Checking an AI hearing test against the usual test for hearing loss

AI summary

From the trial's eligibility criteria. Machine-read from the registry text on this page, so it can be wrong. The registry text below and the trial team decide who can take part.

The trial is looking for adults with sensorineural hearing loss (hearing loss from the inner ear or hearing nerve), with or without tinnitus (ringing or buzzing in the ears). It compares an AI-driven hearing test with the standard hearing test.

Age

18 years or over

Sex

All sexes can join

Disease stage

Diagnosed sensorineural hearing loss, with or without tinnitus at the same time

Main reasons for not taking part

• Diagnosed with, or reporting, hyperacusis (unusual sensitivity to everyday sounds)
• Active ear infection
• Recent ear surgery
• Not enough English to read, understand and complete questionnaires
• Not able to give informed consent

The registry criteria, word for word

What the trial wants to find out

Background and study aims Hearing problems are becoming more common. In the UK around 18 million adults have hearing loss or hear ringing or buzzing in their ears, which is called tinnitus. Many people wait a long time to be seen, partly because there are not enough audiologists - the specialists who test hearing and treat hearing problems. The standard hearing test is called pure-tone audiometry. An audiologist plays quiet sounds at different pitches through headphones and the person presses a button each time they hear one. The test is accurate, but it takes 20 to 30 minutes to complete, and needs a trained audiologist for all of that time. For people with tinnitus and comorbidity it can be especially tiring and slow. The research team has built a version of this test that uses artificial intelligence, called AI-PTA. Instead of following a fixed pattern, it chooses each sound based on the responses already given, which should mean fewer sound presentations are needed to find the same result. The audiologist stays with the person throughout and remains in charge of the test. This study asks three questions. 1. Does AI-PTA give the same hearing results as the standard test? 2. Is it quicker? 3. And do patients and staff find it acceptable? Who can participate? Adults aged 18 or over who have been diagnosed with sensorineural hearing loss, either with or without tinnitus, and who are due to attend the Royal National ENT \& Eastman Dental Hospital in London. Participants need enough English to read and complete questionnaires and to be able to give informed consent. People who have or report hyperacusis (an unusual sensitivity to everyday sounds), an active ear infection, or recent ear surgery cannot take part. What does the study involve? People who are interested are sent information about two weeks before a hospital appointment they already have. Those who decide to take part come in about 30 minutes early, meet a research audiologist, discuss the study, and if happy to take part, sign a consent form. They fill in a short questionnaire about their hearing health. They then have two hearing tests at that visit: the standard test, which is part of their normal care, and the AI-PTA test. Half of the participants have the standard test first and half have AI-PTA first, so that the order does not affect the results. They come back within a week for a shorter appointment, about 40 minutes, to repeat the AI-PTA test. This shows whether it gives the same response twice. They also fill in a questionnaire about what they think of using artificial intelligence in healthcare. That is the end of taking part, though participants are also invited to join an optional group discussion about their opinions on AI in healthcare later on. What are the possible benefits and risks of participating? There is no direct benefit to taking part. The results will help decide whether tests like this can shorten waiting times for hearing assessments in the NHS. The risks are small. Both tests use the same clinical equipment at the same safe sound levels that ordinary hearing tests use, and a qualified clinical audiologist is present throughout and can stop at any point. The main cost to participants is time: about 45 minutes extra at the first appointment and a second visit of about 40 minutes. Some people find hearing tests tiring, and people with tinnitus sometimes notice their tinnitus more during or after a hearing test. Taking part is voluntary, and choosing not to take part, or stopping at any time, will not affect anyone's care. Where is the study run from? Royal National ENT and Eastman Dental Hospitals in London which is part of University College London Hospitals NHS Foundation Trust, and University College London. When is the study starting and how long is it expected to run for? The study was approved in August 2026 and the first participant took part in September 2026. It is expected to run for 12 months, finishing in August 2027. Who is funding the study? The British Society of Otology (part of ENT UK), and the National Institute for Health and Care Research University College London Hospitals Biomedical Research Centre. Who is the main contact? Dr Roulla Katiri,

The 2 groups in the trial

• Standard PTA first

  Participants randomly allocated to receive Standard PTA before Intelligent PTA at visit 1, and the reverse order at visit 2. Randomisation is stratified by tinnitus status, so this arm contains participants from both clinical groups.

  Treatment: Intelligent PTA and Standard pure-tone audiometry (BSA modified Hughson-Westlake)

• AI-PTA first

  Participants randomly allocated to receive Intelligent PTA before Standard PTA at visit 1, and the reverse order at visit 2. Randomisation is stratified by tinnitus status, so this arm contains participants from both clinical groups.

  Treatment: Intelligent PTA and Standard pure-tone audiometry (BSA modified Hughson-Westlake)