Making rapid diagnostic tests available to primary care clinicians, without accompanying support, does not reduce antibiotic prescribing for respiratory infections. That is the central finding from two studies published recently in The Lancet Primary Care from the PRUDENCE programme – believed to be the first trial to evaluate a tailored multi-test strategy for respiratory infections in primary care, matching the test to the patient's symptoms, the season, and their COVID-19 status.
The tests themselves can give a GP new information within minutes – whether a sore throat is caused by group A streptococcus, whether the body's inflammatory response suggests a more serious infection, whether influenza is circulating. But the findings show that information alone is not enough to shift prescribing decisions, and they carry direct implications for how diagnostic technologies are implemented in community healthcare settings.
What the trial found
The PRUDENCE trial enrolled 2,639 patients across 13 European countries who presented with a cough or sore throat and whose clinician was considering prescribing antibiotics. Participants were randomly assigned to receive either usual care alone or usual care plus a tailored point-of-care testing strategy. Tests included C-reactive protein (CRP) blood tests to assess illness severity, rapid throat swabs for group A streptococcus, influenza A and B tests, or a combination.
Antibiotic prescribing rates were virtually identical between the two groups (45.7% in the testing group versus 47.1% in usual care), and patients recovered at the same rate – a median of four days to return to normal activities in both groups. Tests produced valid results in 95% of cases. There were no safety concerns.
Why implementation context matters
A companion qualitative study interviewed 56 patients and 33 clinicians in six countries to understand what happened when tests were introduced into routine consultations.
Tests proved most valuable where clinicians faced genuine diagnostic uncertainty – non-specific symptoms, ambiguous presentations, or situations where they had not yet committed to a prescribing decision. In these cases, test results shifted decisions in both directions.
But several factors limited impact. Clinicians frequently used results to confirm decisions already made rather than to revise them. When test findings conflicted with clinical experience – particularly in cases perceived as textbook bacterial infections – clinicians sometimes questioned the test's accuracy rather than reconsidering their assessment. Patient expectations, concerns about illness severity, and cultural norms around antibiotics also shaped outcomes in ways a test result alone could not address.
Professor Chris Butler, Co-director of the NIHR HRC Community Healthcare, and co-chief investigator of the trial, said: "Point-of-care tests have real potential, but our study shows that diagnostics on their own do not automatically change prescribing decisions. When clinicians are already leaning towards antibiotics, test results often reinforce that choice. To make a meaningful difference, rapid testing must be combined with clear guidance, clinician training, and support to manage patient expectations."
Implications for diagnostic implementation
These findings underscore something those working on health technology adoption in community settings know well: making a tool available is not the same as changing how clinicians use it. The impact of a diagnostic test depends not just on its analytical accuracy, but on how it fits into the consultation – when it is used, how results are interpreted, and whether clinicians feel confident acting on what it tells them.
The studies identify specific approaches that could make point-of-care testing more effective as part of antimicrobial stewardship: clinician training on evidence-based prescribing and test interpretation; communication skills support to help manage patient expectations; clear safety-netting guidance; and strategies to help clinicians navigate situations where test results challenge their initial clinical assessment.
Professor Sarah Tonkin-Crine, University of Oxford, joint senior author of the qualitative study, said:
"Clinicians across six very different countries and health systems described the same patterns – the primacy of clinical intuition, the pressure of patient expectations, the difficulty of acting on a test result that conflicts with your own assessment. Point-of-care testing needs to be part of a broader strategy, one that includes training, communication support, and clear guidance on how to act safely on test results."
The PRUDENCE trial was funded by the Innovative Medicines Initiative 2 Joint Undertaking, as part of the
VALUE-Dx consortium, and led from the Primary Care Clinical Trials Unit at Oxford's Nuffield Department of Primary Care Health Sciences. Oxford researchers involved in the programme receive funding from the NIHR HealthTech Research Centre in Community Healthcare.
Read the trial paper: Point-of-care testing strategy versus usual care to safely reduce antibiotic prescribing for acute respiratory tract infections in primary care (PRUDENCE) – van der Velden AW, Coenen S, Harper E et al. The Lancet Primary Care 2026.
Read the qualitative study: Clinician and patient experiences with point-of-care testing for acute respiratory infections in primary care: a qualitative process evaluation of the PRUDENCE trial – Wanat M, Hoste ME, Anastasaki M et al. The Lancet Primary Care 2026.
The PRUDENCE trial was led from the Primary Care Clinical Trials Unit at the University of Oxford's Nuffield Department of Primary Care Health Sciences, which is supported by the NIHR and its Clinical Research Network. Oxford researchers receive funding from the NIHR HealthTech Research Centre in Community Healthcare (The forerunner of the NIHr HRc Community helathcare) at Oxford Health NHS Foundation Trust and the NIHR Health Protection Research Unit in Healthcare Associated Infections and Antimicrobial Resistance (NIHR207397), a partnership between the UK Health Security Agency and the University of Oxford.