BRENNEN

Industries / Health / Healthcare / Urgent care

X-rays and lab tests

The X-rays and quick tests an urgent care center runs on site, so that a patient can leave the same visit with an answer.

7

Milestones achieved, from 1992 to 2025.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1992

    A rule for ankle X-rays

    Most injured ankles that are X-rayed are not broken. In two university hospital emergency departments, I. G. Stiell and colleagues examined 750 adults with ankle injuries and found a short set of checks (age 55 or over, tenderness at set points on the ankle bones, or being unable to bear weight) that caught all 70 significant ankle fractures and would have cut ankle X-ray series by 36.0%. Put into use at one hospital, the Ottawa ankle rules cut ankle X-rays by 28%, shortened visits from 116 to 80 minutes for patients without a fracture, and missed no fractures. A 2001 survey of 1,769 emergency physicians in five countries found that more than 70% of Canadian and British respondents used the rules often, against fewer than a third in the US, France and Spain.

    LessonPublishing a rule that saves X-rays does not by itself get it used.

    europepmc.orgeuropepmc.orgeuropepmc.org

    Credit

    I. G. Stiell and colleagues

  2. 2009

    A finger-prick test cuts antibiotics

    C-reactive protein (CRP) is a blood protein whose level rises with inflammation, and a point-of-care test measures it during the visit from a drop of blood. In a randomized trial in 20 Dutch general practices with 431 patients who had chest infections, doctors given the test prescribed antibiotics to 31% of patients, against 53% for doctors without it, and patients recovered as well and were as satisfied.

    europepmc.org

    Credit

    J. W. Cals, C. C. Butler and colleagues

  3. 2015

    Molecular flu tests leave labs

    Under US lab law (the Clinical Laboratory Improvement Amendments, CLIA), a "waived" test is one shown to be simple enough, with a low enough risk of wrong results, for staff without lab training to run. In January 2015 the FDA granted the first such waiver to a nucleic acid-based test, one that detects a virus's genetic material: Alere's i Influenza A & B, which gives results in as little as 15 minutes. It had been limited to moderate-complexity labs; the waiver let doctors' offices, emergency rooms and clinics run it. The FDA reviewed data from more than 500 patients.

    mlo-online.com

    Credit

    Alere; US Food and Drug Administration

  4. 2018

    Antibiotics for viral illness

    Researchers at the US Centers for Disease Control and Prevention compared antibiotic prescribing in 2014 insurance claims of people under 65. Antibiotics followed 39.0% of 2.7 million urgent care visits, 13.8% of emergency department visits and 7.1% of medical office visits. For respiratory diagnoses where antibiotics do not help, such as colds, bronchitis, asthma and influenza, urgent care centers prescribed them at 45.7% of visits, against 24.6% in emergency departments, 17.0% in medical offices and 14.4% in retail clinics.

    LessonHow often a drug is given for an illness it cannot treat depended on the setting, not only the illness.

    pmc.ncbi.nlm.nih.gov

    Credit

    Danielle Palms and colleagues, CDC

  5. 2018

    Software marks wrist fractures

    On 24 May 2018 the FDA allowed Imagen to market OsteoDetect, AI software that marks likely fractures of the distal radius, a common wrist fracture, on adult wrist X-rays. The FDA named urgent care among the settings where it may be used, as an aid that does not replace the clinician's reading. It was reviewed through the De Novo route, the FDA's path for a new type of low- to moderate-risk device. The supporting studies were the company's: 1,000 images checked against three hand surgeons, and 24 clinicians who read 200 cases better with it than without.

    fda.gov

    Credit

    Imagen Technologies; US FDA

  6. 2020

    A fast test misses cases

    Abbott's ID NOW COVID-19 test, a molecular test run beside the patient, was authorized under emergency powers in March 2020 and used at clinics and testing sites across the US. Researchers at a New York City academic institution found it negative in a third of samples that Cepheid's Xpert Xpress lab test found positive when swabs were stored in transport liquid, and in 45% when dry nasal swabs were used. On 14 May 2020 the FDA warned that negative results may need confirming in patients with symptoms and required Abbott to run follow-up studies. Abbott said the researchers had used the test in ways it was not designed for.

    LessonA fast result helps only if its negatives hold up in the way the test is actually used.

    europepmc.orgturnto10.com

    Credit

    A. Basu and colleagues; US FDA

  7. 2025

    Fracture AI allowed, with conditions

    On 14 January 2025 England's National Institute for Health and Care Excellence (NICE) said four AI tools may be used in NHS urgent care to help staff find fractures on X-rays while more evidence is gathered: BoneView (Gleamer), Rayvolve (AZmed), RBfracture (Radiobotics) and TechCare Alert (Milvue). A fifth, Qure.ai's qMSK, may be used only in research. NICE's committee found that studies typically showed higher sensitivity without lower specificity, for example emergency physicians' sensitivity rising from 61% to 74% with BoneView, but none of the 16 studies was done in UK urgent care and most looked back at old images. In October 2026 NICE extended the evidence period to about February 2027.

    LessonStudies of readers looking at old images were not enough for routine use; NICE asked for evidence gathered in urgent care itself.

    nice.org.uknice.org.uknice.org.uk

    Credit

    NICE

Watch for

  1. A finger-prick bacteria test

    FebriDx, made by Lumos Diagnostics, measures MxA (myxovirus resistance protein A) and CRP in fingerstick blood in about 10 minutes, to help tell bacterial from non-bacterial respiratory infections. The FDA granted it a CLIA waiver after a study run from December 2024 to July 2025 at nine US sites, including urgent care centers. PHASE Scientific, its US distributor, announced the waiver on 27 March 2026, said the test is in use at Wellstreet Urgent Care and expected to spread across Wellstreet's 140-plus locations, and said it can advance antibiotic stewardship.

    Achieved whenan outside study in urgent care centers shows fewer antibiotic prescriptions for viral illness when the test is used, with no rise in return visits or missed bacterial infections.

    clpmag.comaccessdata.fda.gov

    Claimed by

    PHASE Scientific and Lumos Diagnostics, March 2026

Open

  1. 01

    Quick tests run by newcomers

    An independent bench test of the waived tests urgent care centers run most, using only the positive and negative control swabs that makers sell for their kits, never a patient sample. Volunteers with no lab training, like the medical assistants many clinics hire, run several brands of rapid strep, flu and COVID-19 test from the package insert alone while two cameras film them. Each run is scored step by step against the insert: swab handling, timing, the reading window and how the result is written down.

    The brief
    Credit · unclaimed
  2. 02

    Waived tests with their numbers

    A public, machine-readable table of every FDA CLIA-waived test for the infections urgent care centers test for most: strep throat, flu, COVID-19, RSV, mononucleosis, urinary infections, and the newer bacterial-versus-viral blood tests. For each: maker, what it detects, sample type, time to result, and the accuracy the FDA's decision summary reports, with the number of samples, the comparison method and where the study was run, each figure linked to its page and dated.

    The brief
    Credit · unclaimed

Open 01 · the brief

Quick tests run by newcomers

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An independent bench test of the waived tests urgent care centers run most, using only the positive and negative control swabs that makers sell for their kits, never a patient sample. Volunteers with no lab training, like the medical assistants many clinics hire, run several brands of rapid strep, flu and COVID-19 test from the package insert alone while two cameras film them. Each run is scored step by step against the insert: swab handling, timing, the reading window and how the result is written down.

Why it is still open

CDC's 2005 review of surveys at US sites holding a CLIA Certificate of Waiver, the certificate that lets a site run only waived tests, raised concerns about practices that could lead to testing errors, which it put down partly to high staff turnover and inadequate training. The studies with untrained operators since then each test one maker's product and report accuracy and ease of use, not step errors counted brand against brand: Cepheid's Xpert Xpress Flu/RSV in 2022, judged easy to use from the operators' own feedback, and Roche's cobas SARS-CoV-2 & Influenza A/B test with 30 operators at 10 sites in 2025, a study written by Roche staff. A search found no multi-brand count of step errors.

ebi.ac.ukpmc.ncbi.nlm.nih.govpmc.ncbi.nlm.nih.gov

Built with what exists

The makers' own external control swabs; the FDA CLIA database to pick the tests; two fixed cameras; a step checklist written from each package insert; an AI video model to time-stamp the steps, with a person checking every score.

  1. First step

    Buy two brands of rapid strep A test with their control swabs, write a step checklist from each insert, and film five volunteers running each.

  2. Done looks like

    At least 8 waived tests from at least 5 makers, each run 5 times by each of 20 volunteers; for each test, the share of runs with at least one step error, the share of control results read or recorded wrongly, and the median hands-on time, with the scoring sheets and, with the volunteers' consent, the videos published.

Open 02 · the brief

Waived tests with their numbers

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public, machine-readable table of every FDA CLIA-waived test for the infections urgent care centers test for most: strep throat, flu, COVID-19, RSV, mononucleosis, urinary infections, and the newer bacterial-versus-viral blood tests. For each: maker, what it detects, sample type, time to result, and the accuracy the FDA's decision summary reports, with the number of samples, the comparison method and where the study was run, each figure linked to its page and dated.

Why it is still open

The FDA's CLIA database can be searched by test name, analyte, complexity and document number, but it carries no accuracy figures. Those sit in separate decision summaries, such as FebriDx's, which reports results from nine waived sites. A search found no table that gathers them.

accessdata.fda.govaccessdata.fda.gov

Built with what exists

The FDA CLIA database's download files; 510(k) and CLIA waiver decision summaries as PDFs; pdftotext; an AI model such as Claude to pull figures into a fixed form, with a person checking every figure against its page (Notes: "Seven more flagged").

  1. First step

    Download the CLIA database file, filter it to waived strep A and influenza tests, and fill in the rows for those two infections from their decision summaries.

  2. Done looks like

    At least 150 waived tests in the table, every figure linked to its source page, a second person's check of a random 20% with the error rate published, and monthly updates as the FDA posts new waivers.