
Community diagnostic centres are one of the better ideas in UK healthcare in a decade. They take diagnostics out of the hospital and put it within reach of the people who need it, and they do it at a scale that is already real and still growing. With the 10 Year Health Plan pushing care from hospital to community and a wave of neighbourhood health centres coming behind them, the stakes are only rising, because whatever CDCs prove about decentralised testing is about to be copied many times over. The opportunity in front of the programme is larger than the tests it runs. It is to show the whole system that testing can be close to the patient and rigorously governed at the same time, that decentralised and trustworthy are not enemies but partners. Point-of-care testing, done properly, is how a CDC proves it. Do that, and the centres will not just move diagnostics closer to home. They will model what good looks like for everyone else who is trying to.
Sources and notes
The figures in this article are drawn from published NHS England, GOV.UK, Royal College of Pathologists, UKAS and NICE data and from the relevant international standard. The 170 operating sites, the £237 million for 36 centres, the 108 centres open 12 hours a day and the record of almost 30 million NHS diagnostic tests in 2025 come from the April 2026 GOV.UK announcement; the 14.7 million cumulative CDC total is to end March 2025 and remains the last published cumulative figure; the 96,000 point-of-care tests are for 2024 to 2025. The almost 30 million figure is total NHS England diagnostic activity, not CDC-only and not the same as the CDC totals. The point-of-care share in Figure 3 compares the 96,000 point-of-care tests with about 7.2 million total CDC tests and scans reported for the twelve months to June 2025, periods that overlap but are not identical, so the resulting share is a close approximation rather than an exact ratio. Figures 1 and 3 plot published totals; any line drawn between marked points is indicative. Figure 2 is illustrative decision logic, not a fixed list of tests.
- GOV.UK. NHS patients to get quicker tests and scans closer to home. 13 April 2026: 170 CDCs in operation, a £237 million investment in 36 centres (four new in Gorton, Luton, Boston and Bideford), 108 CDCs open 12 hours a day seven days a week, and a record of almost 30 million NHS diagnostic tests in 2025.
- NHS England. Community diagnostic centres: guidance for planning, design and implementation. The named point-of-care menu a CDC should provide (eGFR, pro-BNP, anticoagulation change of dose, C-reactive protein, hCG pregnancy test, urinalysis), minimum annual activity by site type, and processing other samples through existing pathology routes.
- Royal College of Pathologists. An update from England's Community Diagnostic Centre Programme. 17 July 2025: 96,000 point-of-care tests in 2024 to 2025, the requirement to reach UKAS ISO 15189:2022 accreditation for each modality within two years of full go-live, and the opportunity to integrate more rapid point-of-care testing.
- GOV.UK. 10 Year Health Plan for England: fit for the future. 3 July 2025. The three shifts, including hospital to community, and the Neighbourhood Health Service.
- GOV.UK. Neighbourhood Health Framework. 3 December 2025: 250 neighbourhood health centres by 2035 with 120 by 2030, the commitment to map CDC capacity and planned increases over three years, and the review of GP direct access to diagnostics.
- UKAS. Point of Care Testing (POCT) accreditation, and the publication of ISO 15189:2022. Point-of-care testing assessed under ISO 15189:2022, incorporation of the ISO 22870 requirements, and the 6 December 2025 transition deadline.
- International Organization for Standardization. ISO 15189:2022, Medical laboratories, requirements for quality and competence, which now brings point-of-care testing inside the laboratory quality standard.
- Royal College of Pathologists. Bringing diagnosis closer to patients. Central pathology laboratories must have oversight of point-of-care testing, results and POCT must be UKAS accredited, and IT connectivity should be equal to that of the main laboratory and not dependent on the care setting. See also the College's Point of care testing: a national picture (17 July 2025) on the patient-safety risk of under-resourced, single-handed services.
- NICE. Pneumonia (acute): diagnosis and management, NG237. Point-of-care C-reactive protein testing to guide antibiotic decisions in lower respiratory tract infection, with a level above 100 mg/L prompting antibiotics.
- UK NEQAS. Point of care testing (POCT) external quality assessment, including the point-of-care CRP scheme, and WEQAS POCT EQA services.