Multiplex respiratory testing and the post-pandemic POCT legacy

Think back, if you can bear to, to the spring of 2020. A cardboard box arrives through the letterbox. Inside is a plastic cassette, a tiny buffer tube, a lancet or a swab, and a leaflet. A person with no clinical training swabs their own nose at the kitchen table, drips liquid into a well, waits fifteen minutes, and reads a line. On the strength of that line they decide whether to go to work, whether to visit an elderly parent, whether to sit their exams. Rarely, if ever, had a diagnostic test been placed into so many untrained hands so quickly, or been trusted to change everyday behaviour on such a scale.

We tend to remember the pandemic for the test itself: the lateral flow cassette, the two lines, the anxious fifteen minutes. That is the wrong thing to remember. The cassette was a means. What actually changed, and what has not changed back, is the expectation. A whole population learned that you can have a diagnostic answer in minutes, where you are, without a laboratory in the loop, and that you can act on it immediately. That expectation did not evaporate when the mandates ended. It hardened into the baseline. The public now assumes a fast answer is normal, and clinicians increasingly build their decisions around getting one.

My argument here has an edge to it. The lasting legacy of the pandemic is not the COVID test but the normalisation of the fast answer, and the most important expression of that legacy is the multiplex respiratory panel that tells you in one sample whether it is COVID, influenza or RSV. That capability is genuinely powerful, especially in urgent care in winter. But it raises the quality bar rather than lowering it, because we are moving more complex tests into less controlled settings, and a fast answer that is wrong is worse than a slower answer that is right. Getting the result quickly is the straightforward part. Making it worth trusting is the harder one.

From one line on a strip to four answers from one swab

The arc is worth tracing deliberately, because the two ends of it look deceptively similar and are profoundly different. At one end sits the rapid antigen test of the pandemic: a single-target lateral flow device, cheap, fast, forgiving, designed to be used by anyone. At the other end sits the syndromic multiplex molecular panel: a test that takes a single respiratory sample and simultaneously detects and differentiates SARS-CoV-2, influenza A, influenza B and RSV, and increasingly others besides. One swab in, several distinct answers out.

This is not a marginal upgrade. It reflects a clinical truth that the pandemic forced everyone to confront and then, oddly, made harder to ignore: in the middle of a British winter, a feverish, coughing, breathless patient in front of you could have any of several respiratory viruses, and they can present almost identically. UKHSA surveillance now tracks influenza, COVID-19 and RSV co-circulating through the same weeks of the season, which is precisely the situation a clinician has to untangle at the front door. Knowing which virus it is used to be an academic nicety answered days later by a laboratory, if at all. Now it can be answered at or near the point of care while the patient is still in the department.

The technological leap is that this differentiation now happens on molecular platforms compact and robust enough to sit near the patient rather than only in a central laboratory, and on the better molecular platforms it happens without giving up much analytical accuracy, though performance varies by platform, target and sample. A systematic review and meta-analysis of multiplex PCR platforms put the summary sensitivity for influenza A at around 91 percent with specificity above 98 percent, which is why the near-patient versions of the stronger assays are trusted to carry real clinical weight rather than being treated as a screening curiosity. That is the pandemic legacy made concrete. The population learned to expect fast answers; the diagnostics moved to meet the expectation; and what arrived was not just a faster COVID test but a fundamentally more informative one that answers the actual clinical question, which was never only "is it COVID" but "which of these is it, and what should I do next".

Why differentiation changes what happens next in winter

The reason clinicians reach for these panels is that a differentiated result does not merely satisfy curiosity. It changes management, and it changes it in ways that matter most when the department is full and the beds are scarce. We are not guessing about this. Two UK randomised controlled trials, ResPOC and FluPOC, both run largely in and around Southampton and both using the same multiplex platform, measured what actually happened when a rapid molecular result reached the clinician in hours rather than the next day.

In ResPOC, 720 adults presenting to hospital with acute respiratory illness were randomised to immediate point-of-care testing or routine care. The median time to a result fell from 37.1 hours in the laboratory arm to 2.3 hours at the point of care, and a virus was detected in 45 percent of the tested patients against just 15 percent in the routine-care arm. FluPOC then took the same idea and pointed it squarely at influenza, randomising 613 admitted adults across two winters. The results were stark, and they map directly onto the four decisions that actually matter.

Consider four decisions that a single undifferentiated "respiratory virus, probably" cannot inform, but a specific answer can.

Isolation and cohorting. Knowing whether the patient has influenza, RSV or SARS-CoV-2 tells the department how to place them, whom they can safely share a bay with, and which precautions apply. In FluPOC, 70 percent of patients with confirmed influenza in the testing arm were isolated to a side room, against 38 percent under routine care. In a hospital fighting for every side room in January, that is not administrative tidiness; it is infection control that protects the frailest patients on the ward.

Antiviral decisions. The antiviral you might reach for in confirmed influenza is not the one you reach for in COVID, and neither is appropriate for RSV. FluPOC found that 99 percent of the influenza patients in the testing arm received an appropriate antiviral, against 62 percent in the control arm, and ResPOC saw the same effect earlier, with 91 percent of influenza-positive patients treated with a neuraminidase inhibitor against 65 percent under routine care. A specific, timely result lets you start the right agent early, when early is when it works, or withhold it confidently when it is not indicated. A vague answer leaves you guessing.

Admission and discharge. The virus is only one input, but it is a real one. Here the honesty matters: ResPOC found the average length of stay fell by 1.1 days in the testing arm, a modest but genuine signal, while FluPOC found no difference in length of stay at all. So a differentiated result can sharpen the judgement about who genuinely needs a bed, but it is not a lever that automatically shortens admissions. Anyone selling it as one is overpromising, and the trial evidence will not back them up.

Reassurance and stewardship. A clear viral diagnosis can be the thing that lets you not prescribe an antibiotic, not admit, and not investigate further, with confidence rather than anxiety. Here too the evidence is nuanced rather than triumphant. ResPOC did not reduce the overall proportion of patients given antibiotics at all, but it did increase the number given only a single dose or a course of under 48 hours, from 3 percent to 10 percent for single doses. Diagnostic clarity helps you stop sooner, more than it helps you never start. That is still worth having, but it is a scalpel, not a floodgate.

The pandemic did not teach us to want a test. It taught us to want a fast answer, and then act on it. Multiplex panels are that lesson turned into an instrument.

Notice what all four decisions have in common: they are downstream of the number, not in it. The panel does not admit, isolate, prescribe or reassure. A clinician does, using the result as one input among several. That is the hinge on which the rest of this argument turns.

Complex tests, less controlled settings: the quality bar goes up, not down

Here is the uncomfortable arithmetic of progress. We are taking tests that are more complex than a lateral flow cassette, molecular assays with real sample-handling requirements and real failure modes, and we are moving them out of the central laboratory into emergency departments, urgent care centres, assessment units and beyond. More sophistication, less controlled environment, higher stakes. Every one of those vectors points the same way: the quality requirement rises.

This is counter-intuitive, and it trips people up. The instinct is to assume that because the platform is clever and the workflow is simple to run, the governance can be lighter. The opposite is true. The cleverer the platform and the closer to the patient it sits, the more the surrounding discipline has to compensate for the laboratory that is no longer in the room. A central laboratory has trained scientists, controlled storage, embedded quality control and a culture built around not releasing a bad number. An urgent care corridor at two in the morning has none of that by default. It has to be designed in.

The standards world has already made this explicit, and it is worth saying plainly. The current international standard for medical laboratories, ISO 15189:2022, brings point-of-care testing inside the laboratory quality standard itself, superseding the older, now withdrawn ISO 22870 that used to treat POCT as a thing apart. The message is unambiguous. Testing is testing. Where it happens does not change what a trustworthy result requires. A multiplex respiratory result generated near a patient in the emergency department is the same clinical object as one generated in the core laboratory, and it deserves the same scaffolding: competency, quality control, correct sampling and result governance. What holds up that scaffolding, in practice, is:

• Operator competency. Every person who runs the panel is trained against the instructions for use and formally assessed as competent, with reassessment over time. A near-patient molecular platform is easy to operate and unforgiving of the details, which is exactly the combination that produces confident error.
• Quality control with rules. Controls run at a defined frequency, checked against acceptance criteria, with a clear rule for what happens when a control fails. In a busy department the temptation to skip or to run and not read is real, and it is precisely what governance exists to prevent.
• Correct sampling. The nasopharyngeal or other respiratory sample is among the biggest determinants of whether the answer is right, alongside timing and the assay itself. A poorly collected swab can produce a falsely negative molecular result no matter how good the platform. Sampling technique is a taught, watched, confirmed skill, not an assumption.
• Result governance. The result has to reach the right record, the right clinician, and where relevant the right infection-control and surveillance processes. A brilliant answer that lives only on a screen for ninety seconds and then vanishes has governed nothing.

None of this is a reason to keep multiplex panels out of urgent care. It is the price of putting them there responsibly. The setting is right. The discipline has to travel with the test.

The honest caveats, said out loud

Any senior voice recommending these panels owes the reader the caveats, because the failure mode of a powerful, fast, differentiated result is over-trust. Three honesties matter.

First, a negative does not always mean no infection. A single negative result never fully excludes disease when the pre-test probability is high. Sample timing, sample quality, viral load and the stage of illness all bear on whether a truly infected patient tests negative. The clinician who overrides a strong clinical picture because a strip or a cartridge said negative is making the classic error the pandemic should have inoculated us against, and too often did not.

Second, pre-test probability still governs interpretation. A result is not a verdict handed down independent of the patient. It is evidence to be weighed against everything you already knew before you ran it. In a florid winter presentation, a negative is a weaker signal than it looks; in a low-prevalence, low-suspicion context, a positive deserves more scrutiny before it drives a big decision. Pre-test probability did not stop mattering when molecular platforms arrived at the bedside.

Third, antigen and molecular performance differ, and the difference is clinical. The Cochrane review of rapid antigen tests puts their pooled sensitivity against PCR at about 73 percent in people with symptoms, rising to roughly 81 percent in the first week of illness but falling to around 54 percent in the second week and to 55 percent in people with no symptoms. Specificity is high, around 99 percent, so in a symptomatic patient in season a positive is usually trustworthy, though in a very low-prevalence setting even a high-specificity positive deserves a second look. It is the negative that betrays you. A rapid antigen negative simply carries less reassurance than a negative from a well-run molecular platform, and it carries least exactly when illness is established or when the swab is taken late. Knowing which kind of test produced the answer in front of you is part of knowing what the answer is worth. Treating all "negatives" as equivalent is a category error that the fast-answer culture actively encourages.

A fast answer that is wrong is worse than a slow answer that is right, because a wrong answer that arrives quickly is the one you act on before anyone can stop you.

What this means for your service

If you run, commission or advise on near-patient respiratory testing, the opportunity is real and the direction is right. The work is to capture the clinical benefit without importing the over-trust. Practically, in the order I would tackle it:

• Decide what each result is for before you deploy. Map, for your setting, exactly which management decisions a differentiated result will change: isolation, antiviral, admission, reassurance. A test that changes no decision is a cost, not a capability.
• Build competency around sampling, not just the machine. The platform is the easy part. Train, observe and sign off respiratory sample collection explicitly, and reassess it, because that is where near-patient molecular results are most quietly lost.
• Run quality control with a stop rule. Define the frequency, record the results, check them against acceptance criteria, and above all fix the rule that patient testing pauses when a control fails. In a busy department, the stop rule is the whole point.
• Write the interpretation into the pathway. Put pre-test probability and the negative-does-not-exclude principle into the actual protocol, not just the induction talk. Make it explicit that a negative in a high-suspicion patient is not a discharge licence.
• Govern the result to a record and a clinician. Ensure every result reaches the patient record and the decision-maker, and connect it to infection-control and surveillance processes where relevant. A result that is not captured cannot be governed, audited or learned from.
• Know your assay class. Make sure staff understand whether they are running an antigen or a molecular test and what that means for how much a negative reassures. Bake that distinction into training rather than leaving it to individual memory.

If your team is standing this up from scratch, the free POCT Fundamentals course covers the ground that matters most here: quality control, competency, sampling and escalation for a non-laboratory setting. Our wider training goes deeper for departments taking on molecular and multiplex work, our analyte and test reference helps teams understand what a given result can and cannot tell them, and our consultancy helps services build a proportionate quality system around near-patient testing that protects the result without slowing the clinical benefit.

Keep the fast answer. Earn the right to trust it.

The pandemic gave us something the diagnostics industry had wanted for a generation: a public and a profession that expect answers fast and act on them at the point of need. The multiplex respiratory panel is the mature, clinically useful heir to that legacy, and in a winter emergency department it is a real asset. The trial evidence is genuinely encouraging on detection, the right antiviral and isolation, and honest about the things it does not fix, like length of stay. But the same fast-answer culture that makes it valuable is the culture most likely to over-trust it. The way to honour the legacy is not to slow the answer down. It is to make the answer worth trusting, with competency, quality control, correct sampling and result governance carried right up to the bedside. Speed was always the easy half of the promise. The discipline is the half that keeps a patient safe.

Sources and notes

The clinical impact figures here come from two UK randomised controlled trials of rapid molecular respiratory testing in acute hospital care, ResPOC and FluPOC, both run largely in and around Southampton and both using the same multiplex platform. The rapid antigen sensitivity figures are pooled estimates from a Cochrane review of studies that mostly predate the latest variants, so treat them as the order of magnitude rather than a fixed value. Length-of-stay findings differed between the two trials, which the text reflects deliberately: a real signal in one, none in the other, not a guarantee.

1. Brendish and colleagues. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC). Lancet Respiratory Medicine, 2017 (randomised controlled trial, n=720).
2. Clark and colleagues. Clinical impact of a routine, molecular, point-of-care, test-and-treat strategy for influenza in adults admitted to hospital (FluPOC). Lancet Respiratory Medicine, 2021 (randomised controlled trial, n=613; 202 with confirmed influenza).
3. Dinnes and colleagues. Rapid, point-of-care antigen tests for diagnosis of SARS-CoV-2 infection. Cochrane Database of Systematic Reviews, 2022.
4. Huang and colleagues. Multiplex PCR system for the rapid diagnosis of respiratory virus infection: systematic review and meta-analysis. Clinical Microbiology and Infection, 2018.
5. International Organization for Standardization. ISO 15189:2022, Medical laboratories, requirements for quality and competence, which brings point-of-care testing inside the laboratory quality standard and supersedes the withdrawn ISO 22870.
6. UK Health Security Agency. National flu and COVID-19 surveillance reports, which track co-circulating influenza, COVID-19 and RSV through the winter season.