Self-testing audiometry

Over recent months and years there has been an increase in the number of self-test audiometry systems launched and available on the market, and they seem to offer a cheap way to meet the screening requirements of the Noise Regs. But are they a good alternative to a hearing test being done by someone who knows the subject, be they in-house or an external provider?

It's worth highlighting upfront that as a company who provide workplace audiometry, we do have a horse in this race and self-testing is up against us as a company undertaking hearing tests with a qualified technician. But, if they are a good alternative then that wouldn't stop me saying so, I would just change our provision to go down the self-test route as it would be a lot cheaper for us to provide.

What is self-testing in audiometry?

There are a couple of routes to this. 

  1. Online hearing tests offered by many websites.
  2. Self-testing using audiometry equipment set up on the site for this purpose, such as WorkScreen.

In both situations, the individual operates the test themselves and then receives an automated result based on how well or poorly they have performed.

Online hearing tests

These fall into two types of test, either listening to sounds and speech against a series of background noises, or the more traditional 'here are a load of beeps and press a key when you can or can't hear it'.

Speech against background noise

Despite their non-traditional approach, this first type are mildly helpful at distinguishing problems with hearing as one of the key signs of noise induced hearing loss (NIHL) is increasing difficulty in understanding and following speech in areas with a lot of background noise. However, this is so subjective that it is of no use at all in workplace screening.

It sounds flippant but there is a huge difference between 'listening' and 'hearing'. I am sure many persons of the lady persuasion will agree that their gentlemanly companions are frequently capable of 'hearing' but most definitely are not 'listening'. In tests like this, a big variation can be down to how practiced someone is at actually listening to speech in areas with a lot of background noise. If they have something resembling a life and spend their Saturday evenings in loud pubs and restaurants will be naturally more practiced and better at this than someone who thinks a racy night is a glass of wine and an episode of Midsomer Murders.

There are a lot of online providers offering these, including Beltone's online hearing test, just as a random example. This gives some speech against noise and the point of it is largely to drive traffic to the physical hearing testing centres for a paid-for hearing test and, from the supplier's perspective, hopefully some hearing aids as hey, everyone can always hear a bit better so why not spend some money. Nothing wrong at all with that facility as a sales mechanism, but it is not a substitute for a proper audiometric screening test. Belton are not alone in this and there are many many others out there offering very similar tests, such as Amplifon and Boots. Again, these are not pretending to be proper screening tests and often say things like 'check to see if you need a hearing test' on them.

Online frequency-testing sites

This can be better but is again riddled with too many problems to be of any use in workplace screening. For example, this is an online test by which asks the users to first calibrate the volume by rubbing their hands in front of their nose - I can't even go there, they may as well click their heels together three times, go to Oz and ask the Wicked Witch to set the right volume for them. The test then asks the user to press individual frequencies and say whether they can hear them or not. There is another problem here in that when people are presented with a sound which they may or may not have heard and is right on the limit of their hearing, then if they know the sound actually was there they will err on that side and say yes, whereas in a hearing test we are looking for the quietest sound they can reliably hear, not guess that they can hear. That's a big difference.

This type of test is OK as a fun tool, and to be fair is not presented as a means of doing workplace screening tests at all by, but fun is about as far as it goes for reliability. 

Physical self-test systems

There are also self-test options out there which involve physical product being shipped to the employer and then employees conduct their own tests. The advantage over the online-only style is that at least you are getting a calibrated system (or should be anyway) so the test has far fewer variables than someone plugging their own headphones into their own computer.

Again, without highlighting one for any specific reason, one example system is WorkScreen where a kit is sent to the employer containing a tablet and some headphones, with the attendee then simply following the instructions on the machine.

Screen Shot 2018-05-21 at 08.28.01.png

WorkScreen make some rather bold claims for the system, including 'reducing the administrative burden', that the testing is 'HSE compliant', although other pages then change that slightly to a 'report prepared to HSE guidelines' and that it means less lost work time for the attendees. The boldest claim though is the one about it being 'more accurate than traditional systems' and that's the one which really needs examining closely. 

British Standard for audiometry Cover.png

Firstly, I am a bit suspicious of any company which says their hearing test is 'HSE compliant', when 'HSE compliant' is not the standard for hearing tests in the workplace. The standard which hearing tests should be following is the British Standard for Pure Tone Audiometry - BS EN ISO 8253-1.

That probably sounds a bit of a fudge but there is a big difference between HSE guidelines and the British Standard. The British Standard says how a hearing test should be undertaken to get a valid reliable result - this is the way the test should physically be done. The HSE Guidelines in L108 say how that data should then be interpreted for work. Other than some vague top-level stuff about frequencies, the HSE do not actually say how a hearing test should be done, just what you then do with the data.

There we have the general claim that this self-test system is better and more accurate than traditional audiometry systems so let's look at some elements more closely.

Verification checks.

There is a requirement to do a verification test at the start of every audiometry session. The self-test system may or may not do this, I don't know, although I doubt it as there is no mechanism for this in the kit as it is the headphones which have to be tested and not just the audiometer itself, which means the test is already not compliant with the required standards for a hearing test.

Ear simulators.png

It is possible to automate the verification check on any PC controlled audiometer (which almost all are these days) and I have extensive experience of this. You place the headphones on a special bit of kit which responds to the tones and sends a signal back to the audiometer to confirm the tones are still the right volume. However, they are awful things and despite spending good money on them in the past they ended up sat in a cupboard for years. The reason was that the verification check is looking to make sure the audiometer is responding as it should, but also helps identify any other issues in the complete system. Over the years I had loads of issues where the static automated self-test said everything was OK, but then when you put the headphones on a real person who is moving about problems with static interference or cable connections became apparent. This meant that even if the system does have an automatic verification check built into it, the result are still rather suspect.

Review of the health questionnaire

A big part of audiometry is the pre-test health questionnaire which reviews the hearing health of the attendee before they do the actual hearing test. No two people are the same here - they may answer yes or no to the same questions, but the discussion about their hearing health is absolutely critical to any findings of the hearing test. Why I hear you ask? Firstly because their answers may change the physical test itself, and secondly because you cannot diagnose noise induced hearing loss from an audiogram without a good medical history. Without that medical history being reviewed by someone who knows what they are doing then the audiogram becomes fairly useless. It is way more complex than simple yes/no options.

As an example of their answers changing the test - someone could say they have no hearing in one ear, so in the test you may do a couple of manual verification tones in that ear to confirm this, then just test the one good ear properly. If you don't and the system does a full test it will record a unilateral hearing loss in one ear and throw up a referral. Even worse, what will happen is the tones are played so loudly in the bad ear that the good ear starts to pick them up and the attendee presses the button, recording a result for an ear which is doing little more than keeping their glasses straight.

Otoscopic examination of the ear

The basic requirement for a hearing test requires otoscopic examination of the ears before the headphones go on - the bit where you look down the ears with a bright light thing. You need to know how healthy the ears look before doing the test. Without that you have no idea whether there is any infection, whether the ear drum has any scarring, or how much wax there is in there which may be impacting the result. Without that the hearing test is not compliant and is useless.

Who puts the headphones on?

In a hearing test, as required by the British Standard, you never ever let the attendee put the headphones on themselves - you always put the audiometry headphones on for them. Attendees put them on so they feel comfy whereas as technician puts them on so the speaker unit is placed properly over the ear canal. This sounds trivial but can be the difference between a pass and a fail. How big a difference does this make? These are two results for the same person, one where they put the headphones on themselves and one where the audiometry technician put them on.

 Audiometry headphones placed on head themselves

Audiometry headphones placed on head themselves

 Audiometry headphones placed by the technician

Audiometry headphones placed by the technician

As you can see, when the attendee put the headphones on themselves, they did it so it felt comfortable but the right ear (the red one) was showing increasing losses with increasing frequencies. Once the headphones were adjusted by the technician to make sure they were seated correctly the right ear was back to where it should be. This was the difference between the right ear being classed as OK or classed as having losses which are starting to be of a concern.

It is vital that people do not put the headphones on themselves in a hearing test and this is a fundamental requirement of the standard for pure tone audiometry.

Background noise levels during the audiometry.

The British Standard also places restrictions on the background noise levels in which a hearing test should be undertaken. Some audiometers have a function to measure this built into them which is nice, but the overall total volume is only part of the story and just as important is the type of noise around. It's no good if it is quiet most of the time but then occasionally people can be heard talking or telephones ringing as these are very effective at masking the tones being tested. A technician will see this and work around it but a machine will not. 

How certain was the attendee?

As well as the result the individual obtains what is equally important is how they obtained that result - were they nice and sure and fast or where they hesitant, more unsure and guessing a little more. A technician will pick up on this whereas the automated test will not. If an attendee is looking uncertain, the audiometry technician may vary things like the tone length, or the gap between them to give a little more response time for the attendee, or even change to using a pulsing tone which can be helpful where the attendee has a degree of tinnitus - again, the automated systems will not do this. 

Attendee understanding of the result

The WorkScreen setup says it lets the 'user instantly see their results and also receive a copy of their report by email'. I'm not entirely sure who the 'user' is in that, the client company or the individual undertaking the hearing test. But the result of a hearing test is not the most straight-forward thing to understand so having a technician explain their result to them and how it relates to their lifestyle and health history is vital for their understanding of what it means. We do that and give them a copy of the result there and then, with them being able to ask the technician anything they want or need further explanation on, which can't be done with a self-testing system.

All these sound a bit trivial, but they really aren't. I've been doing hearing tests at work for many many years, and for the last decade owned a consultancy which did over 20,000 of them every year. Those years gave me a lot of experience with both criminal investigations into companies by the HSE for non-compliance with the Noise Regs, and more frequently companies and their insurers getting involved in the defence of spurious claims. In every one of those the first things asked for were proof of competence of the technician who did the test, calibration records, daily verification check records, what the noise levels were in the test environment and what subsequent advice was given to the individual concerned. Without these the tests are useless.  

Claims about it 'reducing the administrative burden'.

If we are asked to do a load of hearing tests for a client, we ask simply for the numbers of people per shift who need to be tested and draw up a schedule of appointments based on that. If the client gives us the attendee names then we will even add the names to the appointments list so all the client has to do is let everyone see that. After the testing we give the client a report along with a list of all the attendee names and when they should next have a test. When those tests are coming due we let the client know and start the process again. There isn't much 'administrative burden' for the client in that. With the self-test system all the organisation has to be done by the client - that is increasing the administrative burden, not reducing it.


Yes, I have a vested interest in this as I provide workplace hearing tests, but if these alternative systems were anywhere near as good as they claim to be then my business model would shift overnight. Why have the expense of building and maintaining mobile clinics and the time and cost of trekking around the country with them if I could just post a system to a client and charge a little bit less but still maintain my margin? If these systems were good I could pay an app developer less than a single mobile unit costs to build, get them to develop a software-based system for me and in that case I'd switch the business over to this model in a heartbeat. But these self-test systems just aren't good enough.

No fence sitting here - these tests are no good for workplace screening, are not compliant with the standards required for workplace audiometry, and should not be used for noise at work health screening.

Relevant links